Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: simpson36 on July 21, 2012, 08:57:08 AM
-
Every time I have seen yet another 'power drawbar' thread, it has been here in the general section, so I'm starting here. If it needs to be moved, so be it.
I've been talking about this for a long time but there are a number of pices to the puzzle that have to come together and finally they are all in the sand box so its time to play.
Requisites were BT30, power drawbar (pneumatic only), Servo driven spindle, access to certain ATC parts. Ths big holdup was the Servo driven spindle, but I finally have that now. The new servo spindle drive is demonstrated in this 'intro to parametrics' video: http://www.youtube.com/watch?v=Cu61oBY5-rw (http://www.youtube.com/watch?v=Cu61oBY5-rw)
I have had the pneumatic cylinder and some other stuff for quite a while from when I was contemplating doing an R8 power draw bar. I abandoned that idea after a lot of research. I recently found a source for the 'fingers' that go on an ATC and hold the BT30 adapters.
A stumbling block was the $500 stud gripper for the BT30. There are two types of gripper that I know of; finger type and ball type. Either needs to be hardened and properly tempered to perform safely. I know how to do that, but I no longer had a furnace (sold with the prototyping shop many moons ago). So I have aquirde a furnace. I still have to install the new automated controls, but the 50A 220V wiring is completed.
Meanwhile, I havecompleted the first amd most essential component, the above mentioned stud gripper. I went with the ball type. It works perfectly so all I need to do is turn doen the sleeve and harden the components. I have successfully completed my first scratch built spindle for my 4th axis, so I will use the same techniwues and new tools to make the mill spindle. Compared to making the stud gripper, everything alse should be fairly simple.
(http://www.thecubestudio.com/4thAxisSpindleProject/BT30GripPartsVertWEB.jpg)
The center piece (hold the balls) is S7 tool steel while the 'sleeve' is A2. The sleeve OD is 'mill finish' as it has not been machined yet.
(http://www.thecubestudio.com/4thAxisSpindleProject/BT30GripPartsHorizWEB.jpg)
THe next two photos show the gripped in open and closed positions. I have colored the bals red becuase otherwise it was very difficult to see what was actually happening. This was the first round prototype and it works perfectly 8) so these actual parts will be hardened and used in the spindle.
(http://www.thecubestudio.com/4thAxisSpindleProject/BT30GripRedBallWEB.jpg)
(http://www.thecubestudio.com/4thAxisSpindleProject/BT30GripRedBall02WEB.jpg)
-
Simpson,
Beautiful work, as always! I will be very interested to see how this progresses, as I'm currently building an ATC for my mill - TTS/R8 for now, but still plan to switch to ISO30 down the road.
Regards,
Ray L.
-
Thanks Ray,
This will probably be the last post for a while on the BT30 spindle because I have other work to complete before I can re-focus on this project. The gripper is a central character in this play so really making and testing the gripper mechanism could be considered part of the feasibility study. Put in simple terms; no gripper - no spindle. So I wanted to accomplish a working gripper before I dedicate the time and resources to a spindle/drawbar project.
Below are photos of the completed gripper parts. After heat treating the sleeve was tempered only slightly for max hardness. The ball holder was tempered to a typical 'cold work' spec for S7 After heat treat, the parts were pickled and blasted to remove scale. Small S7 parts and A2 parts are air quenched and typically there is only minor distortion. The ball holder was fine and the threads still worked perfectly, so it just got a black oxide treatment and was completed
The sleeve, as expected, had minor distortion. The ID does not have to be perfectly concentric and the ball holder still slid nicely in the sleeve, so the ID was simply re polished for smooth action and low wear on the balls. However, again as expected, the OD had to be ground to bring it to size and concentricity for pressing into the spindle nose.
My research turned up a figure of around 600lbs drawbar tension for BT30, so I will design the mechanism for 600 lb pull and 1000lb release on 100psi shop air. The cylinder is rated to 200psi, so there will be plenty of headroom. I will test the gripper at 2,500lb pull so that in case I have to go up to 1,000 pull on the holding force, I will still have a 2.5x test.
Making the new mill spindle will be simple compared to the new 4th axis spindle that I recently completed. There are decisions to be made about cartridge vs open frame and of course the type of bearings.
OK, on to the photos. First is just a shot of the completed gripper pieces:
(http://www.thecubestudio.com/4thAxisSpindleProject/BT30GripPartsFinalWEB.jpg)
This is a shot of the setup for grinding the OD of the sleeve. The sleeve is bolted to the same custom made mandrel for both machining and grinding the OD.
(http://www.thecubestudio.com/4thAxisSpindleProject/BT30GripSleeveGrindWEB.jpg)
This next shot is the 4th axis spindle getting 'cored', but a similar operation, albeit not nearly as big a hole, will be done on the new mill spindle. This is the toughext part of the job. Once that hole is thru the piece, the rest is relatively easy. I was only barely able to get this hole made. The 4th axis motor stalled several times and the stress twisted the mill column. The final drive on the current 4th axis is 5.76:1 and the column is bolted together and not fully boxed (this is just a temporary mill). To resolve, the new 4th axis final drive is 12:1 and the column (now realigned) will be welded and boxed in. This will defeat my 'easy shipping' plan for this mill, but there are no other options.
(http://www.thecubestudio.com/4thAxisSpindleProject/DeepDrill02WEB.jpg)
-
Simpson,
Beautiful work, again. FWIW - I recall some time back seeing a spec of 1300# for the drawbar tension on ISO30.
Not sure if you're aware of the power drawbar I did a while back. I used a small stepper motor, driving through a high-ratio planetary gearbox. Using current limiting on the stepper driver, I can control both tightening and loosening torque. I'm currently running TTS, but this design let's me swap TTS or R8 tools, simple based on how many turns I turn the drawbar. And, of course, it will work if/when I switch to a 30-taper spindle. Because of the high ratio, with a 400 oz-in stepper, I can get up to 75 ft-lbs drawbar torque - far more than needed. It has worked absolutely perfectly from day one. I'm now working on a 10-tool carousel ATC.
Regards,
Ray L.
-
Beautiful work, again. FWIW - I recall some time back seeing a spec of 1300# for the drawbar tension on ISO30.
Thanks for the kudo, Ray. One thing I have learned in researching power drawbars; if you want another spec on tension, just wait 30 seconds . . LOL! I can recall several people trying to do R8 drawbars with about 1,000 and I had not done any homework at that point, but it *seemed* pretty weak to me. I had one 4th axis buyer who has a power drawbar that he purchased somewhere and he told me that it would not hold the TTS tools. He had no idea how much force it used so I had him measure on of the Belleville springs and tell me the arrangement of the springs. . . . turned out he had something like 500 - 600 lbs . . . mystery solved.
There are two specs on BT30; above and below a certain RPM (6k as I recall, but I'd have to pull my notes). I cannot speak to any other taper standard, but it seems logical that it would go up as the holders got larger. The reason the tension is so high on R8 and TTS is that they rely on friction to transfer torque. BT30 has drive dogs . . which are something of a mixed blessing
Not sure if you're aware of the power drawbar I did a while back. I used a small stepper motor, driving through a high-ratio planetary gearbox. Using current limiting on the stepper driver, I can control both tightening and loosening torque. I'm currently running TTS, but this design let's me swap TTS or R8 tools, simple based on how many turns I turn the drawbar. And, of course, it will work if/when I switch to a 30-taper spindle. Because of the high ratio, with a 400 oz-in stepper, I can get up to 75 ft-lbs drawbar torque - far more than needed. It has worked absolutely perfectly from day one. I'm now working on a 10-tool carousel ATC.
Yes, I am aware of your 'silk purse'. I almost folowed you down the R8 road on another project that seems to me was quite a while prior to your current working model. It did not seem doable to me because I *think* you were looking at a single pneumatic cylinder at the time. Not positive on that . . . just a recollection. There a lot of bodies along that roadside. It would appear that you have now taken a novel approach to an old problem and succeeded where many have failed. Kudos to you on that!
Couple of FWIW back at you; pneumatic R8 drawbar IS doable, but by my calculations say it would require an articulated stacked cylinder. I have a box with all of the materials to make just such an animal; but ultimately I judged it to be a poor investment not because it would not work, but because it would not be compatible with any reasonable ATC design. I'll be able to use most of that stuff now when I revisit this BT30 project. Second FWIW; I am running both R8 and BT30 in the same spindle . . I did not realize this was impossible until it was too late . . just saying.
-
Hey Steve. Interesting new project to follow. Always nice to see your work and the way you present it.
Dan
P.S.
If you could resize a bit your photos it would make it easier to watch them as they are too wide to fit on the screen.
-
Simpson,
The spec I saw was for ISO30, which does not have drive dogs, so probably 600 pounds would make sense with the dogs. Mine will be ISO30, as that's the only spindle available for my machine.
Actually, I think I was the first to actually quantify the drawbar tension required for TTS, and wrote it up here about 3 years ago. At the time, even Tormach did not know, and one of their engineers asked me to keep them posted on my findings. They have since arrived at the same number I did - about 2500 pounds tension, or 25 ft-lbs drawbar torque (for a properly-lubed drawbar). R8 can be somewhat less, due to the taper. I did make a single-cylinder air drawbar with mechanical advantage, aid it kinda worked, but was far from ideal, so I scrapped it. It can easily be done with a stacked, large diameter air cylinder - several people have done it on CNCZone over the last year or so, using triple-stacked 3-4" cylinders - but I always leaned towards and air-over-hydraulic system, for compactness. Then, I came up with the stepper design, and never looked back. It's handled the most aggressive cuts I could do on TTS with no pullout ever. And being able to do TTS, ISO30 or R8 with the same PDB is a nice bonus.
Regards,
Ray L.
-
IF memory serves me I remember 30 taper drawbar pressures ranging from from 900-1300 lbs depending on manf. and useage. The higher pressure were normally on machines that did a LOT of heavy hogging all day long.
You do realize that r8 does NOT have the same taper as 30 taper ??? Just checking(;-)
(;-) TP
-
The spec I saw was for ISO30, which does not have drive dogs, so probably 600 pounds would make sense with the dogs. Mine will be ISO30, as that's the only spindle available for my machine.
Perhaps ISO30 (which I know nothing about) and BT30 are similar enough that you could still use BT30 adapters by just adding drive dogs to the spindle. Just thinking out loud. I see lots of stuff for BT30 and I can't say the same for ISO30 . . . of course, I never look for ISO30, so that may have something to do with it . . . :-[
-
IF memory serves me I remember 30 taper drawbar pressures ranging from from 900-1300 lbs depending on manf. and useage. The higher pressure were normally on machines that did a LOT of heavy hogging all day long.
I think BT30 is around 1000 if you run over a certain RPM, but I would have to drag out my notes before I would argue the point . . or I could just wait another 30 seconds :P BT30 adapters are often balanced to some outrageous RPM, so I suspect running under 6K would put most BT30 guys to sleep. I have 7HP peak 3000RPM motor, and while I would love to run 10k, a 3.x overdrive is practically trying to make a speeder using belts . . maybe not such a good idea. My bearing choices would also narrow considerably because of he seals. My motor is 140V rated and it is dining on 180V so I do get some extra revs. I will be making all new pulleys and still have a dual range setup, but I think a 2x overdrive will be adequate. I will need to move to a 1" wide belt on the underdrive . . . . do not ask.
I already have a die grinder mount for really fast stuff like routing PCB and so on. The die grinder runs 32k as I recall. Really I want a turbine, but I do not have enough air.
You do realize that r8 does NOT have the same taper as 30 taper ??? Just checking(;-)
Yes, I am aware of the . . . minuscule . . . difference. I re-ground the short R8 taper to BT30. The BT30 does not get full support with the arrangement so it is not 'ideal' strictly speaking . . . but neither is the TTS arrangement, in my view.
My turn; you do realize that R8 are spring collets?? 10 brownie points if you know why that makes a difference. Incidentally, I'm not recommending anything . . I'm just saying ;).
-
3000 @7hp is really not a good combination, at 3000 rpm you are still limited to a certain chip load so the extra HP does not add much to the combination. You tend to just generate more heat that you cannot get rid of easily.
IF you set up the spindle with low drawbar pressure AND HIGH torque loads on the cutter you will probably RATTLE the tool holder in the spindle under heavy loads
Now gear it 2:1 then at 6000 rpm you are getting somewhere. BUT even that low your bearings, Sindle bearing alignment, grease and preload better be spot on(;-)
R8 collets now that is an exercise in anger management(;-)
Just a thought, (;- )TP
-
Actually, I think I was the first to actually quantify the drawbar tension required for TTS, and wrote it up here about 3 years ago.
Yes, I remember those threads. It was a bit like discussing religion. Required more patience that I can muster.
Then, I came up with the stepper design, and never looked back. It's handled the most aggressive cuts I could do on TTS with no pullout ever. And being able to do TTS, ISO30 or R8 with the same PDB is a nice bonus.
I have a few questions about your machine if you don't mind sharing on a forum.
*How do you accomplish releasing the collets? On my R8 design, I provided an adjustable gap to effect an 'impact' load to release the collet.
*How do you prevent the collets from rotating in the spindle nose?
*Do you turn the drawbar itself or a nut at the top of the drawbar? With TTS, I can conceive of either method, but with R8, I would imagine there is only one choice.
*How do you control the torque on the stepper. Are there drives with that capability, or are you controlling the supply to the drive?
-
Just a thought, (;- )TP
Your thoughts are always interesting and welcome. What I am hearing is that there are a lot of practical examples to consider. It all boils down to 'where the rubber hits the road' or in context; 'where the tool hits the workpiece'. Everything before and after is academic.
7HP is peak, so I have to design for that amount of power, even though it is momentary. Chiploads would be calculated on continuous power, which is considerably less. The motor is rated in the neighborhood of 7 amp cont.
The question of drawbar tension is also academic with regard to 'how much is enough', so why not just go high and 'play it safe'? Once again there are practical considerations, in this mechanism there is a 'jesus nut' which is the pull stud. These little guys are stupidly expensive in spite of the tiny bit of material and very simple geometry. The cost is in the testing of each part. Fatigue life is probably too big a topic for a hobby forum, but suffice to say that any metal part that is repeatedly stressed and then released has a fatigue life. Airframes are retired after a certain number of hours for that reason and so *should* pull studs. The relationship between the amount of stress and fatigue life is not linear.
An analogy that most reader can identify with is fuel consumption. If you drive 60 instead of 50, you might expect to use 20% more fuel for the 20% increase in velocity, but that's not the case. And the delta in fuel consumption between 50 and 60 is not the same as between 90 and 100, even though the delta in speed is the same. Travelling at very high speed has benefits and costs associated with those benefits. There is a lot of focus on 'too low' drawbar tention, but there is also a 'too high' drawbar tension which ultimately can do much more harm.
Just a thought back at ya, (;- ) SS
-
*How do you accomplish releasing the collets? On my R8 design, I provided an adjustable gap to effect an 'impact' load to release the collet.
The entire PDB mechanism moves up and down. When activated, it is pushed down by air cylinders, applying a few hundred pounds of down-force, which is enough to pop the collet free, provided the taper is kept well-lubricated. Once the taper is popped, the TTS holder simply falls out.
*How do you prevent the collets from rotating in the spindle nose?
With R8, the alignment pin in the spindle takes care of that, though I suspect just the friction on the taper would do it as well. I'll find out someday when I get around to removing the pin.
*Do you turn the drawbar itself or a nut at the top of the drawbar? With TTS, I can conceive of either method, but with R8, I would imagine there is only one choice.
I'm using the stock drawbar, with a 3/4" 12-point socket on the PDB spindle.
*How do you control the torque on the stepper. Are there drives with that capability, or are you controlling the supply to the drive?
Any decent stepper drive will have current limiting that is controlled by the value of a resistor. Reduce current, and torque reduces more or less proportionally. I set the current limit while tightening to a value that results in 25 ft-lbs drawbar torque, which is about 2500 pounds drawbar tension, when the drawbar and taper are lubricated (with a thin coat of anti-sieze), and the TTS shank and collet bore are clean and dry. When loosening, I use full current, which give about 75 ft-lbs torque with the 400 oz-in stepper I currently have. The current limit resistor is switched by a relay.
Regards,
Ray L.
-
You can never have TOO much drawbar tension(;-). But you do have to consider design parameters. A real pull stud(stud only) is designed for up in the neighborhood of 100,000 lbs of drawbar pressure before failure. The upper end level depends on the material and design used before the stud stretches or the drawbar stretches and releases the pressure and the tool holder.
The other side of the design is releasing the high drawbar pressure. Also fitting in the Very TALL spring stack as high pressure calls for a very tall stack to get pressure and overall travel.
The overal weakness of MOST power drawbar systems is the SPRINGS that retain the pressure. The more you use them the sooner they loose tension or Break springs. Not unusual to change out stacks twice a year on 24/7 machines .
I would personally shoot for the 900-1000lbs level.
You mileage may vary, (;-) TP
-
The current limit resistor is switched by a relay.
I am aware of the resistor set power limits, but a couple of your descriptions lead me to believe you were changing the force in real time. Some of the industrial drives have commandable parameters that can be set 'on-the-fly', but I recall that you were using Gecko (or equiv), which do not have this capability, so I suspected that you had something more up your sleeve.
Relay switched resistors . . very clever indeed. 8)
-
You can never have TOO much drawbar tension(;-). But you do have to consider design parameters. A real pull stud(stud only) is designed for up in the neighborhood of 100,000 lbs of drawbar pressure before failure. The upper end level depends on the material and design used before the stud stretches or the drawbar stretches and releases the pressure and the tool holder.
An interesting statement considering the range of sizes these things have. My calculator has you off by an order of magnitude for the cross section of a BT30 stud . . . and that would be with a zero safety factor. This is one for the 'agree to disagree' file.
The other side of the design is releasing the high drawbar pressure. Also fitting in the Very TALL spring stack as high pressure calls for a very tall stack to get pressure and overall travel.
The overal weakness of MOST power drawbar systems is the SPRINGS that retain the pressure. The more you use them the sooner they loose tension or Break springs. Not unusual to change out stacks twice a year on 24/7 machines .
Working with only 'Catalog knowledge' of Belleville springs, and having no resources to draw from, I am in WAG mode for this aspect of the design. I am aware of the tall stacks of springs. I was thinking the travel could be had by simply using a few large diameter springs at the top of the spindle rather than a long stack of small diameter springs inside the spindle. However, I have only seen the guts of a couple of spindles, and so far none have taken this approach. Have to wonder why. Other than redundancy, it is hard to find a major advantage to the tall stack. It there a hole in this logic?
-
When I tried to do a Belleville system for my knee mill, I could not find springs with characteristics that would allow me to get the force and travel I needed without a fairly long stack (about 30 springs, about 3", IIRC), even when I went to pretty large diameter (~1.75", IIRC). And, unless you're only running very low speed, even a fairly short stack will need to be contained within some kind of housing, to prevent it from whipping, with ruinous results, when the spindle is on. That housing will then need to be balanced. Perhaps you can find more suitable springs, but I could not. I also found such a short stack of springs, for all practical purposes, could not be compressed beyond their rated force. The friction goes up exponentially beyond that point, making any additional travel come at the expense of extremely high incremental force for a very small incremental travel, even when the springs are well greased. Using a very tall stack allows the response to be nearly linear, as each spring need only move a few thou. VMCs seem to use stacks of about 150 springs, I assume for this very reason.
Regards,
Ray L.
-
One too many zeros I am afraid. Ray has it covered on spring stacks. To get pressure AND needed travel it takes a STACK of bevels to get it all done.
(;-) TP
-
Thanks again for the replies.
I had not considered the disc-to-disc friction on the Bellevilles . . . certainly a consideration.
Although in the final analysis it may be the only workable solution, the whole belleville stack idea just does not appeal to me. Is there some reason that, considering the small size of the BT30, one could not use a valve spring. Not from Grandma's Chevy, but a racing spring or a diesel spring. I have not looked at diesel springs yet, but Racing Valve springs are available at 750lb rate with over 1,100lb open pressure and 1" travel. These are meant for about 350lb seat pressure, but the gripper needs only about .23" movement to release so that would leave .77" compression on the spring for about 750*.77 + 350 or approx 930lbs constant pressure and 1,100lbs max at full release. This seems doable with a pocket in the top of the spindle to center the spring and just using the racing spring retainer that would normally be used with the spring.
The relatively small diameter and mass of the coil spring being held at center I do not *think* would pose a balance problem, but the spindle would be balanced anyway, so that's not a concern. At that high compression, the coil would loose tension over time, but I would wager it would be an improvement over replacing bellevilles twice a year.
Reliable, predictable, balanced, frictionless linear pressure. For the little BT30, it seems very doable. Have I missed another bole in the logic?
-
I agree about Bellevilles, which is why I gave up on that whole approach. I expect for your sub-1000 pound target tension, you *might* could find a valve spring or die spring that would suffice. For the 2500# I was aiming for on my attempt, I couldn't find anything close that wasn't really huge OD. Even at 1000#, most seem to up in the 2-3" diameter range. Up to about 5000 RPM, balance will be fairly forgiving. Above that, I think despite a perhaps small-ish diameter, you'd need to at least provide very positive centering for the spring to prevent vibration.
Regards,
Ray L.
-
MIGHT work, But the closer to coil bind you take the spring the more unreliable it gets. I also build race engines. I would definatly suggest a screw down threaded adjuster as you WILL have to readjust the height over time to keep the pressures correct.
Compressing the dual spring packs up to 1000 lbs to install can be a scary deal even with a air/hydraulic spring compressor design to do it safely.
Would not be my first choice but could work.
Bevel spring packs are used around the world every day for CNC machines why reinvent the wheel? WITH your number of tool changes I don't think spring changeouts will be a concern. The choice of spring is that for a reason. Might want to ask a manf WHY they use it.
(;-) TP
-
I already mentioned that the spring would be in a pocket machined in the top of the spindle and and held by a valve spring keeper. I would thread the top of the rod and run threaded nut/flange down onto the valve keeper so there is no need for a spring compres or at all and adjustability would be provided simply by turning that nut/flange. The spring pressure could actually be measured/monitored by the air cylinder, but that's another topic.
If everything was done 'the way they always have been done', we would still be riding horses to work every day . . or worse . . . we'd still be lion food. Edison said something like "I have not failed, I have simply eliminated 10,000 material candidates so far" . Education, experiece, research . . only take you so far, and eventually, like the movie says, you just gotta go 'WTF' and give it a try.
One of my favorite sayings, which I had framed in my offices years ago was "A ship is safe in port, but that is not what ships are built for." But I digress . . .
http://www.summitracing.com/parts/SCC-SRP1363-1/Application/?prefilter=1
This guy seems like a perfect candidate. This spring exceeds the spec. I am leaning this way unless a 'game changer' that I have not though of comes up. I am not concerned with easily solvable retention or installation issues. This spring would only need to be compressed about an inch. I have not done the math on a belleville stack so I don't know at this point how much you would need to compress that stack, but the inch or so required by the coil spring is acceptable in my view. Changing to a lighter or heavier coil spring or replacing the spring would be very simple. The stack designes I have seen so far have no provision for changing the stack without removing the spindle and I am not keen on that idea. I am performance oriented and if the coil arrangement will perform with out the disadvantages of a long stack of springs, then I would be inclined to look at it more carefully. I have more to learn about this, but you guys have been a big help, and I appreciate all of the comments.
Thanks everyone!
-
I have the same 4 ball gripper setup and BT30 as you on a BP clone. ( I was the one who emailed you about grinding my spindle). My quill is locked and I drive the knee for the Z axis so my spring choices were better than If I was using the quill. I settled into these Belleville springs from mcmaster 9712K84. I'm running 30 springs. This gives me 1000 lbs working pressure and leaves me about .375" of travel before my air cyl won't move it. I only need .225” for release. I made a sleeve that goes 1.5" into the drawbar bolt hole and then steps out to the bore size of the top step pulley. then the sleeve is turned down to accommodate the springs. I’ve been running this setup at 6000 rpm the last couple of day’s with no real vibration issues. In fact it seems to be better than my stock drawbar. I’ve got a job I need to get through but unless someone chimes in why that valve spring won’t work I’m going to give it a whack. My working height is around 2.5” of spring so If I can reduce that with a valve spring I’m all for it.
As a side note I’m running drive dog adapter on the spindle nose but I wanted to see how much grip the BT30 had. I removed the dogs and put a piece of .625 cold rolled in a set screw holder. I cranked the screw down fairly hard and tried to turn the rod with vise grips. The rod spun in the holder after a lot of pressure but the holder never turned in the spindle. Not scientific but gave me a little more confidence in my rig.
I have 100 pieces to machine with 4 setups with 4 tool changes each setup so that should be a nice test.
Derek
-
Derek, thanks for the replay. What you have is exactly what I was referring to in that you have a much smaller stack of much larger springs. Your stack is under 3" and arranged on top of the spindle instead of being 12" long and extending thru the spindle. Very good to have a working confirmation of my theory.
Coincidentally, I was slogging thru some numbers on the BV springs here this morning on similar springs and your selection produces the best numbers I have so far. Great choice!
Your finding without the dogs also reinforces the 600# spec I have for under 6k. drawbar tension. That test needed only a torque wrench to be 'scientific'.
There was a comment about re-venting the wheel which I have pondered further. It occurs to me that if BV had some distinct advantage over coils, valve springs would have been replaced long ago. Now strictly to speculate; it seems to me that based on the manner in which Bellvilles flex, they would have a very short fatigue life. That notion is reinforced by another comment about the frequency of replacement, however, by primary objection is the containment issue which is made more complicated by the BV spring OD changing as it flexes.
-
I think the reason the industry uses Bellevilles is the high drawbar forces they need to hit. I was seeing numbers in the 5000 lb range. I also realized early in my design the the more washers you have the less movement each washer needs to make and the less overall pressure applied to the equipment. Also like you said working bellevilles will tend to fatigue them so the more you have the better.
One thought I had before your valve spring idea was to reduce the length of the splined section on my spindle and replace that area with bellevilles. I would have to reduce the diameter to fit in the bore of the step pulley but I think there are some washers in that size range that would work. The math looks good on the valve spring so I'm going to go that direction first and see how it plays out.
A car running that spring probably works it more in one 1/4 mile pass than I would in a year or two of tool changes :)
-
HEY guys go for it. Just a note the referenced spring is a triple set of springs not a single stack.
When you implied using a auto retainer setup I assumed you meant using the keepers as well.
When you do a static rotational grip test it is NOT the same as a running ,cutting spindle. The vibration and side force of cutting effect the holding power of the taper.
You can spend a lifetime inventing new ways to do old things with little improved outcome of the results OR a life time making new parts and gizmos. I am with the patent people 99.9% of everything important has already been invented.
Sling some chips, (;-) TP
-
You can spend a lifetime inventing new ways to do old things with little improved outcome of the results OR a life time making new parts and gizmos. I am with the patent people 99.9% of everything important has already been invented.
Sling some chips, (;-) TP
Yea but where's the fun in that :)
Seriously if I was running normal low RPM spindle speeds I would stick with my current setup. The lower overall height is appealing enough for me to make a stab at it.
The only reason I'm running it without the drive lugs is I have an opportunity to do this 100 piece job and I don't have my indexable VFD yet. I'm just doing light cuts in aluminum so slipping won't be a problem. I'm cobbing up a quick 6 position tray type holder and I'll auto tool changing!
-
I'm cobbing up a quick 6 position tray type holder and I'll auto tool changing!
What are you using to hold the BT30 adapters in yout tool tray?
-
right now they are just sitting in holes I pocketed. I'm designing a really quick and dirty cover with an air cylinder that pops open to expose the tools.
The end game is to have a 10 tool carousel. I'll be using a traditional spring loaded fork arrangement on that. It's all designed and ready to build I just don't have the time. I'm using an old servo motor and spare drive I have to run it. Trying to keep it as simple as I can and work with stuff I have on hand.
-
Short update; with my new 'Mega' 4th axis project completed, I am turning my attention not to completing this BT30 spindle for my new mill.
Two things became apparent while building the new D1-4 Cam loc spindle for the 'Mega Duty' 4th axis; 1) more power was needed to drive the 1" drill bit thru hard material and 2) the mill column would need to be reinforced and welded up to resist twisting during the drilling operation.
The previous 4th axis prototype had only slightly less torque than needed for the 1" drill thru. The previous 'Super Duty' had just under 6:1 reduction and the new 'Mega Duty' has 12:1, so I am not anticipating any problems at all with the drilling. I have taken down the mill column and boxed it in with 1/2" 1018 CR plate and welded it up. The column should no longer get twisted by the drilling force. The mods should allow the pre-hardened material to be drilled thru the entire 9" length from one side at 1" diameter. The material will be here tomorrow.
I had hoped top avoid the big hole by using a coil spring. This would require only a hole large enough for the rod, but when I went back to complete the spindle design, I found that this was not going to be possible for BT30 because of the small size of the adapters. I think it may be doable in the larger sizes, but I have not looked at that. I considered briefly moving to a larger size, but decided against that solution. Therefor I will be using the dreaded belleville spring stack for this first spindle due to the small adapter size. That is the situation, even though it does not sound particularly logical.
-
The entire Gripper assembly is now completed and tested. Here are some photos of the complete assembly and the test rig. I will be doing a full YouTube video on this project, and I have a small fragment showing the testing of this completed gripper. I'll post a link when I have it on my site.
(http://www.thecubestudio.com/4thAxisSpindleProject/BT30StudGripperTestRigWEB.jpg)
(http://www.thecubestudio.com/4thAxisSpindleProject/BT30GripperAssemblyNotationsWEB.jpg)
-
Looks great!
It seems you were able to fit a goodly amount of Bellevilles in there so your ejection force won't have to be as large as mine.
Just a FYI I'm running the same setup as you except all my parts are heat treated and tempered A2 ( I seem to recall your using A2 and S7). I just finished a run on a bunch of castings with over 400 tool changes and it never missed a beat. I have another set to run and after that I'm going to do a tear down and see how everything is looking.
Switching to the BT30 was one of the best things I've done to this mill in a while. I'm loving it!
-
The Bellevilles are working and there are ways around the spring-to-spring friction issue, but I still do not like the whole concept. The BT30 is so small that the gripper sleeve has to be installed from the opposite end of the spindle, which leaves a much larger hole in the spindle leaving the need for a 'platform' to be created for the base of a coil spring. Plus I would have to reanalyse the stresses for a thinner spindle wall (larger hole). There was a spring washer available which met all of the requirements and has a path to a heavier 'sister' as an upgrade path, so that was the easy way out for this first iteration. As you can see from the video fragment, the assembly releases at .20" with the stack set to 1,000lbs and there is an additional; .135" of travel after that. The spring pack binds at about 1,940lbs. It should take about 1,400lbs to release. I have not calculated the max available pull, but I know it is way above what is needed, so the heavier sister will not be needed and I will design the pneumatics for a 1,800lb release @ 90psi.
The gripper will normally run at about 600lbs. I have tested the gripper mechanism to 3,000lbs and there was zero dimensional change, so I'm calling it a done-deal. I will be making a much larger spindle immediately following this one and I will use the coil spring in that design. Interest seems to be waning in this project, so I may not post beyond this, but I will be making a video of the entire development process. After I have presented it to the folks who are behind this whole effort, I will post it up on YouTube.
www.thecubestudio.com/BT30SpindleProject/BT30GripTestFragment.wmv (http://www.thecubestudio.com/BT30SpindleProject/BT30GripTestFragment.wmv)
-
Just a FYI I'm running the same setup as you except all my parts are heat treated and tempered A2 ( I seem to recall your using A2 and S7). I just finished a run on a bunch of castings with over 400 tool changes and it never missed a beat. I have another set to run and after that I'm going to do a tear down and see how everything is looking.
Each tool steel, or any material for that matter, has characteristics that are good in some applications and not-so-good in others. A2 when hardness is the primary requirement. S7 is tougher and shock resistant. A2 would outlive the S7 (used for the ball holder), but would be more likely to fracture. These are not super stressed parts, and should not break nor wear out in our lifetimes, so either is acceptable. The pull stud is the only scary part of the mechanism as a whole.
I watched your video. Very clever indeed. The opening lid was really a surprise! You should spice that up just for fun. Paint it green and put some big teeth and some crocodile eyes on it. I have to build the spindle and the pneumatics, which are fairly straightforward tasks, and then attack the ATC. I have looked at a lot of arrangements. Mohammad comes to the mountain and vice-versa. Somewhere between ideal, realistic and practical is where It will probably end up.
-
have you seen this, it's a denford triac bt30 head.
http://www.denfordata.com/bb/viewtopic.php?t=1953
-
The pull stud is the only scary part of the mechanism as a whole.
You know I've thought about this a lot. If we only have 1000 lbs of up force then 1001 pounds of downforce is going to start to move the tool into the work. But more than that it's going to let the tool holder wobble which in most cases will be catastrophic to the tool. I can understand the concern about pull studs in regular VMCs that are running 2-5K pounds of force but I just can't see it being a problem at the levels we're working at. Feel free to punch holes in this theory as I'm wondering about all this as well!
I watched your video. Very clever indeed. The opening lid was really a surprise! You should spice that up just for fun. Paint it green and put some big teeth and some crocodile eyes on it. I have to build the spindle and the pneumatics, which are fairly straightforward tasks, and then attack the ATC. I have looked at a lot of arrangements. Mohammad comes to the mountain and vice-versa. Somewhere between ideal, realistic and practical is where It will probably end up.
I was thinking of running a MIDI that went "Bwa Ha Ha Ha" every time it opens and closes! It reminded me of the toy bank coffin that lifted and the hand popped out to grab the coin.
Before I started this maching job I was convinced that mounting my carousel on the end of the table was the best way to go. Keep it simple. Well after watching all of the unnecessary X axis motion with my tool tray I saw the value in flying in the carousel. I think I've settled into the best choice for me which is the ATC is positioned to the left of the spindle and up high enough so my longest tool will be above the spindle nose. This is 8.5". The motion is down 8.5" to bring the fork level with the BT30 groove. Over 8" to engage the fork. Down again 3.5" to clear the pull stud to the spindle. Rotate carousel. Up to engage and then reverse out. Bimba makes a nice 3 position round body cylinder. I used their online configuration tool and it was only $75.00 for a 3/4" cylinder.
Now I just need to find the time to make it!
-
You may want to consider that you'll need to move the x axis to one end when you're changing tools to avoid the tools striking the job. Unless you have sufficient z movement to allow this.
-
You may want to consider that you'll need to move the x axis to one end when you're changing tools to avoid the tools striking the job. Unless you have sufficient z movement to allow this.
Hi
I'm driving the knee so I have a fair amount of travel. I also noticed that efficient tool changing depends entirely on the job your running. Tool length, Height of work etc. So that being said I figured using a macro for the tool change that picks up the tool change position DROs. I would then specify where I needed the table to be in order for the tools to clear. Why move the Z 12" if only moving it 6 will get the job done. At least that's todays theory Check back with me in a couple of hours and I may have a different take :D
Derek
-
sounds like the best way to go.
-
Question to anyone: What are you using for a safety interlock to prevent the power drawbar from releasing if the spindle is turning?
Speaking for myself, I do not trust MACH with such an important job, so I would be looking for something autonomous, or activated by the servo drive and not by MACH. I am thinking about a spring loaded solenoid that must retract a mechanical interference from the mechanism before the drawbar can release. Perhaps steal some energy from the motor to drive a relay to keep the 'lock' in place until the motor has zero energy.
The drive I am using can sequence everything with an internal program, but I am looking also for a more 'generic' and autonomous solution which would work with less sophisticated drives.
-
The only reasonable thing I've been able to come up with is making the ram button on the end of the air cylinder rotate. Then have a wire that the 24VDC that runs my estop system run through it like it was another estop switch. If the ram comes down while the spindle is turning then the button will turn break the wire or contact and kill the 24V. My estop system is totally separate from Mach. when you kill the 24V you kill everything. Servos, VFD,coolant and the air valve for the draw bar. The air valve is a spring return so when you kill the power the default is up.
Not elegant but for an electronics ditch digger like me it will get it done. I've been working with Mach daily since it was master5 and never witnessed it activate an output on its own.
Having said that it wouldn't surprise me at all if it happened tomorrow. That's why my estop has nothing to do with mach.
Speaking of drives what are you running? I'm looking at the Delta VFD-M series. it's in the right price range , takes single phase input and has simple indexing to align the drive lugs.
-
the relay that enables the vfd/spindle drive can be used to isolate the solenoid operating the drawbar.
-
the relay that enables the vfd/spindle drive can be used to isolate the solenoid operating the drawbar.
In this case, I think the drive would need to be disabled in order for the drawbar to work. With the drive disabled there would be no mechanism (other then a spindle lock) to make sure the spindle is stationary.
-
ah i see it's the need for the spindle to be stationery for the drawbar to operate you'tre worried about. Add a tacho to the spindle and use this to operate a relay that disables the drawbar.
-
The only reasonable thing I've been able to come up with is making the ram button on the end of the air cylinder rotate. Then have a wire that the 24VDC that runs my estop system run through it like it was another estop switch. If the ram comes down while the spindle is turning then the button will turn break the wire or contact and kill the 24V. My estop system is totally separate from Mach. when you kill the 24V you kill everything. Servos, VFD,coolant and the air valve for the draw bar. The air valve is a spring return so when you kill the power the default is up.
Not elegant but for an electronics ditch digger like me it will get it done. I've been working with Mach daily since it was master5 and never witnessed it activate an output on its own.
Having said that it wouldn't surprise me at all if it happened tomorrow. That's why my estop has nothing to do with mach.
I am running 62 dev version because that is what works with the PoKeys board that I am messing with. You cannot use the 'stable' version with PoKeys and 62 has broken B and C axis feedrates and also has a tendency to either have fatal errors or just up and quit without warning. I have a screen macro that simply lights an LED and activates an output. The LED always lights, but the output would go high sometimes and sometimes not. When Mach starts and exits (intentionally or otherwise) andwhen the PC starts up or shuts down, or when Smoothstepper has a tantrum, it is hard to predict what any pin is going to do. My safety interlock will be a mechanical device which will provide a positive physical impairment of the drawbar movement. I just have not decided how to arrange it.
Speaking of drives what are you running? I'm looking at the Delta VFD-M series. it's in the right price range , takes single phase input and has simple indexing to align the drive lugs.
My spindle motor is using a Copley Controls Xenus drive. 165V 40A on single phase 120V. Will also run on single phase 220V, but the bus voltage also doubles. The motor is 140V 37A peak.
-
ah i see it's the need for the spindle to be stationery for the drawbar to operate you'tre worried about. Add a tacho to the spindle and use this to operate a relay that disables the drawbar.
There is a spindle lock that the motor cannot overpower, and the motor actually has a tacometer built into it. It puts out 7V per thousand RPM if I remember right. That is an interesting option, thanks!
For BT30 or any other scheme that has drive dogs, the spindle must home and stay homed. Not all spindles have locks and I am looking for the most universal solution . . i.e one that the typical mill could use . . although really I don;t know what 'typical' means :-[
There is that functional issue, certainly, but what I am focused on is preventing the adapter from being ejected unexpectedly at 7,000 RPM with a sharp tool in it. I am beginning to get the impression now that a safety interlock is a foreign topic.
-
Well I was planning on using pokeys to run the ATC so it looks like I'll be in the same boat as you! On this machine I'm not running the smooth stepper so that should remove one area of uncertainty. I'm running a smooth stepper on my router and have experienced some quirky actions for sure.
-
There is that functional issue, certainly, but what I am focused on is preventing the adapter from being ejected unexpectedly at 7,000 RPM with a sharp tool in it. I am beginning to get the impression now that a safety interlock is a foreign topic.
[/quote]
then my earlier reply will sort that.
"the relay that enables the vfd/spindle drive can be used to isolate the solenoid operating the drawbar."
-
ah i see it's the need for the spindle to be stationery for the drawbar to operate you'tre worried about. Add a tacho to the spindle and use this to operate a relay that disables the drawbar.
Ultimately you would want this as a separate entity from Mach. Then at some point you have to decide what do you trust more. Mach or the circuitry your using to monitor the tach and activate the relay.
Not slamming your idea just trying to draw it to the next conclusion. Ultimately I like this idea if it was safe.
-
Denford use this on their spindle interlock to keep the door closed when the spindle is running.
-
Right now I have a relay that turns the vfd on and off through Mach. If I routed the air valve in a way that when that relay is energized there would be no power available to the air valve so even if the output was turned on in mach the drawbar would stay put.
Basically a variation of what angel tech suggested a few posts back. I think for my purposes this may be the best route.
-
you just replace the existing relay with a douible pole relay and use the second set of contacts to isolate the drawbar air solenoid.
-
That sounds too simple :)
-
If the relay turns off the drive and allows the drawbar to move simultaneously, there will still be a 'danger zone' if the spindle was running a 7K and the drawbar was set to move. The adapter would release as soon as the VFD was disabled, but it would still be spinning at 7K.
The Tach idea would eliminate that possibility. Also, as mentioned previously, I don't see how you can disable the drive and then change the tools. What would keep the spindle stationary?
It occurs to me that there has to be some sort of home sensor on the spindle in order to change tools that have drive dogs, so unless I am missing something, why not just watch the home signal? i.e. WHILE <continuous home sensor signal>, do the tool change, WEND. The spindle would have to be stopped and any movement would halt the tool change . . . which would be a good thing, presumably.
I think this is the hot setup for the spindle, so that just leaves every other movement . . . LOL! Actually, I have magnetic sensors on the air cylinder, so it will be fairly simple to detect that the rod has reached the correct release and grab locations. If it fails to pick up a tool, it would move past the correct grab spot, and that should be pretty easy to detect also.
-
Good point.
The Delta VFD-m uses an input, usually a proximity sensor for indexing. You can tell it what angle from the sensor you want it to stop at. This is the same drive that David DeCaussin uses on his machine. If you watch the video you can see it stops, then turns to index. Not sure if it locks the position though. My thinking at this point is to ultimately have the VFD dictate whether or not the draw bar can engage. That may be a dream or something as simple as using an intelligent output on the drive. I'll have to see what I can come up with when I get the drive.
If you don't trust Mach-pokeys-smoothstepper to not turn on (or off ) an output then you can't really trust it for any other method to prevent unintentional draw bar movement.
-
Not familiar with VFDs but I use the Zero speed output of my servo drives to enable the drawbar, maybe VFD's have such a feature?
I do mine in the PLC, also have it interlocked the other way so the spindle cant start if the drawbar is energised.
Hood
-
Good point.
The Delta VFD-m uses an input, usually a proximity sensor for indexing. You can tell it what angle from the sensor you want it to stop at. This is the same drive that David DeCaussin uses on his machine. If you watch the video you can see it stops, then turns to index. Not sure if it locks the position though. My thinking at this point is to ultimately have the VFD dictate whether or not the draw bar can engage. That may be a dream or something as simple as using an intelligent output on the drive. I'll have to see what I can come up with when I get the drive.
If the drive can signal, that would be a good option to consider. Most likely I will do my own setup that way, but it is useful to get as many ideas on the table as possible and then pick up the one that best suits each machine.
If you don't trust Mach-pokeys-smoothstepper to not turn on (or off ) an output then you can't really trust it for any other method to prevent unintentional draw bar movement.
It is not possible to get MACH out of the loop completely since the tool change command initiates in the G-code and (typically) MACH is controlling the spindle. My objective is to have, in addition to an 'enable' function, a positive mechanical interlock that will prevent the drawbar from moving unintentionally, regardless of the source. I suppose the reasoning would be similar to the folks who have their E-stop setup outside of MACH, but still triggerable by MACH.
[/quote]
-
Not familiar with DVDs but I use the Zero speed output of my servo drives to enable the drawbar, maybe VF D's have such a feature?
I do mine in the PLC, also have it interlocked the other way so the spindle cant start if the drawbar is energised.
Hood
For my own machine, I will use the drive as you have done. Also I will be using a separate PLC to control the sequencing and watch the various sensors. Did you write your own code for the PLC?
As with my motor controller, I am trying to anticipate what might be out there so that I can accommodate as many setups as possible, and then make the ATC controller configurable. If you click on the lower right image on this page www.theInTurn.com (http://www.theInTurn.com) you will see the configurator for the 4th axis motor controller. Note that certain safety features are optional, like monitoring the signal, and how to react in the event of a failure. Most likely I will do something similar for the PDB/ATC controller and let the user pick the source of the interlock; drive, home sensor, tach, etc.
Do you have a mechanical lockout on the drawbar, or do you rely on the 'enable' function?
-
LOL... Hood said "Not familiar with VFDs" and somehow in your quote it appeared as "Not familiar with DVDs" ;)
Dan
-
Ha ha now that is naughty, editing quotes ;D
My PLC is a Koyo DL06 and I programme it with the graphical ladder logic which I find very easy compared to script.
My solenoid can only act if the PLC is getting a zero speed signal from the spindle drive so it is partly hardware and partly software I suppose. Drive and PLC ladder being the software, drive output/plc input and relay output from the PLC being the hardware. I dont have it via Mach at all as I dont have an ATC so my drawbar is just via a button on the panel, that may change in the future if I get time to servo the knee or even better pick up a decent bedmill :)
The Spindles Enable is also via a switch on the panel which goes to the PLC. If the PLC is seeing that the draw bars relay is active then it will not pass the spindle enable signal to the drive.
If I rotate the spindle by hand when the drawbar is active it will drop out the solenoid right away, not that that is really any use but it shows the ladder is working as it should :)
Hood
-
Ha ha now that is naughty, editing quotes ;D
The Devil made me do it! Actually it was the spell checker and an itchy trigger finger. Why the spell checker looks at quotes is something of a mystery.
My PLC is a Koyo DL06 and I programme it with the graphical ladder logic which I find very easy compared to script.
My solenoid can only act if the PLC is getting a zero speed signal from the spindle drive so it is partly hardware and partly software I suppose. Drive and PLC ladder being the software, drive output/plc input and relay output from the PLC being the hardware. I dont have it via Mach at all as I dont have an ATC so my drawbar is just via a button on the panel, that may change in the future if I get time to servo the knee or even better pick up a decent bedmill :)
This is the type of thing I am looking to protect against; accidentally hitting a button, MACH or the PC misbehaving. I have a pneumatic lock on both the 4th axis and the spindle and when things go awry, sometimes I hear the familiar 'pop pop' which is harmless in this case, but would not be if the 'pop' was the drawbar releasing a tool. I plan to have the interlocks that you have wisely included on your mill and also have . . . uh . . I'll call it a 'lock' to differentiate from 'enable'. The lock will be engaged at all times via spring load. This will be a mechanical interference which will physically prevent the cylinder from contacting the drawbar. It will have to be retracted vie electric solenoid before the drawbar can be released.
The sequence will probably be something like:
*receive tool change command from MACH (or manual override)
*wait for spindle to stop rotating
* - - - 'disable' - - drive. This is a topic in itself, and could be:
- engage spindle lock, disable drive
- or-
- leave drive enabled to hold spindle position and disconnect (via swapaxis board) the step stream so that the drive will not rotate
*enable Drawbar
*check sensors (including drive outputs - zero speed or equiv), drawbar position, and whatever else might be a good idea to check
*return tool
*sensor check - is tool in correct position?
*retract Drawbar safety lock
*release tool
*position commanded tool
*check elevations (Z axis or carousel) - is spindle nose all the way on the tool?
*grab tool
*Sensor check - is drawbar in correct position for a seated adapter?
*extend drawbar safety lock
The Spindles Enable is also via a switch on the panel which goes to the PLC. If the PLC is seeing that the draw bars relay is active then it will not pass the spindle enable signal to the drive.
If I rotate the spindle by hand when the drawbar is active it will drop out the solenoid right away, not that that is really any use but it shows the ladder is working as it should :)
From this I assume that your drive is disabled during a tool change? You press a panel button to disable the drive, then press another panel button to activate the (Pneumatic?) drawbar release, presumably while physically holding the adapter by hand? Swap tools manually, lining up the drive dogs by eye, and then press a panel button to grab the new tool?
If I have that right, then I think this is also what Ray does with his new motor driven drawbar setup. Adding the ATC removes the human supervision that you have now. I'm just wondering what methods people use (or think about using) to replace the operator's eyes and good judgement when automating this particular process since it is potentially extremely dangerous. It will be a lot easier to incorporate in the design than to go back and add later.
-
When I press the drawbar button it automatically takes the enable away from the spindle drive so only one button needed to press.
The Spindle disable button is just used when I have my hands in about the work for any reason, have the same on the lathe. Both have brakes on the spindles and this unlocks the brake as well.
Drawbar is air over oil.
Hood
-
Oh and if I ever get round to putting a servo on the knee and building an ATC then I will just use the homing feature of my drive to locate the spindle but if the drive didnt support that it would be easy enough just to home to an opto for me.
Hood
-
No surprise that your setup seems well thought out for a manual change power draw bar. Did you build it or did it come with?
Is your spindle motor a direct 1:1 drive or is there a reduction?
It seems to me that anything other than 1:1 1:2 or 2:1 would defeat the drive's homing as the spindle would stop at a random spot when the motor homed. My current temporary mill has odd ratios on the dual range head. I suppose they made sense at the time, but at this point ???. I will need to use the opto which is still there left over from tach duty quite some time ago.
For the new mill I am thinking about doing a 1:2 and 2:1 ratios for low and high. In this scheme, the motor can home on the encoder index and the tool will always be at the correct azimuth. Those ratios would also provide a nice low range for big drills and big taps and about 7K max spindle speed.
How is the knee related to the ATC? Would you mount the tool holders on the table?
-
The power drawbar was standard on the Beaver mills, the later ones than mine were single action intensifiers where mine was a bang bang type. It needed five or six pulses to get enough to release the tool and it took 3 or 4 seconds. I got a bigger intensifier from a friend and now mine is single acting and releases the tool in well under a second.
I have an 8Kw servo on the spindle so I could have increased the speed, in fact I did at first but there is a gearbox that is to change between 1:1 and 10:1 ratios for main and back gear. The gearbox didnt sound happy at the higher revs so I just put it back 1:1 on the pulley. I am thinking on the hoof here, the drive can be set to move any amount of encoder counts off of the index after homing so likely I could work out the amount needed even if I had a weird reduction ratio.
Regarding the ATC, I would want the knee as an axis so I could drop the work well out the way so I could swing in a carousel or maybe even some form of dual arm changer, there just wouldnt be enough room without doing that as the quill only has 150mm of travel and some of the tools I use are longer, in fact a lot longer. The knee already has a motor on it, just an induction though, but it would be easy enough to fit a servo instead. I will do it at some point but just too many other things on the go at the moment. I will likely do away with the quill as the Z when I do or maybe just use it for picking up the tools. The great thing about the drawbar setup on mine is the cylinder moves with the quill so it can be operated at any position of the quill.
Hood
-
It would be extremely unlikely for the encoder to happen to be in the correct position by chance, so Degrees-off-of-index is needed to locate the actual desired home position relative to the random position of the physical encoder install, but I don't think it would be usefull in homing a spindle with reduction (other than the few ratios I cited).
For example, take a simple 3:1 reduction.
With the motor homed, the spindle could be at 0,120, or 240 degrees. Each position would require a different number of motor degrees to reach 0 (or 180) for a tool change, and I don't *think* the drive keeps track of encoder rolovers and even if it did, it would be a nightmare to extract that info . . again that just 'on the hoof" :P
Even with 1:1, 1:2, 2:1 etc, ratios, the degrees from index would still be useful to move the spindle from 'opto home' to the required tool change orientation. I think I have that feature on the Xenus, but now that you've brought it up and I've noodled over it a bit, I can see where it will be needed to operate the ATC from a PLC. I had to write some tracking code for my balancer and that encoder is absolute, so it was tolerable . .barely. I think I would prefer to get poked in the eye with a sharp stick to attempting that on a PLC with an incremental encoder. Kflop comes to mind because they have a lot of libraries. Perhaps I will take a peek at that.
What is clear is that this aspect of the process is going to need some more scrutiny.
Really I did not have an appreciation of how complex it would be to develop an ATC. Or more to the point a safe ATC. The mechanisms needed are only a small challenge, but the function will be a large one, methinks. This is another of those cases where one of your random comments is going to probably save me a lot of headache down the road. :-*
-
My drives can be set up to roll over at any amount of encoder counts, it is in fact the way I use it for my lathes turret, it has a 20:1 gearbox on the motor so the drive is set up to roll over at 160,000 pulses. As the drive is also an indexing drive I can set up up to 64 positions and call any one of them from a combination of 6 inputs, life is easy :)
Big problem you have is you are trying to design in safety features and you have no control over the hardware it is going to be fitted to. I think likely the best you are going to manage is to have a setup that requires certain signals and leave it up to the user to provide them whether that be by real signals or trickery. Either way you have done your best to make things safe and it will be the end user that has failed if they dont do as needed.
Hood
-
Hood; thanks for all of the info. You are a little over my head with the last post which only shows that I have a lot of study to do yet on the best way to skin this cat.
I have a tiny update and a short side trail for this thread. First the update: I have decided to go the cartridge route and make a complete housing for the spindle. This solves some problems and creates some problems. It seems DIY prefer aluminum for the housing, but I prefer steel for a number of reasons. The problem is that a chunk of steel large enough to create a flange and hold the massive lower bearing AND long enough is over US$200 and heavy as hell so shipping is also scary. IN the photo below, are the raw materials that I will be turning into a spindle. Purchasing an appropriate size piece of stock for the nose and the body of the cartridge put the cost under US$50, but it introduces the problem of attaching the two pieces together. After weighing options, I have decided to furnace braze the two pieces together. I have a furnace large enough to hold the parts and I have acquired a furnace brazing paste that contains flux and metal powder which is applied ti a specific slip fit between the parts. I am well aware of the furnace brazing process, but have never done it myself, so this will be interesting. If anyone has experience with this process, please post any tips or tricks, or specs on the slip clearance if you know what it is. I know there is a spec, but have been unable to find it so far.
(http://www.theCUBEstudio.com/BT30SpindleProject/MaterialsWEB.jpg)
(http://www.theCUBEstudio.com/BT30SpindleProject/MaterialsStackedWEB.jpg)
-
OK now for the side trail. This is related to the BT30 spindle project because the device will end up inside the head of the new mill. For now, I have built the device and it will be installed under my current mill head.
Problem: need to drill bolt patterns in parts machined on the 4th axis. Taking the parts out and drilling the bolt pattern on the mill is a PIA, especially since the 4th axis has indexing capability and can easily rotate and hold any part for drilling. The mill had revolves and can face the 4th axis, but that is more time consuming than moving the part to the mill for drilling because the head has to be re trammed each time.
Solution: horizontal drilling spindle on (for now) or in (the new mill) the mill head.
To accomplish this, you first purchase one of these - new:
(http://www.theCUBEstudio.com/GearBoxProject/GearBox.jpg)
The you grab one of these off Ebay for something less than the new price of US$285 plus something like $80 for the MT adapter. The chuck is the largest Jacobs 'Super Chuck' with capacity to 3/4". The huge MT adapter was trashed and unusable, but for my purposes, it was perfect because it was big enough to cut down to 20mm.
(http://www.theCUBEstudio.com/GearBoxProject/Chuck.jpg)
Next you reduce the diameter of the MT adapter to a 20mm press fit and duplicate the snap ring grooves and pin holes of the original shaft.
Move the bearings and gear over to the new shaft:
(http://www.theCUBEstudio.com/GearBoxProject/ShaftAssembledWEB.jpg)
Rig a method to preload the bearings and reassemble. Now you have a horizontal drilling capability that can hold tools up to 3/4" and can drill and hard tap a workpiece held in the 4th axis. ONe of the first uses will be making the BT30 spindle housing.
(http://www.theCUBEstudio.com/GearBoxProject/GearBoxAssembledWEB.jpg)
-
Steve,
Fascinating approach! Can't wait to see how it works out!
Regards,
Ray L.
-
Like your solution! Especially the fact it turned out relatively inexpensive. There are such heads ready available, but are very expensive.
By the way, don't see a reason for preloading the bearings if all it's going to is drilling. I think there would be much more backlash in the gears that could be a problem for rigid tapping.
Dan
-
Like your solution! Especially the fact it turned out relatively inexpensive. There are such heads ready available, but are very expensive.
I think you are talking about the right angle gearbox? I don't know of any that would fit onto an X3 spindle, but you're right, the Bridgeport angle heads are a bit pricey. An interesting advantage of mine is that I still have the vertical spindle active while the horizontal is running, so given enough space, I could still have the mill spindle participating in the machining process. I have two mounting schemes and will implement both, one is horizontal mounted behind the spindle and belt driven. This has the advantage of being strong and rigid and allowing different ratios.
The second scheme mounts the gearbox on the spindle (ala Bridgeport) but actually bolts to a collar that in turn bolts to the bottom of the head. Advantage here is that the drill can now be rotated to any position in the horizontal plane. Disadvantages are that it consumes the spindle, interferes with the tooling bar and is a relatively weak mount because the stress has to run all the way thru the aluminum gearbox casting. Still, If I need angled holes, I can make them.
By the way, don't see a reason for preloading the bearings if all it's going to is drilling. I think there would be much more backlash in the gears that could be a problem for rigid tapping.
Dan
Can't have the shaft flopping around if I expect to hard tap or do any light milling. The gear lash is a good point. Had no occurred to me yet. On first blush, it seems I could measure the backlash and compensate in the tapping macro. Worst case is I would get a bit more H than is built into the tap. For most applications, this would not be an issue. We shall see. I have to get it mounted first and then work out any . . . um . . . less than optimal characteristics . . . that may arise.
-
My drives can be set up to roll over at any amount of encoder counts, it is in fact the way I use it for my lathes turret, it has a 20:1 gearbox on the motor so the drive is set up to roll over at 160,000 pulses. As the drive is also an indexing drive I can set up up to 64 positions and call any one of them from a combination of 6 inputs, life is easy :)
Can you explain the relationship between the gearbox ratio and the rollover point. I've noodled over this on and off for a while now and it is not coming clear. I'm sure it is one of those obvious things that somehow become invisible if you think about it too hard. :-[
Big problem you have is you are trying to design in safety features and you have no control over the hardware it is going to be fitted to. I think likely the best you are going to manage is to have a setup that requires certain signals and leave it up to the user to provide them whether that be by real signals or trickery. Either way you have done your best to make things safe and it will be the end user that has failed if they dont do as needed.
Well, you hit the nail on the head, Mr. Hood. I already have some trouble with people modifying stuff in unsafe ways. I've been asked already for the design, including buying it, by some folks who are well known around here, but I have not shared anything other than what is in this thread. I am a bit paranoid about liability because I live in the USA where people successfully sue because the coffee was too hot and burned them when they spilled it in their lap while driving . . . . absurd, but true. The more I contemplate your statement, the more I lean toward keeping this exclusively for my own use.
-
Yes, Steve, I was talking about the Bridgeport right angle head attachment.
Regarding the backlash in rigid tapping, compensating it in software should work great. I do rigid tapping on a lathe that has about 0.1mm backlash on the Z and compensate it in Mach and it works very well.
Dan
-
Steve, what I was talking about is the capability of my drives to use the incremental encoders as, I suppose, you would an absolute encoder.
I have a gearbox on my turrets motor that is actually 25:1 reduction (not 20:1 previously mentioned) and my encoder on the motor is 2000 (8000ppr) so I tell the drive that it should roll-over at 200.000 pulses (25 x 8000).
The motor turns 25 times but the turret turns once and the count in the drive goes from 0 to 200,000 then starts again. That means as far as the drive is concerned 360 degrees is 200,000 pulses. Because I can set things up this way then I can index to any position on the turret and the drive knows exactly where to go, doesn’t matter if the turret has rotated 30 full turns or not as the count would roll-over at one rev (200,000 pulses). In reality that never happens as there is another nice feature, the drive can be set to take the shortest route between index positions and thus it can go both forward and backwards depending on the index you call, this makes it ideal for a turret or an ATC.
The tricky part was the ladder, it took me a while to get my head round how to accomplish it but it worked out well in the end.
I keep the logic side of the drive alive so I never have to home it unless switch off at the mains or there is a power cut. Even if the drive faults for some reason, say for example some stupid person pushes a boring bar too far into the holder such that it sticks out the back and then when indexing it jams against the saddle and faults the drive (Don’t know who would ever be so stupid to do such a thing ;) ) Anyway even if the drive faults it just needs a reset and it still knows exactly where it is as it has been keeping track of the encoder counts.
If I lose power then all I need to do is home the turret, I have it homing to a low current value so I have it clamped when homing and thus it doesn’t move. To home after a power failure I just unclamp, manually rotate to tool 1 and then clamp then home and that’s it set.
Hood
-
NEW VIDEO!!
Hood: Thanks! Now it makes sense. Very clever use of available features - as usual.
Dan: the amount of play in the gears did not make a noticeable difference in the tapped threads. The tap went in and out smoothly, no binding, no aftercutting, and the screw feels (I know this is subjective) the same as threads tapped in the mill spindle. Check it out.
http://www.youtube.com/watch?v=Dd_M6Li_OrA&feature=youtu.be (http://www.youtube.com/watch?v=Dd_M6Li_OrA&feature=youtu.be)
-
Looks very good, Steve!
Dan
-
Just finished another (unrelated) project and I am back on the BT30 spindle. Carving has begun on the turkey. The part is currently flipped over and I am working on the nose, which includes the nose beargin pad, the BT30 taper and threads on the nose for a bearing retainer. Next is another flip and a steady rest for boring the 8.5" deep hole to final size. Then grinding the bearing pads to complete. The BT30 taper will be final ground in-place on the completed spindle.
Progress photos:
Initial OD turn including body, top bearing pad and drive end.
(http://www.thecubestudio.com/BT30SpindleProject/SpindleTurnInitialWEB.jpg)
And now drilling hard material 1" diameter 8.5" deep. The operation stalled the previous 'Super Duty' 4th axis and it had to be done in three steps. That was one of the motivations to create the new 'Mega Duty' 4th axis. The previous 4th axis could do the job with a bigger motor, but 750watt is the largest motor Mitsu makes that runs on single phase power. The new 4th axis munched thru this operation with no problem. Coolant fed drill. No pilot hole.
(http://www.thecubestudio.com/BT30SpindleProject/SpindleDrillWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/SpindleFirstCutPosedWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/SpindleHoleDepthWEB.jpg)
-
. . . . another nice feature, the drive can be set to take the shortest route between index positions and thus it can go both forward and backwards depending on the index you call, this makes it ideal for a turret or an ATC.
Hood
It does sound ideal. I was directed (by BR, I think) to a video where a fellow (founder of Fanuc or something like that) had made a very nice ATC and apparently went on to make an entire mill. He used a Geneva mechanism in his as I recall. That was the only thing I did not like about it. I will check to see if Mitsu has a similar feature to what you describe as I have a spare 400 watt Mitsu motor and drive on a shelf. The Mitsu has the real time auto tuning which is particularly good for the 4th axis, but that feature has little advantage for a tool turret and what you describe does seem like the hot setup! Especially since it is independent of MACH!
The drive you are talking about is Allan Bradley, is that correct? (in case the Mitsu does not have such a feature). How do you talk to the drive?
-
The guy you are talking about would likely be Dave DeCausin (sp?) one of the founders of Fadal.
Yes the drive is a DSD but it needs to be capable of Indexing, the standard drives wont do that.
I talk via the PLC and the ladder has most of the integration in it for the unclamp/rotate/clamp timing. If you simply just wanted to index you could simply have some inputs and it would index to the selection you called.
The inpouts are seen in the drive as follows
Pre-Select 0 = 1
Pre-Select 1 = 2
Pre-Select 2= 4
Pre-Select 3 = 8
Pre-Select 4=16
Pre-Select 5 = 32
So as you can see depending on which inputs you use you can select any of the 64 index positions, well really only 63 as index 0 is not really valid for this operation.
You also need an input set up as Start Index and likely Start Homing, also you may have Reset Drive or whatever. The drive has 8 digital inputs but the nice thing is you can set them up for multiple uses. That means if you are doing an operation that requires two inputs you dont have to waste two inputs if they can legally be seen by the drive at the same time, for example you could have the start Index set up with the Presets.
Have attached a few screenshots of the drive setup, first shows the Index presets I have set, have opened the first three positions so you can see the configuration in the drive.
Next is the Inputs, as you can see I only need 4 presets as I only have 12 positions so there is no need for me to double up on the inputs.
Last is the outputs setup.
Heres a vid of my first tests with the turret I made from scratch (well curvic coupling wasnt home made ;) )
http://www.youtube.com/watch?v=hRTfxuXf6yk
Hood
-
Very nice. Thanks for taking the time to explain.
So you are using digital inputs. Mitsu has both digital inputs and you can also talk to the drive, but thru RS422, so a converter is needed. I have acquired the converter, but it is still in a box . . . somewhere . .
I also have a couple of Copley drives and they are straight RS232 serial and have a text based command set. Very easy to use, but no digital input option like the Mitsu. Each has its plus and minus.
You have given me a lot of good info to think about as the ATC project grows closer. I will have the spindle done soon.
Q: if you don't mind sharing, how are you locking the turret? It appears to be very heavy and takes a moment to settle. Do you have a damper on it?
-
The turret plate is 305mm dia (not including the bar/drill holders which are not in that vid) and 60mm thick but not sure of the weight but its heavy ;D
It locks down onto a curcvic coupling and the clamp/unclamp is by hydraulics, no damper.
In these vids I didnt have a keyway broached in the upper pulley and it moved around slightly (I was waiting on a broach the right size arriving but couldnt wait to test ;D ) so I had a delay set in the PLC before clamping to make sure it settled before clamping.
Hood
-
Thanks again for the info, Hood. You and BR and Himmy and many others need to be recognized occasionally for the vast amount of time you all spend helping others.
So Kudos to all and please know that it is much appreciated.
Quick update on the spindle. Spindle proper is completed except for grinding the bearing pads and drill/tap for the drive dogs.
Progress photo here:
(http://www.thecubestudio.com/BT30SpindleProject/SpindleFinishedWEB.jpg)
-
Couple more progress photos:
The latest addition to my 'mini machinign center' is the horizontal spindle. Here making quick work of the bolt patterns and countersinking.
(http://www.thecubestudio.com/BT30SpindleProject/HousingDrillBoltPatternWEB.jpg)
Final fix on the friggin' spring stack. I despise these things.
(http://www.thecubestudio.com/BT30SpindleProject/GripperAssemblyFinalWEB.jpg)
Housing parts:
(http://www.thecubestudio.com/BT30SpindleProject/HousingPartsWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/HousingAssemblyWEB.jpg)
And finally the whole gang in a 'family' photo:
(http://www.thecubestudio.com/BT30SpindleProject/ComponentsAllWEB.jpg)
Some detail items and final assembly and the spindle is complete.
-
HIYA Steve, looking good. A couple of questions.
Do you plan to preload the bearings? or let them float?
Is your spring pack release mechanism going to put full load on the bearings?
(;-) TP
-
The lower bearing is dual row angular contact and is preloaded. It is ABEC3. The spindle nose has a threaded retainer, so if I chose to use separate ABEC7 angular contact set, the retainer would clamp them together (preload is already set on these bearing sets). In either case, the upper bearing is pressed in (lightly) to the case and floats on the spindle (snug). The housing is steel, to the expansion difference will be only the delta in temp from the spindle to the case, which will be minimal.
A C3 top bearing would probably have enough clearance to not need to float, but I did not do the calculations on it because I elected to float the top bearing anyway.
The lower bearing will take the full force of the release, but it is well within the static load spec on the bearing. Again, I have not done calculations to quantify, but I would not be surprised if the force required to fault the Z axis is greater than the release pressure. 400 watt motor thru belt reduction to a relatively fine thread ball screw. The new mill will have a 750 watt Z axis motor because the head will be much heavier.
Edit: incidentally, I am of the opinion that the big CNC manuf spec a certain finish on the Belleville springs. If I could do so, I *might* be inclined to use them again . . unless the cost was prohibitive. I will never use off-the-shelf commercially available Bellevilles again for this application. In fact, I am considering selling this prototype BT30 and making a larger spindle . . . for which I would be able to use the coil spring.
-
Steve what holds the lower bearing bearing set snug in the spindle housing?
In the old days we coated the bevells with high zinc grease and never saw a problem with years of service between change outs. Sometimes that rougher looking surface is better at holding the lube than a super slick finish.
Lookin Good, (;-) TP
-
The lower bearing is pressfit on the spindle and in the housing. I use Loctite 609 bearing set as well if it is a steel bearing in an aluminum housing. I will use that here as a precaution. Mechanicallty there are pins that prevent the bearing from moving and lastly, I have decided to preload the upper bearing in order to take the release stress off the main bearing. So altogether I think that is 4 methods of retaining the bearing.
This wraps up the spindle project. If anyone is interested in purchasing this prototype then I will build a larger one, otherwise I will use it in the new mill as-is. PM me if interested.
Here are progress shots of the completed spindle. All that remains is the drive pulleys, which attach to the steel clamp collar shown.
Two sets of holes in the clamp collar are to mount items above and below the collar.
(http://www.thecubestudio.com/BT30SpindleProject/CompletedSpindleBackWEB.jpg)
From the front, you can see the Labarenth seal and the mounts for the drive dogs.
(http://www.thecubestudio.com/BT30SpindleProject/CompletedSpindleFrontWEB.jpg)
-
You were talking about the finish on the disc springs, have you had them plated or is that the way you got them? The only ones I have ever seen have no plating at all and are black (maybe some sort of passivation?)
The way the cylinder and springs are designed on the Beaver mill is quite nice I think, spindle, drawbar and cylinder are all in one unit thus no pressure on bearings at all as far as I can see.
Hood
-
You were talking about the finish on the disc springs, have you had them plated or is that the way you got them? The only ones I have ever seen have no plating at all and are black (maybe some sort of passivation?)
Black Oxide, most likely. By 'finish' I was referring to the smoothness. These springs came from McMaster-Carr. Some are very nice and cleanly punched and some have ragged edges or raised ridges at the OD from the punch die. On a brand name American made part, these imperfections would have been cleaned off, but with No-Name suppliers like McMaster, you don't know where most items are coming from . . although you can easily guess.
Raised ridges from the punch operation were left on the edges of the springs. These uneven raised edges, when facing each other in the stack, refused to stay aligned. The ridges would slip off each other driving the spring sideways with a very great force. The only solution was to remove this ridge and radius the ID slightly - by hand - on each of 60 springs. Now they are behaving, but that is WAY too much time to put into that component. For anyone contemplating a build like this, I would recommend buying the springs from the manuf, or at least know the brand name of the springs. However, reality check; I found springs that were nearly $4 EACH. That makes a US$240 stack. The stack I used was a little over $40 and the coil spring was a bit over $50. Functionally, the springs are fine and in any un-stacked application, the ridge would be insignificant, but for stacking, the quality of the finish must be very good or the springs will splay out.
The way the cylinder and springs are designed on the Beaver mill is quite nice I think, spindle, drawbar and cylinder are all in one unit thus no pressure on bearings at all as far as I can see.
Hood
My R8 power drawbar did not put pressure on the bearings. But those are much smaller bearings. I just went thru the exercise of calculating the leadscrew force and the 400 watt motor faults at about 650lbs and the 750watt motor would push the head with about 1,250lbs. The drawbar is supposed to be run at about 600lbs, but even running it at 1,000 lbs, the release force is not much more than the spindle would see from the axis drive, albeit a lot more frequently. This is all academic really, because the bottom line is that the release is about 1,400 lbs and the bearing is rated over 10,000 lbs, so there is no concern over hurting the bearing.
The R8 drawbar was a different story with a much higher release force and a much smaller bearing. Actually, one can only imagine the actual force on a typical R8 spindle bearing that is generated by the shock force of hammering on the drawbar to release the collet.
Still, There is a large steel collar on the drive end of the spindle and it would be easy enough to just fork on that to eliminate release pressure on the bearings. Perhaps I will do that. This is a first prototype so it is 'legal' for me to change my mind 50 times as I go thru the build . . ;)
-
Ah Ok i see what you are saying now. The ones I got for my mill were direct from Mubea and were very nice and smooth. I have just looked back to see how much they were and for the size I need, 23mm OD x 10.2mm ID x 0.9 mm Thickness it cost £7.84 per 100 (min quantity) , think I needed about 112 but ordered 350 as a friend also wanted some for two of his machines.
Have looked at Mubea and its phosphated and oiled finish.
Yes I often thought that about a manual (or impact power drawbar) on the Bridgeports and the torture the bearings got but it didnt seem to harm them.
Hood
-
Heres a pic of the ones I had extra, bit dirty with sitting in the drawer but you can see they are decnt finish.
Hood
-
The ones I got for my mill were direct from Mubea and were very nice and smooth. I have just looked back to see how much they were and for the size I need, 23mm OD x 10.2mm ID x 0.9 mm Thickness it cost £7.84 per 100 (min quantity) , think I needed about 112 but ordered 350 as a friend also wanted some for two of his machines.
Have looked at Mubea and its phosphated and oiled finish.
Wow, dirt cheap. Is that a price recently or 10 years ago? Your springs are not that far from mine, except less than half as thick, which would explain twice as many, perhaps. I would speculate that you have the springs stacked a couple in each direction and not alternating each spring? I went with the thicker spring to eliminate the spring to spring friction inherent in 'nesting' springs in the same direction. I have gotten numerous sizes if Bellevilles that were nicely finished, but always thinner springs like yours. These are the thickest springs I have ever worked with (compared to the small diameter), and also the roughest. Live and learn, as they say.
-
It was the 29th March last year I ordered them.
Mine are stacked in 4's with the first and last set facing their respective seats.
Attached is a pic of the drawbar with the original springs, probably 25 or so years old, they were still fairly good but had lost some of their tension.
Hood
-
The springs I got from mcmaster were very smooth. Almost like they were run through a vibratory finisher. They weren't shiny like yours but a dull grey. My springs are .090" thick so maybe you crossed a threshold thickness wise of what could be punched smoothly. Or their dies were worn out. I'm still thinking that your idea of using a valve spring is the way to go. My rig is working really well but I still plan to give the valve spring idea a whack.
The spindle looks amazing by the way.
Derek
-
The springs I got from mcmaster were very smooth. Almost like they were run through a vibratory finisher. They weren't shiny like yours but a dull grey. My springs are .090" thick so maybe you crossed a threshold thickness wise of what could be punched smoothly. Or their dies were worn out. I'm still thinking that your idea of using a valve spring is the way to go. My rig is working really well but I still plan to give the valve spring idea a whack.
Mine are also .090 thick. Worn out dies is probably the cause, as you said. In hindsight, I should have just returned them and purchased elsewhere, but having no experience with spring stacks, I did not anticipate the problems . Incidentally, the springs were black, the shiny surface is from me polishing the stupid things. They should work fine now, but it was just a lot of unnecessary work to get there. I hate having my time wasted . . . probably more than anything else.
What part number did you purchase, if you don't mind sharing that info.
Next round I will go with the coil spring, but I only need one spindle, so unless somebody buys this one, I will be using it and not making any more.
-
I used 9712K84
Next round I will go with the coil spring, but I only need one spindle, so unless somebody buys this one, I will be using it and not making any more.
That's how I feel about what I have as well. When the carousel is done and it's time to change out the Bellevilles I'll probably revisit the valve spring idea.
Derek
-
The reason I asked about retaining the lower bearing is that you will be pushing down on the spindle each tool change and over time and heat the housing may expand enought to loosen up the press fit and cause the spindle to drop. Most spindles I have seen had a screw in retainer ring to preload and lock the bearing into the housing. Preload in this case is taken care of . Locking it into the housing under preload may be something to think about.
Just a thought, (;-) TP
-
The reason I asked about retaining the lower bearing is that you will be pushing down on the spindle each tool change and over time and heat the housing may expand enought to loosen up the press fit and cause the spindle to drop. Most spindles I have seen had a screw in retainer ring to preload and lock the bearing into the housing. Preload in this case is taken care of . Locking it into the housing under preload may be something to think about.
Just a thought, (;-) TP
A very good thought. In case you think I don't listen to peoples ideas, your comment earlier on this topic got me thinking and I realized that while the design will work OK with the current bearing, it would not be ideal for an ABEC 7 set, which the design is supposed to accommodate, so have modified the design by splitting the labyrinth plate into two pieces. The outer one (the original one that is in the photos ) is now bolted to the housing to . . <drum roll> . . retain the bearing.
A new inner ring will overlap the outer ring to create the labyrinth seal. So, preload or no preload on the upper bearing, the lower bearing is retained independently.
The original plate (now the retainer) is completed except for hogging out the middle. I am making the new inner ring now. Off to the gym now, so no more pictures until tomorrow.
-
OK, two more pics and we're done. Here are the bearing retainer parts and the assembled spindle nose. All that is left are a couple of drive dogs. Anybody know if the dogs are hardened on a typical BT30 spindle?
I will need drive pulleys and a disc brake rotor for the top of the spindle, but I am note sure what arrangement I want for those yet, so I suppose the power draw bar is up next.
(http://www.thecubestudio.com/BT30SpindleProject/RetainerPartsWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/RetainerAssembledWEB.jpg)
-
Steve,
It's a thing of beauty! A few things I'm not clear on:
Are the 4 Philips screws what retains the lower bearing?
What are the two "dimples" on the inner perimeter of the (assumed) bearing retainer?
Don't know if BT30 dogs are hardened, but it certainly would not hurt. I wouldn't think it would be really necessary, however. In theory, there should be no movement between the tool holder and spindle, and even if there is, it'll be once per tool, and the dogs will make contact, then should not move again until the tool is removed and re-installed.
Can't wait to see what you come up with for a PDB!
Regards,
Ray L.
-
Steve,
It's a thing of beauty! A few things I'm not clear on:
Are the 4 Philips screws what retains the lower bearing?
Yes, the screws, in part, retain the bearing. The screws shown are temporary until I grab some socket flat heads in that size. You know how it goes, I have ever size above and below . . .
I say 'in part' because there is also a press fit as well as being 'glued' in with Loctite bearing retainer.
Edit: oh yes, I almost forgot: mechanically there were also already pins thru the housing below the bearing.
However, taken alone, and using only rough numbers because we are not even close to the wall, the bearing is rated around 10k and 4 of the proper fasteners will resist about 9.5k, so even without the press fit, glue, or preload on the top bearing, the lower bearing would be toast by the time the screws popped.
What are the two "dimples" on the inner perimeter of the (assumed) bearing retainer?
What hump?
kidding . . . those are just the remnants of the spanner pin holes. What is now the 'retainer' used to be the full plate with the labyrinth seal at the edge, so it had spanner pin holes for 'wrenching'. A new smaller plate takes its place, this time with the seal between the outer and inner disks.
-
Beautyfull job Steve, (;-)
(;-) TP
-
HIYA Steve I don't know how you are going to compare price wise but this is about what you are building.
http://www.tormach.com/store/index.php?app=ecom&ns=prodshow&ref=30505
(;-) TP
-
HIYA Steve I don't know how you are going to compare price wise but this is about what you are building.
http://www.tormach.com/store/index.php?app=ecom&ns=prodshow&ref=30505
(;-) TP
I have been aware if the Tormac spindle for a long time. It was quickly eliminated for two reasons.
1) it is toy sized. I know it is difficult to tell from a photo, but for scale compare the BT30 taper to the housing size on my spindle and the Tormac.
2) the Tormac has only a drawbar and no option for anything else.
Price? I don't think the Tormac spindle is equivalent to what I built. The price of the Tormac would get you the ABEC 7 bearing set for a 55mm shaft. I did a lot of bargain hunting and there is still more than the price of the Tormac just in parts and materials in my spindle. I would have built a new spindle long ago except for the $500 stud gripper. I have mentioned this before, asked about it, done extensive searching, best I eved did was I found one for about $400, and remeber this is just for the gripper fingers, nothing else. I never found a used gripper and never found a ball type gripper new or used.
Probably I will not produce the spindle so I have not worked up a price, but it would be more than the useless Tormac toy and less than TECO or equiv precision sindles. I would sell the prototype for a reasonable offer only because I am contemplating going with a larger adapter size and it is not possible to recut the spindle for a larges adapter. Once I start using the BT30 spindle, I will be collecting more and more BT30 tooling which will make it a lot more expensive to switch to a larger size down the road. On the other hand it is always difficult to sell my prototypes and I have not listed the spindle, nor my 4th axis prototype. I know it is idiotic, but they are like my babies and I always want to keep everything. :'(
-
Couldnt agree more, the difference between the two is night and day.
Hood
-
AH ok so what you really wanted was a 40 series spindle size. Now I understand. SHould be plenty of used 40 series spindle assy out there that just need a rebuild.
(;-) TP
-
I don't just want a 40 series spindle, and I cannot use an off-the-shelf spindle. It is not that simple.
What I want is an ATC. The spindle is just a piece of a much larger puzzle, as is the PDB. Actually, the ATC itself is part of a larger project. The concept of an ATC is not limited to swinging adapters around. A tool turret is also an ATC, in my view, and I am contemplating having both. Although in the 'mini-machining-center' configuration, the tools in the mill spindle can be single point tools for turning operations (see image below), my InTurn™ 4th axis' unique turning capability would not be fully served by a convectional milling machine ATC. The 4th axis wants tools pointing at it. That is why I did the Horizontal Spindle project. I need the big spindle because it will hold not only single cutting tools for the 4th axis, but potentially also a tool turret. I have the static loading capability that I need, but BT30 is starting to look maybe a bit too small to hold a tool turret, There are other solutions to that problem, if indeed it becomes a problem, but most would require a new spindle design with a nose specifically capable of holding a turret, either by a larger adapter or a mount on the spindle shaft itself.
Hopefully that clears up some of the mystery.
(http://www.thecubestudio.com/BT30SpindleProject/SpindleTurnInitialWEB.jpg)
-
YES that clears up some of the mystery.
Will this thing make coffee in the morning as well(;-).
Have you considered just adding the turret mechanism to the spindle housing and just index the spindle then relock the turret solid each tool change. You could add the index slots to the spindle housing then use the Power draw bar to lift the turret off the slots then use the spindle motor to rotate the turret then release the power draw bar to engage the slots again to lock it down.
That way you do not need a huge servo OR brake mechanism JUST to hold the turret in position
When NOT in lathe mode the spindle then uses a standard BT30/40 tool holder. IF you are going to move up to 40 you might want to consider Cat40 as they are more common in the used market than BTs.
Just a thought, (;-) TP
-
Will this thing make coffee in the morning as well(;-).
No, but there is a Margarita adapter for the ATC. :D
Have you considered just adding the turret mechanism to the spindle housing and just index the spindle then relock the turret solid each tool change. You could add the index slots to the spindle housing then use the Power draw bar to lift the turret off the slots then use the spindle motor to rotate the turret then release the power draw bar to engage the slots again to lock it down.
When you say spindle 'housing' I think you mean the spindle itself? If so, then yes, attaching to the spindle nose may be the way to go. I really wanted to avoid having the spindle stick down past the bottom surface of the head, but that may be what I'll need to do.
That way you do not need a huge servo OR brake mechanism JUST to hold the turret in position
You have a knack for making suggestions right AFTER I have finished building something . . :( However, notching would not be adequate for the spindle because fine adjustments are needed for the tools. Something that is extremely cool is to cut up to a shoulder in one direction and then just rotate the tool and cut up to another shoulder in the opposite direction. Saves a tool change! But the rotation is not exactly 180 degrees, hence notching would be an impediment. Besides that, I have already added a substantial rotor (see photos) and will use one of the 4th axis calipers with it. You can't move the 4th axis spindle or the mill spindle when locked, trust me on this.
When NOT in lathe mode the spindle then uses a standard BT30/40 tool holder. IF you are going to move up to 40 you might want to consider Cat40 as they are more common in the used market than BTs.
The jury is still out on this question. I'm working on other stuff right now so I will revisit this question later and consider all suggestions. I have completed the PDB design. Quite unique I will say . . extremely compact. It mounts on the head, so I will need to mock up a head to do any testing. Hood expressed that he liked that on his mill a lot of stuff follows the head around, I have that as a design goal. We shall see.
Incidentally, 'your' bearing retainer has come in handy as a leg up on incorporating an oil seal. SKF has some contact seals good for about 7k RPM, so I am going to use that instead of the labyrinth with the current bearing because I do quite a lot of grinding. I have added a positive pressure purge to the housing already, but that can just be plugged. The labyrinth and purge is still necessary for high RPM with the precision bearings though, so it is good to have also.
Spindle lock rotor:
(http://www.thecubestudio.com/BT30SpindleProject/SpindleLockBottomWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/SpindleLockTopWEB.jpg)
-
To adjust the toolpoint on center you always have the Y axis available to adjust it.
Just a thought, (;-) TP
-
Drawbar Actuator is completed. Met and actually far exceeded all design goals.
Specs:
Travel: .25" (selectable)
Working Downforce @ 100psi with .25" travel: 5,281lbs (in service, this will be adjusted down to about 1,500lbs via pressure reg)
Max Downforce: 9,127lbs
Apply time: < 1 sec (configurable via tubing size or orifice, can be almost instantaneous, but imposes shock load)
Retract time: approx 2 sec (configurable as above, soft retract is intentional to limit shock load)
Electric Solenoid operated positive mechanical lockout. Actuator cannot release drawbar unless solenoid is activated first to withdraw mechanical interference.
Currently presses against spindle bearings (top bearing if pre-loaded). Can be installed in a floating configuration and forked to whatever is available on the spindle for zero bearing load. This actuator generates nearly zero shock load so bearing loads would be the compressed spring force only.
Overall dimensions installed orientation:
Length 9.35"
Width 4.56" (solenoid adds approx .65" to this dim on one side)
Height 4.49"
(http://www.thecubestudio.com/BT30SpindleProject/ActuatorAssembledSideWEB.jpg)
-
Steve,
I guess that should be enough force. :-)
So..... I'm guessing you have a two-stack 4" cylinder, generating about 2500#, with a 1:2 cam arrangement pivoting on the shaft in the lower RH corner of the photo?
I'm curious why the release is so slow? Is it a single-acting cylinder, and that's just the bleed-down time?
-
Steve,
I guess that should be enough force. :-)
So..... I'm guessing you have a two-stack 4" cylinder, generating about 2500#, with a 1:2 cam arrangement pivoting on the shaft in the lower RH corner of the photo?
Close. Two stack 3", 4:1, no pivot.
Compactness was a design goal. Actually, I did three complete designs using different methodology and selected this one as the best all around.
I'm curious why the release is so slow? Is it a single-acting cylinder, and that's just the bleed-down time?
Double acting. I added an edit to explain that this is configurable and I have it set conservatively. Essentially instant speed is available, but the cost might be high, so other than torture testing, I will want to run it as gently as is practical in actual service. I have a restrictor on the exhaust.
-
I guess that should be enough force. :-)
Some pretty outrageous drawbar force numbers have been batted about for the big mills. It is hard to know what the facts are because the 'specs' never come with any references, but in any case, I am not developing these machines just for benchtop mills. Everything can be used in 24/7 production shop floor. The actuator can easily push 5k with a 2:1 safety factor. There is an intentional (and easily replaceable) weak link that fails just under 10k (hence the max force spec).
-
Ok you have my curiousity up. IF the cylinder is 90 deg to the pushrod how do you actuate it without a pivot?? What is the pin in the picture for??
I can think of a few ways but just curious as to what you picked.
(;-) TP
-
Spindle completed. Actuator completed except for lockout solenoid which is on order.
I made up a test rig from scrap that duplicates part of the future head. The initial test was added to the previous fragment. and can be seen here. If you already say the first fragment, skip to 00:40 sec.
That wraps up the spindle and PDB. The ATC is next, but I have had some other stuff come up so it will be a while before I can resume this project.
A big THANK YOU to all who participated in this thread. :) :-*
www.thecubestudio.com/BT30SpindleProject/BT30SpindleTestRig.wmv (http://www.thecubestudio.com/BT30SpindleProject/BT30SpindleTestRig.wmv)
-
Excellant Job Steve you make us proud (;-).
(;-) TP
-
Looks like its going to be the dogs danglers :)
Hood
-
Update: I have settled on an arrangement for the ATC and completed the preliminary design.
So far, I have built one prototype tool storage 'pod' (for lack of a better term) for testing and while it works as expected, there *might* be a tendency for the release mechanism to jamb in the long term, so I have made some changes, adding to reliability but also to the cost, as is usually the case.
After reviewing a lot of existing ATC configurations, both DIY and commercial, I chose to not move the entire tool storage device and instead use a single swing arm to transfer tools from storage to the spindle. here are some design goals which are met with the design thus far:
1) safety - accommodate sensors at each step in the process for monitoring purposes.
2) configurable - choice of 6 tool turret or unlimited number of tools via chain belt
3) self contained - optionally the ATC (with turret only) can be mounted entirely on the mill head (credit to Hood for this goal),
4) Tool tapers are completely covered and protected by the tool storage 'pods'.
5) positive mechanical retention of the tools in the storage device - no reliance on material spring - i.e. nylon 'fingers' or equiv.
6) tool release from storage pod is initiated by the swing arm 'claw', and then only after the claw has firm control of the tool.
7) fast - have not quantified this yet, but definitely not 10 to 20 seconds as typical DIY changers require.
Feedback always appreciated, especially if you have a favorite ATC arrangement and can cite the advantages. I will post a photo of the prototype storage pod a little later today perhaps.
-
Sounds like it will be nice, are you having a dual sided arm or just using a single arm?
,
I am likely not going to have to bother getting the knee on the Beaver servo'd in preparation for a toolchanger, now as I will likely be buying a Chiron FZ12 Mill shortly, going to see it on Sunday so I can see if I reckon the motors I have will fit. Now they are fast, something under 1 second toolchanges. ;D
Hood
-
Sounds like it will be nice, are you having a dual sided arm or just using a single arm?
,
I am likely not going to have to bother getting the knee on the Beaver servo'd in preparation for a toolchanger, now as I will likely be buying a Chiron FZ12 Mill shortly, going to see it on Sunday so I can see if I reckon the motors I have will fit. Now they are fast, something under 1 second toolchanges. ;D
Hood
Single arm mounted on the head. The swing is over 180 degrees making articulation rather complicated, so the arm will probably be servo powered. Expensive option, but one that also provides for accurate positioning and also for sensing jams, collisions, or empty tool pods.
The arm will raise and lower on a single linear slide via pneumatic cylinder. I am unsure at this point where to put the sensors for vertical movement.
I doubt there will be sub second tool changes, but 5 seconds would be great.
The tool change speed will be improved significantly in a configuration where the turret is carried on the mill head, but with a tool change possible at any Z position, special attention will need to be paid to tool clearance in the g-code to make sure there is room under the head for the new tool to swing in. Have not looked at that Gorilla yet.
-
Looking forward to seeing what you come up with! I'm finally getting off my lazy a$$, and getting back on my ATC - a 10-tool carousel that will handle TTS initially, and, hopefully, eventually move to 30-taper. I considered a "remote" carousel and transfer arm, but it seemed to get too complex too quickly. My current carousel design is really quite simple, with only two air cylinders for actuation, also fully inter-locked. I'm using a Geneva driven by a simple gearmotor for tool selection. With luck, I should have it working in another week or so.
Regards,
Ray L.
-
On the topic of 'Any-Z-position-tool-changes', it has just occurred to me that if there is access in MACH to the tool table, the operator could enter the highest point on the workpiece stock and the ATC controller could raise the head to safe level for the commanded tool after checking the available Z travel. If the travel is exceeded, the machine could fault or suspend and wait for operator intervention.
-
The toolchanger I was planning was similar with the exception that the arm was going to be dual sided and it could swing round take a tool out, drop swing 180 and raise and place the new tool in. It would really have neeed three levels though as the toolchanger would be higher.
Regarding the sensors, the Chiron has them on the cylinders, there is a cylinder for each tool, each having a prox sensor top and bottom and then two more cylinders either side that lower and raise the basket. It looks at first a very complex set up but when you study it is actually quite simple and seems reliable as the only thing my friend has had to do in the 15 or so years he has had it is replace pipes as the originals started splitting.
Heres a vid of the style on the Chiron.
http://www.youtube.com/watch?v=97KQ5DKSuNw
Hood
-
I considered a "remote" carousel and transfer arm, but it seemed to get too complex too quickly.
Agreed. However, the goal of unlimited tools eliminates a moving storage device as a possibility. Moving a carousel is easy enough, as you correctly point out, but moving a 6 or 7 (or more) foot long chain belt and associated drive mechanism would be impossible by any practical method that I can think of. Another consideration on this aspect is the cover s for the tool storage. Since the arm moves up and down, I think I can sidestep needing an opening access door . . . . hopefully . . maybe . .
My current carousel design is really quite simple, with only two air cylinders for actuation, also fully inter-locked.
how are you sensing a successful tool change
I'm using a Geneva driven by a simple gearmotor for tool selection.
Simple and effective for a carousel. NG for the chain belt option, but a servo drive could be moved over with out much difficulty and the programming would not need to be significantly different . . perhaps even handled by a config param.
With luck, I should have it working in another week or so.
A week! Jeez, It will take me a week to figure out how this new coffee maker works . . :D
-
Heres a vid of the style on the Chiron.
http://www.youtube.com/watch?v=97KQ5DKSuNw
Hood
THAT is the single coolest toolchanger EVER! :-)
Regards,
Ray L.
-
Its something else when you see it in the flesh ;D
Hood
-
I considered a "remote" carousel and transfer arm, but it seemed to get too complex too quickly.
Agreed. However, the goal of unlimited tools eliminates a moving storage device as a possibility. Moving a carousel is easy enough, as you correctly point out, but moving a 6 or 7 (or more) foot long chain belt and associated drive mechanism would be impossible by any practical method that I can think of. Another consideration on this aspect is the cover s for the tool storage. Since the arm moves up and down, I think I can sidestep needing an opening access door . . . . hopefully . . maybe . .
My current carousel design is really quite simple, with only two air cylinders for actuation, also fully inter-locked.
how are you sensing a successful tool change
By the carousel position. The whole carousel drops down, moves under the spindle, then raises up to insert the tool. If the tool does not correctly enter the spindle, the carousel will either hang up, and throw a fault, or the tool "fork" will break off, which will not be detected, but should be both VERY rare, and relatively harmless (I hope).
I'm using a Geneva driven by a simple gearmotor for tool selection.
Simple and effective for a carousel. NG for the chain belt option, but a servo drive could be moved over with out much difficulty and the programming would not need to be significantly different . . perhaps even handled by a config param.
Actually, no reason why a Geneva can't be used to move a belt just as well....
With luck, I should have it working in another week or so.
A week! Jeez, It will take me a week to figure out how this new coffee maker works . . :D
Well, I started on it a couple of months ago, but it's been on the shelf for a while. I got the geneva, carousel disc, and all tool holders done and working - that all took just one day to machine. I just have to make the mounting arm and housing.
-
Certainly is an advantage to a double sided swing arm. I dought you will make fast tool changes with a single side arm . You have to get the old tool put it in the slot then rotate to the new tool and load it.
The double allows hot swapping the tools, it is as fast as it gets.
Super fast tool changes always add complexity to the tool changer. Thinking out loud when the machine is running in auto what is the difference between 5 secs and 35 secs ?? It is no longer YOUR time being consumed with teh TC. You can be off drinking coffee(;-).
Just a thought, (;-) TP
-
Actually, no reason why a Geneva can't be used to move a belt just as well....
Geneva drive for a belt is doable, but lacks high speed random access, needing to stop at each position in sequence, including empty pods. With an intelligent drive such as the one described by Hood, you can simply command the drive to go to a spot and not have to track the position as you would with a Geneva.
I also wonder if it might be possible (in a practical sense), to have variable width tool 'slots' on a chain belt. i.e. why have the same real estate consumed for a 1/4" drill bit as a 6" face mill? I have not given this much thought yet, but if it is doable in some reasonable way, it could make the difference between getting 20 tools on getting 25 tools on the same belt.
-
Heres a vid of the style on the Chiron.
Friggin' AWESOME! Straight out of Star Wars. . . . . I think I got some wood watching that vid . . (not sure how that will translate internationally . .LOL!)
Anyway, given a 50k budget . . . and a staff . . I might be able to come up with something like that . . . . however, if I fall somewhere between that monster and Tormach I shall be well pleased.
-
Regarding the sensors, the Chiron has them on the cylinders, there is a cylinder for each tool, each having a prox sensor top and bottom and then two more cylinders either side that lower and raise the basket. It looks at first a very complex set up but when you study it is actually quite simple and seems reliable as the only thing my friend has had to do in the 15 or so years he has had it is replace pipes as the originals started splitting.
Heres a vid of the style on the Chiron.
http://www.youtube.com/watch?v=97KQ5DKSuNw
Hood
That's one cool tool changer. Definitely the fastest I have seen.
If the basket can't rotate then setup is not an easy task, trying to reach the tools on the rear.
Dan
-
Certainly is an advantage to a double sided swing arm. I dought you will make fast tool changes with a single side arm . You have to get the old tool put it in the slot then rotate to the new tool and load it.
Certainly a valid point. I looked at a dual arm . . . longingly . . . salivating . . however, while both arms work, they also both swing and a 360 degree sweep is required. A single arm requires only about 200 (approx). There just was no enough room for a full sweep and still be mounted on the mill head, so I had to choose.
Super fast tool changes always add complexity to the tool changer. Thinking out loud when the machine is running in auto what is the difference between 5 secs and 35 secs ?? It is no longer YOUR time being consumed with teh TC. You can be off drinking coffee(;-).
So . . . is fast important or not important? ??? With my planned arrangement, there will not be anything approaching 'super fast', but as you point out, it will also not be super complex nor super expensive. So between a Smart Car and a Lamborghini, I'll be happy to end up with a Camaro . . although actually I like the Challenger much better.
Making 30 parts in one hour instead of 25 parts would be significant enough for my purposes. I only have one product that I make a few hundred at a time.
Just a thought, (;-) TP
[/quote]
-
Steve,
Like Terry, I too question the real-world value of a super-fast toolchange time. I can't imagine that it would save more than a few minutes per day at the cost of significantly increased complexity, and opportunity for disaster, due to higher speeds, and generally higher precision required. I think you can quickly get to the point where doubling speed will take 10X the time and 2X the cost, to make it as reliable as a slower, simpler machine.
Re: the speed of a Geneva, this, too, I think is a bit of a red herring. Perhaps a high-speed Geneva will always be a bit slower than a max-performance direct-driven mechanism, but I think the overall difference in toolchange time would still be small. I'm running mine VERY conservatively, and it runs under a second per tool pocket. So, even with a 24-tool carousel, max seek time would be 12 seconds. I'm sure I could easily double that. Beyond that point, you have to start worrying about the tools flying out as the thing moves. The biggest job I ever did only used about 12 tools, and that one took hours to run, so an additional 3 minutes in seek time would be lost in the noise.
Certainly doing a variable-tool-size chain-driven machine is quite do-able, but, again, at considerable cost in terms of complexity. You'd have to not only program each tool length, but now the "pocket" width as well. And, setting up a job will mean a fair amount of disassembly/re-assembly to move the tools to their required positions. Or, have pre-configured slots of different sizes, but then you have to deal with mapping tool numbers so each tool lands in an appropriately-sized pocket. My biggest tool is a 4" face mill, so I just sized all the pockets accordingly.
Regards,
Ray L.
-
Heres a vid of the style on the Chiron.
Friggin' AWESOME! Straight out of Star Wars. . . . . I think I got some wood watching that vid . . (not sure how that will translate internationally . .LOL!)
Anyway, given a 50k budget . . . and a staff . . I might be able to come up with something like that . . . . however, if I fall somewhere between that monster and Tormach I shall be well pleased.
Ha ha yes it translates almost word for word, seeing it on you tube is cool but seeing it when standing right in front is amazing and to think they have been doing that same toolchange since at least the mid to late 80's, the one I am likely getting is I think 1987 or 88.
-
That's one cool tool changer. Definitely the fastest I have seen.
If the basket can't rotate then setup is not an easy task, trying to reach the tools on the rear.
Dan
It doesnt rotate Dan and although I havent seen a tool holder being changed I would imagine you just call an arm to drop but not call the main cylinder to raise up.
Hood
-
Heres a vid of the style on the Chiron.
http://www.youtube.com/watch?v=97KQ5DKSuNw
Hood
Very cool ! I saw these at IMTS.
I especially like the part load/unload gripper set up as a tool.
Thanks Hood,
Russ
-
Steve
Heres a site that gave me ideas when I was dreaming about how I was going to do mine, now I dont have to (assuming I buy the Chiron ;D ) May give you some food for thought, I especially like the hinge down holders.
http://www.miksch.de/en/tool-changing-systems.html
Hood
-
Very cool ! I saw these at IMTS.
I especially like the part load/unload gripper set up as a tool.
Thanks Hood,
Russ
Yes and the beauty of the basket changer is nothing rotates round so having air lines to things liike grippers dont pose a problem.
Hood
-
Steve,
Like Terry, I too question the real-world value of a super-fast toolchange time. I can't imagine that it would save more than a few minutes per day at the cost of significantly increased complexity, and opportunity for disaster, due to higher speeds, and generally higher precision required. I think you can quickly get to the point where doubling speed will take 10X the time and 2X the cost, to make it as reliable as a slower, simpler machine.
Re: the speed of a Geneva, this, too, I think is a bit of a red herring. Perhaps a high-speed Geneva will always be a bit slower than a max-performance direct-driven mechanism, but I think the overall difference in toolchange time would still be small. I'm running mine VERY conservatively, and it runs under a second per tool pocket. So, even with a 24-tool carousel, max seek time would be 12 seconds. I'm sure I could easily double that. Beyond that point, you have to start worrying about the tools flying out as the thing moves. The biggest job I ever did only used about 12 tools, and that one took hours to run, so an additional 3 minutes in seek time would be lost in the noise.
I'm not sure if our posts 'crossed in the mail' or if you misunderstood. My setup will not be 'super fast', just fast. And even that is a subjective term. I think the only number I mentioned was 5 seconds, but really I won't know what it will tolerate until it is running and I can push it to failure. What I don't want is 20 second tool changes. Seconds add up when you are multiplying them by orders of magnitude. If you doubt that, just run Vectrix generated peck drilling code for 30 deep holes . . . . ::)
Certainly doing a variable-tool-size chain-driven machine is quite do-able, but, again, at considerable cost in terms of complexity. You'd have to not only program each tool length, but now the "pocket" width as well. And, setting up a job will mean a fair amount of disassembly/re-assembly to move the tools to their required positions. Or, have pre-configured slots of different sizes, but then you have to deal with mapping tool numbers so each tool lands in an appropriately-sized pocket. My biggest tool is a 4" face mill, so I just sized all the pockets accordingly.
The jury is still out on this and I have not really thought it thru. Thanks for you observations. Every comment helps.
The idea probably falls apart without a smart servo drive that can find the tool spots on its own. I had in mind just a certain number of 'small' spots and 'wide' spots. You put your 4" face mill in an existing wide spot and your 1/4" drill in a 'small' spot. The only disassembly would be to alter the ratio of small to large to suit your typical jobs. The position of the tools can be defined as a distance from 'home' and could be calculated easily thus; 3 wide spots x 4" and 2 small slots x 3"; that tool is at 18". All you would need is a table of tool positions stored in the drive or the controller.
It would just be a way to shoehorn a couple more tools on a belt which may be all you need for a specific job. If I build in the capability, that doesn't mean it has to be used. Even if the slot width was variable, they could all be set the same, so no harm no foul . . . . IF it turns out to be easy to implement. If it does result in a lot of complexity, then I would agree with you that it may not be worth the effort, but I don't know that yet. Certainly the tool postions could be completely random and be probed, but that is most likely over the doable-but-not-practical line.
The chain belt option is a definate, but variable spacing is just a discussion point at the moment.
-
Steve
Heres a site that gave me ideas when I was dreaming about how I was going to do mine, now I dont have to (assuming I buy the Chiron ;D ) May give you some food for thought, I especially like the hinge down holders.
http://www.miksch.de/en/tool-changing-systems.html
Hood
Couple things, Hood;
First, thanks for the link. But a full half second for a tool change? Please . . . way to slow . . :D OK seriously cool stuff. This is pretty much similar to what I have in mind . . except on a whole different level.
Second, on their 'magazine', they seem to be linking individual tool 'pods' to each other rather than attaching them to a separate chain . . interesting . . and so much for the variable slot width idea . . :-\
Third, since you won't need that old Beaver mill, you can just put it in a big box with a ribbon . . my address is . . . . . ;D
-
Ha ha well afraid I like my Balding Beaver so its staying. Have to get rid of the Bridgeport and one of the manual lathes to make room though. Sad day in a way as I had to sell my CBR600 to buy the lathe when I was starting out and kind of have a soft spot for it even though I rarely use it.
Hood
-
Second, on their 'magazine', they seem to be linking individual tool 'pods' to each other rather than attaching them to a separate chain . . interesting . . and so much for the variable slot width idea . . :-\
Could you not make up double width pots for such situations?
Hood
-
Could you not make up double width pots for such situations?
Hood
Now that's creative thinking! 8) I suppose you could indeed . . so long at the 'double wide' could get around the corners.
-
Could you not make up double width pots for such situations?
Hood
Now that's creative thinking! 8) I suppose you could indeed . . so long at the 'double wide' could get around the corners.
Surely a man of your proven capabilities could make it hinge ;D After all the pot itself will be the same size, just the lugs out each side would be longer.
Hood
-
Picture worth thousand words:
This it the first prototype 'pod'. There would be 6 (or more) of these bolted to the outside of the turret wheel or to the chain belt
The body survived the cut, but the lever is being replaced by a modified design to be more 'jam proof' . . most likely the coil spring will be replaced as well. Note that the tool taper is completely covered and protected and that the tool cannot release from the pod without the catch being retracted (by the claw on the arm).
The pod body is a little complex, but the entire part; OD turn, flat one side, drill and counterbore two 1/4" mounting holes, center drill, 3/4" drill thru center, taper bore center, drill and tap latch pivot mounting hole, can be done with a single setup and a single program on my InTurn™ 4th axis . . provided I had a working ATC . . . Catch-22.
(http://www.thecubestudio.com/BT30SpindleProject/PodFrontWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/PodLatchWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/PodTaperWEB.jpg)
-
Looks a nice design however may just throw out another idea. My friends Beaver mills changer uses balls and a small spring in a similar fashion to your drawbar. There is a small air cylinder above the tool release position so when the correct tool is in position and the clamp is gripping it the cylinder pushes down on the spring and releases the pullstud, I thought it was quite a nice idea and the spring pressure needed is small so any old valve spring would be more than enough.
Hood
-
Heres a video of a similar setup that CNCIT on the forum has oon his Beaver mill, you can quite clearly see the cylinder/spring setup.
Hood
-
Looks a nice design however may just throw out another idea. My friends Beaver mills changer uses balls and a small spring in a similar fashion to your drawbar. There is a small air cylinder above the tool release position so when the correct tool is in position and the clamp is gripping it the cylinder pushes down on the spring and releases the pullstud, I thought it was quite a nice idea and the spring pressure needed is small so any old valve spring would be more than enough.
Hood
Very interesting. This sounds like it would be a bit more complicated to build and certainly additional cost (so what else is new), but it would solve the *potential* jamming concern. The claw could trip the air cylinder, thereby retaining the automatic 'interlock'. Hmmmm
BTW, it is BR's job to make good suggestions right AFTER the part has been completed . . if you HELP him . . I'll never finish anything . . :D
Hey wait a minute . . did you just move me from one PDB to SEVEN ??!! :o
-
Heres a video of a similar setup that CNCIT on the forum has oon his Beaver mill, you can quite clearly see the cylinder/spring setup.
Hood
Oops . . no link.
-
Seven, thought you were going to have a 40 tool magazine ;D
Hood
-
Ha ha ok dope slap needed.
http://www.youtube.com/watch?v=t45eE6aOWrI
Hood
-
Ha ha ok dope slap needed.
http://www.youtube.com/watch?v=t45eE6aOWrI
Hood
That video is so close to what I have in mind I can hear the Twilight Zone theme music in my head. The claw does not extend, and the carousel would not need to stop at each tool, . . and it won't be yellow . . . but otherwise this is extremely close to what I will build.
I've been sitting here doodling and already have a prelim design for a simple ball catch. Very interesting possibility . . . . but 40 ball grippers? :'(
-
. . . but 40 ball grippers? :'(
Better than having two ball grippers, that may hurt ;D
Hood
-
The ball gripper pod design went fast at first, but bogged down when run against the design goals. . . positive mechanical lockout and sensors turned out to be a very big challenge. I am not satisfied relying on only electronic safeguards. In my view that is like substituting a proximity warning for a belt guard. Electronics can fail too easily.
After a couple of hours looking at three different methods, one of which is in the photos, everything finally came together with the ball gripper method. I figured out a bulletproof mechanical lockout and a sensor location and the thing is still buildable for less than a million dollars . :P I shall put a Scottish flag on the finished product.
This project is headed for a significant pause because I am ordering material to make a batch of the new 4th axis next week. When that arrives, I won't have time to play. The BT30 spindle is also sold and ironically, I will be making an identical spindle as I was able to resolve the turret issue. I agreed to the sale so I'm going to honor that, of course, but it will be a setback, schedule wise, to make another spindle for this project.
If I can get a new gripper equipped pod made (NOTHING from the previous prototype pod is useable now), I will post it here. I would like to get one finished so that I can do the proof concept stuff at least, but we'll see how it goes.
Thanks once again for all of the comments and ideas.
-
HIYA Steve I WAS being nice and NOT saying anything about the design else you may never finish it.
(;-) TP
-
Post some pics and I am sure we can then say it looks good but what about........ ;D
Hood
-
OR maybe you might consider ???
(;-) TP
-
Jeez! It would save time if I just post the drawings and get 'approval' before building the parts . . . LOL!
-
I have my RED pencil ready. LOL
(;-) TP
-
I am just surprised it took you so long to come to that conclusion ;D
Hood
-
Hood, do you know the spring pressure on those tool carriers? If not then the cyl dia and psi would be enough.
I found a spring that will fit the pods. As installed would be 45lb closed 72lbs open. Seems like plenty, but would be nice to have real world confirm. Probably the heaviest too I would ever have in there would be only a few pounds, and there is a positive mechanical lock preventing the gripper from moving until the claw grabs it. The claw releases the mechanical interlock after it has the adapter by the groove. Only then can the air cyl move the gripper to release the tool.
Thx.
Anyone, I did all of the calculations on the drawbar gripper, but I don't think I need to do any of that for these little guys. I'm not planning to use a sleeve or to harden anything. Just 2 steel balls running right on the 7075-T7 body and held in a 12L14 ball holder should last years I would think.
-
Well I am going through to see the Chiron tomorrow so I will measure the cylinder Dia ( if I remember ;)
Really though all you need is enough pressure that you cant easily pull it out or it rattle out with vibration so I wouldnt imagine it would have to be seriously strong.
Hood
-
REMOVE THIS POST
-
While I admire where this has gone and found some of the conversation enlightening, the chances of anyone else getting a working spindle at a reasonable price doesn't look possible and it's a shame becuase I think you've really hit on somehting good.
I kind of consider myself a spindle aficionado, I have the luxury of working with a spindle expert who does this as part of his daily routine and I've gotten pretty good at it.
Here is some of my recent work on ATC spindles, what I started last week is an advanced version of the BT15 spindle cartridge with the disc springs internal and should be done in a couple of weeks.
Yes, even a sherline BT15 ATC conversion happened in the process which consists of a shaft, petal-clamp, disc springs and drawbar utilizing timing belt/pulleys and has sufficient tool clamping force to prevent slipping at 382lbs of pulling force.
As you can see in one of the pictures, they are real petal-clamps, nitride coated for strength and durability (over-kill, I found out).
I'm fighting with some Delta VFD's at the moment (I can't figure out how to make them do orientation) but as soon as I'm done I'd like to take what I've collected here and come up with a BT30 spindle cartridge that people can buy if we can figure out what a reasonable/fair price is I should be able to have something to purchase in about 30 days.
I do have a spindle expert at my disposal and I've made several different designs myself as well as a couple of BT30 spindles that were expensive but I believe I can produce something at a reasonable price based on the cumulative experiences and knowledge.
-
Here is a quick basic image of a design with good characteristics, the mounting flange can be located anywhere on the body and isn't currently shown because I'm not sure what placement would be better suited for most applications and running off 25pcs in various locations would be costly and time consuming for a production run.
Typically the two dogs are made from 42Cr with a hardness of HRC50-55 but most people use 4140 without hardening and I've heard little in the way of distortion or wear due to usage.
If not the image, here's a drawing that shows another design that can be produced for a reasonable price.
-
Poking around for BT30 stuff and came across this complete spindle and actuator. Good comp.
http://www.ebay.com/itm/BT-30-8-000-RPM-SPINDLE-with-AIR-CYLINDER-ASSEMBLY-/300743758615?pt=LH_DefaultDomain_0&hash=item4605b99317
DWalsh: caution about posting drawings . . . the teacher has a big RED pencil . . . a corner chair with dunce cap . . . and possibly a wooden ruler . . and a paddle :o
-
Another BT30 spindle . . this one in need of TLC, but the price is right . . . .
http://www.ebay.com/itm/150874092526?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1423.l2649
-
Poking around for BT30 stuff and came across this complete spindle and actuator. Good comp.
http://www.ebay.com/itm/BT-30-8-000-RPM-SPINDLE-with-AIR-CYLINDER-ASSEMBLY-/300743758615?pt=LH_DefaultDomain_0&hash=item4605b99317
DWalsh: caution about posting drawings . . . the teacher has a big RED pencil . . . a corner chair with dunce cap . . . and possibly a wooden ruler . . and a paddle :o
Now that is interesting/funny but I produced this spindle for Greg Heyen of Servo Products Co. LTD, the hybrid pneumatic/hydraulic cylinder I get out of China and comes assembled with the switches, valves and other parts as pictured for cheaper shipped than I can buy the parts in the USA.
-
Steve,
The cylinder is 45mm OD. I took a couple of pics so you could see the setup closer, springs are not that strong and I managed to push them down fairly easily with the palm of my hand.
Hood
-
Steve,
Just for your amusement, here is the current state of my ATC - I finally got some time to work on it yesterday. It's mounted on the machine, and, operating it manually, it appears to work perfectly - the forks slide on and off the tools very nicely, and the tools slip in and out of the spindle effortlessly. I hope in the next few days to get the top plate of the housing done, which will let me mount the geneva motor, and see how that all works under power. Then, it's just a matter of mounting the two air cylinders - one long one to swing the ATC to the spindle, and back to the park position behind the table, and the second to push the carousel down for ejecting tools from the spindle. I'll also need to re-make my Y/Z way cover, to make it narrower so the ATC can get back as far as possible when parked, and re-make my coolant "collar", which right now mounts to the bottom of the spindle, and interferes with the ATC.
Regards,
Ray L.
-
Steve,
The cylinder is 45mm OD. I took a couple of pics so you could see the setup closer, springs are not that strong and I managed to push them down fairly easily with the palm of my hand.
Hood
Mr.Hood, thanks SO much! Looks like I was in the ball park with my WAG. This morning I ordered materials to make the pods and the carousel. Now I can get the springs and other hardware bits on order.
Curiosity, are you going to buy the mill?
-
Steve,
Just for your amusement, here is the current state of my ATC
Well, if I am to be amused, I must get out my RED pencil >:D . . . kidding, of course. Your work is always first class and logically conceived.
Interesting thing is that what you have is something of a hybrid between the typical moving carousel arrangements that I have seen and with what I am doing. You move the carousel . . so far fairly conventional . . . . but as always there is the 'Kabibble Twist' on it. You move it with an arm! That removes one of my objections to the travelling carousel concept . . that being the carousel hanging over the table. This is the first arrangement I have seen with an arm swinging the carousel. Very clever indeed.
Couple questions if you don't mind (not to worry no RED pencil here);
1) Another of my quibbles with the DIY setups is lack of protection for the taper (or 'stud' I suppose in your case). Do you have any plan to protect those surfaces from swarf? Do you consider that a must have or optional?
2) What kind of sensors are you planning? At what points in the process do you recon a confirmation is required before moving to the next step?
3) Probably you would get around to this eventually, but I am impatient so . . what will drive the Geneva?
4) did you make the holders in the carousel or are they a Tormach part?
5) Will you use a sparate PLC to control the sequencing or will it be integrated into your Kflop system?
6) How do you do the thumbnail - to - full size photos?
-
Curiosity, are you going to buy the mill?
Yes, decided to go for it, started a thread in the Show and Tell, looks a bit grubby but I suppose for a 1987 machine its not too bad. Mechanical wise it seems fine and really thats all I am worried about.
Hood
-
DWalsh: caution about posting drawings . . . the teacher has a big RED pencil . . . a corner chair with dunce cap . . . and possibly a wooden ruler . . and a paddle :o
Do you really see anything wrong with that drawing?
It's almost a copy of the Okuma BT30 spindle minus some features to make it cost affordable to produce and sell for a reasonable price.
I sell the BT30 spindle with matching cylinder to Servo Products for $1600.00 delivered (they are less than 15 minutes away from me).
The BT15 spindle is a work of art internally, a magnetically activated internal coolant/air valve for coolant through spindle and blowout (put a magnet on the body in one of two places and it activates the valve), PTFE high pressure seals and 60,000RPM P42 (JIS class 4 dimensional accuracy with JIS class 2 running accuracy) NTN sealed 25 degree angular contact bearing (not cheap) are all hidden inside so externally it doesn't look like much, I've made changes to the design and moved the disc springs internal without sacrificing anything so it now looks like the traditional cartridge and I should have 10 assembled in the next two or three weeks, maybe a little longer if NTN doesn't deliver the bearings on schedule.
-
Steve,
Just for your amusement, here is the current state of my ATC - I finally got some time to work on it yesterday. It's mounted on the machine, and, operating it manually, it appears to work perfectly - the forks slide on and off the tools very nicely, and the tools slip in and out of the spindle effortlessly. I hope in the next few days to get the top plate of the housing done, which will let me mount the geneva motor, and see how that all works under power. Then, it's just a matter of mounting the two air cylinders - one long one to swing the ATC to the spindle, and back to the park position behind the table, and the second to push the carousel down for ejecting tools from the spindle. I'll also need to re-make my Y/Z way cover, to make it narrower so the ATC can get back as far as possible when parked, and re-make my coolant "collar", which right now mounts to the bottom of the spindle, and interferes with the ATC.
Regards,
Ray L.
Nice piece of work Ray, well thought out and implemented.
-
Steve,
Just for your amusement, here is the current state of my ATC
Well, if I am to be amused, I must get out my RED pencil >:D . . . kidding, of course. Your work is always first class and logically conceived.
Interesting thing is that what you have is something of a hybrid between the typical moving carousel arrangements that I have seen and with what I am doing. You move the carousel . . so far fairly conventional . . . . but as always there is the 'Kabibble Twist' on it. You move it with an arm! That removes one of my objections to the travelling carousel concept . . that being the carousel hanging over the table. This is the first arrangement I have seen with an arm swinging the carousel. Very clever indeed.
Couple questions if you don't mind (not to worry no RED pencil here);
1) Another of my quibbles with the DIY setups is lack of protection for the taper (or 'stud' I suppose in your case). Do you have any plan to protect those surfaces from swarf? Do you consider that a must have or optional?
Yes, the carousel will be enclosed (a circular top cover, and 4" high side covering, so the tool holders are completely enclosed top and sides, with only longer tools exposed. There will be a rotating door to uncover the selected tool, operated by a simple cable as the carousel swings into position. I think this, alone, will suffice for keeping crud away from the tool holders. If not, it will be easy to rig a Plexi or sheet metal shield between the back of the table and the front of the carousel when it's parked.
2) What kind of sensors are you planning? At what points in the process do you recon a confirmation is required before moving to the next step?
I haven't worked out all the details, but I will be using a combination of microswitches and Hall sensors. Switches will be used to ensure the arm moves to both extremes of motion. This will allow me to stop operation if the arm jams for any reason. Also to ensure the carousel "lift" mechanism moves full travel, so I can stop operation if the tool does not move fully into, or out of, the spindle. There are Hall sensors on the Geneva to indicate it is "in position", and also one to indicate it is positioned at tool 1. This will allow me to "home" the carousel, and track position during moves. I also expect to have a hall sensor to detect a toolholder is present in the selected position, so I can tell that it correctly was loaded into the spindle, or extracted from the spindle before proceding. I think that should be enough....
3) Probably you would get around to this eventually, but I am impatient so . . what will drive the Geneva?
The Geneva is driven by a simple DC gearmotor, about $35 from McMaster. I hope to have that mounted and working no later than tomorrow. I did prototype it a few months ago, and it worked perfectly - very smooth and quiet, and very precise positioning, with near zero torque required to rotate it. The motor I'm using is not blindingly fast, but adequate - about 1 second per tool. This should make the worst-case toolchange time maybe 15 seconds,. and the more typical (sequential) probably under 10 seconds.
4) did you make the holders in the carousel or are they a Tormach part?
Made my own, though they are very similar to the Tormach (kinda have to be....). They're cheap and easy to make, and a designed-in failure point in case of a jam. They're attached with #10 Nylon screws, so a jam should simply rip the fork off the carousel, breaking only the screws. Note that I could very easily fabricate a new carousel disc, and tool forks (no more than half a day's work), and switch the whole thing over to 30-taper. No other changes should be required, other than perhaps extending the vertical travel of the carousel, which would just mean swapping out the air cylinder.
5) Will you use a sparate PLC to control the sequencing or will it be integrated into your Kflop system?
Sequencing will be controlled entirely by the KFlop for now. Until I run out of I/Os (and I'm not that far....), there's really no reason to do otherwise. Once it's working, moving the functionality to a dedicated micro would be really trivial, and I'd probably just go to a simple RS232/RS485 serial interface (as I'm already doing on my pendant).
6) How do you do the thumbnail - to - full size photos?
I don't.... I just attach the photos using the built-in function on the forum.
-
DWalsh: caution about posting drawings . . . the teacher has a big RED pencil . . . a corner chair with dunce cap . . . and possibly a wooden ruler . . and a paddle :o
Do you really see anything wrong with that drawing?
Not at all. Just a joke based on the guys tauting me a few posts back. BR was the actual target, in a sense. I assumed you had read thru there and would get the tease. First rule of plumbing and so on. Sorry it was unclear. No offense intended. :-X
-
DWalsh: caution about posting drawings . . . the teacher has a big RED pencil . . . a corner chair with dunce cap . . . and possibly a wooden ruler . . and a paddle :o
Do you really see anything wrong with that drawing?
Not at all. Just a joke based on the guys tauting me a few posts back. BR was the actual target, in a sense. I assumed you had read thru there and would get the tease. First rule of plumbing and so on. Sorry it was unclear. No offense intended. :-X
I wasn't sure if it was a pun or a serious remark.
Seems there are some talented people involved in this thread, a refreshing change.
-
Making good progress on the ATC. The whole carousel and "lift" mechanism is working now. The only major missing pieces are the air cylinder that swings the ATC into position under the quill, the side covers for the carousel, and all the "sensors". Lots of little details to take care of (like that ugly stack of cr@p on top of the air cylinder in the last photo), but I see no reason this will not work very nicely. Operating it manually, it very reliably locks onto the tool, yanks it out, and puts it back in.
Regards,
Ray L.
-
DOn't forget to build (2) (;-) I will send a BIG box.
(;-) TP
-
Here is an interesting approach to a power draw bar:
They get by with only 600# drawbar tension, but they have a clamp collar that grabs the spindle nose to transfer torque for high power tooling. R8 is a vexing problam and while it would not be useful with an ATC, I think this is a very clever solution to speed up manual tool changes. Apparently so did the patent office.
Video: (there are several, but this will get you started)
http://www.youtube.com/watch?v=YaiZC93ASJ8&feature=relmfu (http://www.youtube.com/watch?v=YaiZC93ASJ8&feature=relmfu)
Comapny site:
http://mach-1tooling.com/ (http://mach-1tooling.com/)
-
Making good progress on the ATC. The whole carousel and "lift" mechanism is working now.
Very nice.
The only advice I would offer is to put an upper support on the arm pivot pin. That is a long lever and when you get that carousel loaded with tools, it will put quite a moment load on the pin. Aluminum is not the best choice for the mount, in my view, due to the modulus of aluminum. Your mechanism moves a load vertically at the end of a long arm. This arrangement will induce significant vibration and aluminum has little ability to damp that. An upper support on the pivot pin would address both the moment on the pin and the vibration (bouncing) of the loaded carousel.
An upper support would convert the moment on the lower bracket to compression and put the upper bracket in tension. The pin force would be converted from moment to shear. Moment loads are a bitch to calculate, while shear is easy, so there is an ancillary benefit as well.
edit: one other comment; You probably already have something in mind for this, but just as an observation, the shafts running thru the arm at the carousel would benefit from bellows or a scraper type o-ring to keep the nasties (evidenced by the wall behind the carousel) from jamming the movement.
-
the trouble with R8 in a manual/semi auto change, is you need soo much more room under the tool to make the change.
The 30 or 40 taper allows you to take the tool sideways a lot earlier than the R8.
-
the trouble with R8 in a manual/semi auto change, is you need soo much more room under the tool to make the change.
The 30 or 40 taper allows you to take the tool sideways a lot earlier than the R8.
I do not know this for a fact, but my experience with the BT30 taper had led me to speculate that the R8 design has the concentric ring spaced relatively far above the taper in order to prevent the taper from racking due to inaccuracy in the drawbar straightness or thread precision.
If anyone has some reference material on the original R8 design goals, I would really love to see it.
-
Very nice.
The only advice I would offer is to put an upper support on the arm pivot pin. That is a long lever and when you get that carousel loaded with tools, it will put quite a moment load on the pin. Aluminum is not the best choice for the mount, in my view, due to the modulus of aluminum. Your mechanism moves a load vertically at the end of a long arm. This arrangement will induce significant vibration and aluminum has little ability to damp that. An upper support on the pivot pin would address both the moment on the pin and the vibration (bouncing) of the loaded carousel.
An upper support would convert the moment on the lower bracket to compression and put the upper bracket in tension. The pin force would be converted from moment to shear. Moment loads are a bitch to calculate, while shear is easy, so there is an ancillary benefit as well.
The whole mount bracket, and probably the arm, will be re-made once it's all working. I already had to modify it once, so it's a little "hacky" already. Though it is surprisingly strong just as it is. Also, the small amount of slop/flex it allows is not really a problem, as there will be a tapered pin and socket on the side of the head to ensure precise alignment of the ATC to the head. I wanted to support the pivot pin with a U-shaped bracket, but didn't have any material large enough to make that, so built what I could with what I had on hand. Most of the time, when making something like this, I'll do a quick and dirty prototype (what this is), to verify dimensions, fit and functionality, then refine the design, and re-make most/all of the parts for the final "product".
edit: one other comment; You probably already have something in mind for this, but just as an observation, the shafts running thru the arm at the carousel would benefit from bellows or a scraper type o-ring to keep the nasties (evidenced by the wall behind the carousel) from jamming the movement.
The clearances are tight enough I doubt anything could get in there, but I will do something to at least prevent accumulation of crud up there. Something as simple as a sheet-metal shield, covering the "front" (when parked) and top, would do the job. It would also be easy to fit lip seals on the top and bottom of the arm. Note the "main" shaft does not move, only the two smaller "guide rods" move.
Got a nice "bonus" - yesterday PM I spent a few hours tweaking a few of the parts, and adjusting things up carefully. I was originally afraid I might have to lose up to 1/2" of quill travel, but now have it to where I'm losing only about 1/10", and, with a simple mod to the lower support for my quill drive (just counter-sinking some screws), I can easily recover that 1/10", and lose no quill travel at all (I currently have just a hair under 5").
-
Made a bit more progress today - only had a couple of hours of shop time. I rigged up a power supply for the Geneva motor, and did several dozen complete "manually sequenced" toolchanges (pushing a button to power the Geneva, flipping a valve to actuate the lift). I see no significant problems whatsoever - it seems to work very nicely, even when fully loaded with heavy tools. I got one half of the guide pin done - a 1/2" tapered pin - mounted on the side of the head, and just need to make the "socket" part that goes on side of the pivot arm. This will ensure absolutely precise vertical positioning of the forks, and provide a hard stop for the pivot arm.
Regards,
Ray L.
-
Lots of nice spindle and tool changer activity here.
I keep waiting for someone to set up a SCARA robot to do tool changes. Ray's got half of an arm on his. One more arm and the robot is ready. It can load and unload your parts when it isn't too busy changing tools. It can make sure the coolant nozzle is pointed at just the right spot for each tool. Or, it can sweep the chips at the end of the day with a little brush, LOL.
Cheers,
BW
www.cnccookbook.com
PS Mostly, joking, but a friend has a SCARA and they're remarkably simple and useful. He uses it to move parts around on his lathe. For example, to flip them so he can machine the part that had been chucked.
-
PS Mostly, joking, but a friend has a SCARA and they're remarkably simple and useful. He uses it to move parts around on his lathe. For example, to flip them so he can machine the part that had been chucked.
Any video, Bob? Would love to see that-
Dave
-
So I have gotten lots of Kudos in this thread. However, sooner or later, we all take our 'turn in the barrel' and I thought I should reveal one of mine on this project. If I make a change to incorporate a good suggestion, I don't count that as a fail, but here is a clear SNAFU.
There are a lot . . . a LOT . . . of reads of this thread and not much participation, so I conclude there are a lot of newbees lurking around who don't speak up or ask questions. To those folks this post is dedicated. Even a well seasoned engineer does not always hit a home run each time up. Don't be discouraged if your project does not work at first. Just keep on truckin' and make it happen.
OK, moving on to the fail. Below are photos of the claw that grabs the BT30 adapters and swings them around between the carousel and the spindle nose. The challenge in this part is to get it to snap over the adapter without excessive force (which may not be available from a small servo motor), and yet have a sufficient hold on the tool to function properly and safely (#1 priority). The current prototype meets those goals, but only up to a certain arm speed. For a typical DIY project, it would be completely adequate at a nice slow pace, but this ATC is slated for 'kick ass' performance and I do not want any impediments to that goal. Well at least not designed in.
The design shown works perfectly as expected, but the hold on the tool failed testing at the speeds I intend to run the arm. Sudden extreme decel at the end of the swing (the spindle nose) can dislodge the adapter (and whatever expensive and razor sharp tool it is holding). Finding a powerful spring that could be somehow mounted on the unusually shaped claw without interfering with other operations was a very big challenge. Ultimately as you see, I cut a channel to house the heaviest spring I could find. The swinging finger on the claw snaps smartly, but is unable to retain the tool during a sudden stop. There is no real estate on the part to allow for a larger spring and the back of the claw is off limits because it 'parks' against the mill column to clear the rotating carousel.
Result: back to the drawing board. Ashes to ashes, dust to dust and scrap to scrap. The claw started as chunks of 7075 scrap and it is now very expensive scrap.
The replacement design is completed, but not fabricated. The solution was to add a rocker in place of the swing finger. The rocker stays in either open or closed position via a high powered 'over center' spring plunger arrangement. Most important, the new claw has a positive mechanical lock. It can grip the adapter simply by moving over the groves in the same manner as the failed claw, but it is not able to release the tool until a mechanical interference is withdrawn. Similar in function to the interlock on the drawbar actuator.
Photos will be in the next post . . if I can figure out how to do the thumbnails.
-
Could not get the thumbnails to work. Anybody have a quick 'thumbnails for dummies' course?
(http://www.thecubestudio.com/BT30SpindleProject/ClawAndArmWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/ClawStartWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/ClawEndWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/ClawEdgeWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/ClawSpringWEB.jpg)
-
Could not get the thumbnails to work. Anybody have a quick 'thumbnails for dummies' course?
Yes, you can upload your pics right from your PC by using the following method:
When replying, click the Additional option drop down arrow (highlighted in the pic below).
To get the maximum amount of photos per post (4), click the "more attachments" option also highlighted below in the pic.
If you have more than 4 photos, you'll have to post twice as the forum only allows 4 photos per post.
Hope that helps- ;)
Dave
-
Thanks Dave.
What your describe is exactly what I did, but when I clicked to post, it just hung. Maybe I did not wait long enough. Is it a long process? I have 6Mb/sec upload speed and I uploaded those and a few more to my site in only a few seconds.
-
Steve, the retention mechansims being used on most VMC's for the carousel look pretty strong to me, and most are simple umbrella changes that aren't going to move as fast as the intermediate swing arm style you're building.
Check out these two videos to get a better idea:
http://www.youtube.com/watch?v=RQ4UT3IwTUg&feature=channel&list=UL
http://www.youtube.com/watch?v=aTwcBMvcqjA&feature=relmfu
And this one for a shopmade intermediate arm:
http://www.youtube.com/watch?v=KEBNB2d-syg
Tool holders are being flung hard.
Cheers,
BW
www.cnccookbook.com
PS Dave, I'll hunt for some similar video. My buddy never lets me video because he doesn't want customers getting irritated that their parts are showing up on the Internet.
-
Possibly the forum itself, as I noticed many "lag" times when I'm here. Heck, even if I'm changing threads to look at something else, it lags for a bit.
Give it another go and I can delete it if it gets messed up for you.
Dave
-
PS Dave, I'll hunt for some similar video. My buddy never lets me video because he doesn't want customers getting irritated that their parts are showing up on the Internet.
Completely understood, Bob!
Dave
-
Steve,
If it were me, I'd design the arm such that the centrifugal force during rotation *could not* allow the tool to come out. The last link Bob provided shows one way of doing this, with the tool being inserted into the arm from the side, rather than the end. A very simple retention device can then be used.
That same video offers an approach I'd never seen before, that could be quite attractive to many hobbyists - rather than having a full toolchanger, just a means of "queuing up" the next tool. The mechanism in the video is very simple and elegant. I may have to build such a device, just for grins.....
Regards,
Ray L.
-
Dave, here is a SCARA robot loading and unloading a lathe:
http://vimeo.com/3355086
Not my friend's setup, but similar. The thing I like is the SCARA's have such simple geometry and construction.
Ray, yup, make the centrifugal force work for you. Use a stronger spring in the extractor too!
I have a big collection of tool changer vids and other details I have been studying. Eventually I will upload to CNCCookbook. Not gonna happen until I finish a few of the giant open 55 gallon drums of worms around the shop!
Specifically, I am just putting finishing touches on installing the mill in a new Flood Enclosure and will shortly start fabricating a new head to mount my Tormach BT30 Cartridge Spindle. Just order a Delta VFD-M and will shortly start experimenting with spindle position control.
Cheers,
BW
www.cnccookbook.com
-
That is a great video, Bob thanks for sharing.
Looked like a wreck was just about to happen at the end there though! lol!
Dave
-
Bob;
Thanks for the links. The retention of the holders in the carousel is not an issue. On mine, the BT tool holders will be completely enclosed in a sealed pseudo-spindle and released via an simplified ball gripper. Credit for the ball gripper idea goes to Mr. Hood. The 'pods' (my fancy name for the carousel mounted holders) each gave a positive mechanical lockout on the gripper that prevents release until a mechanical interference is withdrawn.
As to the speedy littel dual arm gizmo, it *appears* that there is a stud retaining the holders in the arm. It is hard to figure out exactly what the camming is doing even pausing the vid, but there is only a ridge to fit into the holder grove on one side of the arm pocket . . if that made any sense. The stud "I think* must retract to release the holder. So . . if I have that observation correct, then the holders are positively locked into the arm socket by that retractable stud. This is what I was referring to as a 'positive mechanical retention', or 'interlock'. Something that prevents the arm letting go of the holder until some specific action it taken.
Ray;
Centrifugal force is not the problem. The claw already surrounds the holder. The problem is the sudden decel at the end of the arm sweep in the direction where the claw opening leads the motion. The sudden stop generates thru inertia, a force greater than what the claw spring can hold. If I noodled over is long enough I could probably increase the spring pressure, but there is a point where I would need hydraulics to get the darn thing on the holder in the first place, so I abandoned that whole method, even though it is by far the preeminent method used in almost all of the DIY videos that I studied. The industrial stuff is just too friggin' fast to actually divine what's going on.
Critique; the tiny dual sided arm that Bob linked to has one major drawback that you should fix if you do build one; the cam is exposed. I would fit a bellows on that thing or I don't think it will perform very long. In my working life, fault analysis was my specialty. You might be surprised how many machine failures were caused by poor design relative to keeping foreign matter out of the mechanisms.
-
Possibly the forum itself, as I noticed many "lag" times when I'm here. Heck, even if I'm changing threads to look at something else, it lags for a bit.
Give it another go and I can delete it if it gets messed up for you.
Dave
THX, Dave. Will try again on next post.
-
Dave, here is a SCARA robot loading and unloading a lathe:
Not my friend's setup, but similar. The thing I like is the SCARA's have such simple geometry and construction.
Nice, but this is the one I want for changing material in my machines:
http://www.youtube.com/watch?v=lTGOjWv9fcQ&feature=related (http://www.youtube.com/watch?v=lTGOjWv9fcQ&feature=related)
OK, state of technology . . and I found this quite some time ago . . here is the most amazing demonstration of robotic dextrerity I have come across to date:
http://www.youtube.com/watch?v=-KxjVlaLBmk&feature=related (http://www.youtube.com/watch?v=-KxjVlaLBmk&feature=related)
Ray, yup, make the centrifugal force work for you. Use a stronger spring in the extractor too!
Nope, centrifugal force is not the problem . . and stiffer springs cause as many problems as they solve. See my response to Ray on this.
-
Credit for the ball gripper idea goes to Mr. Hood.
No credit to me for the idea, Balding Engineering or whoever made their toolchanger for them is where the credit is due, I only pointed it out ;)
Hood
-
No credit to me for the idea, . . . . I only pointed it out ;)
That is what I credit you for; bringing the idea to my project. If you gave me a Cream Crowdie , I would thank you, even if you did not make it yourself . . . . unless there was already a bite out of it when you gave it to me . . . . in which case . . .
-
Here are the components of the first prototype carousel tool holder pod using the ball gripper arrangement. It works fine, but I have an idea for a simplified variation on the concept that I will try before committing to this one. It is still a ball gripper, but with a snap in / snap out mechanism that has an adjustable retention pressure. The device still has a mechanical safety interlock and therefor cannot release a tool inadvertently. Should do the same job, but be much simpler and does not need an actuator to release the tool. We shall see . . . .
The components are laid out in the photo as an 'exploded view' and they go together in the order shown. The only thing that is not visible is an o-ring groove at the bottom where the adapter is pulled up into the taper. Since there is not a separate hardened sleeve, the gripper is actually a chamber inside the body. Challenging to machine, but I was unable to get a clear photo of it.
Air cylinders, worm gear drives, raw materials, a variety of springs, special fasteners and other assorted components are arriving for the ATC project, but I am getting short on time to work on it before I have to address other priorities. I may get a new claw made, but probably that will be all for a while.
Now to try this thumbnail photo attaching again . . .
-
Do you sleep?
Thats looking great.
I disassembled my ball gripper after about 800 tool changes and there were no signs at all of wear or distortion. Starting to test my spring forks to work out those details. Even if I stall on the tool changer (I won't) changing over to the BT30 and air drawbar was the best mod I've made to this mill since I converted it to Mach. Manual tool changes are so… Civilized :)
Again amazing work your doing.
Derek
-
Steve,
I don't understand the black lever at the bottom of the photo. I'm assuming this is a positive lock, that prevents the ball gripper from coming down and releasing the tool, but it's not clear how it fits. I assume the pivot goes into the slot on the side of the housing, but where does the "tang" on the RH side go?
Regards,
Ray L.
-
Now to try this thumbnail photo attaching again . . .
congrats ;)
Dave
-
Steve,
I don't understand the black lever at the bottom of the photo. I'm assuming this is a positive lock, that prevents the ball gripper from coming down and releasing the tool, but it's not clear how it fits. I assume the pivot goes into the slot on the side of the housing, but where does the "tang" on the RH side go?
Regards,
Ray L.
My bad. I have the lever in the photo upside down . . or backwards, depending on the perspective. It works exactly as you surmised. The 'beak' if you will, pokes thru the hole at the top and interferes with the ball holder descending into the body, this preventing the tool release. The 'tail' reaches down to the bottom of the body and is spring loaded. The spring is not in the photo because those parts are not here yet. The spring goes in the bottom of the pivot slot just below the pivot pin (also not in the photo).
The claw engages the slot on the adapter and then in the last 1/4" or so of travel, it pushed in the tail of the lever, thereby removing the interlock and allowing the air cylinder to release the tool.
You are uber observant, so I will presume your next question will be why is there a 'tooth' on the 'beak'. The 'beak' is retracted, the plunger is then pushed in (down) by a small air cylinder. The plunger not only 'releases' the ball grip, it also physically ejects the adapter about .060". The next action in the sequence if to the arm to descend with the toolholder in-hand, so to speak. At this pint in the sequence there are two options:
1) let the plunger up and lower the arm simultaneously and the plunger will raise all the way to the snap ring stop. It will then need to be depressed again by the air cylinder in order to return the tool
2) lower the arm first and then let the plunger up. When the claw is lowered from the pod, the lever re-opens via its own return spring and the 'beak' pokes back into the chamber. Letting the plunger up at this point results in the 'tooth' catching the top edge of the plunger and holding the gripper in the 'open' position.
In the second configuration, the tool is returned simply by the claw raising the holder into the pod. The claw re-compresses the 'tail' as it raises the last 1/4" or so, thereby releasing the plunger and locking the holder into the pod.
Hopefully that all made sense.
-
Do you sleep?
Thats looking great.
I disassembled my ball gripper after about 800 tool changes and there were no signs at all of wear or distortion. Starting to test my spring forks to work out those details. Even if I stall on the tool changer (I won't) changing over to the BT30 and air drawbar was the best mod I've made to this mill since I converted it to Mach. Manual tool changes are so… Civilized :)
Again amazing work your doing.
Derek
THX! At this point I am not considering the ATC design to be proprietary and there is nothing in it (thus far) that is beyond the capabilities of a competent DIY. I may be inclined to share the drawing of the final design claw and pod. Not a promise at this point, but I don't see a patent in the ATC so far, and I am doing it on my own dime, so no NDA on that part of the project.
The current pod actually works perfectly. My concern with the first prototype was the potential for jamming. The current gripper arrangement does not tend to jam, but it needs to be a little more bulletproof. More robust. Probably the interlock lever's 'beak' should be hardened, which would put it out of DIY. I have already refined that design a bit from what is in the photo and the areas that need to be stronger can be handled with different materials, but I don't want to spend any more time on it for now since I have a new method in mind. Something of a hybrid between the two approaches. See my response to Ray (Kabibble) for a description.
-
Steve,
I don't understand the black lever at the bottom of the photo. I'm assuming this is a positive lock, that prevents the ball gripper from coming down and releasing the tool, but it's not clear how it fits. I assume the pivot goes into the slot on the side of the housing, but where does the "tang" on the RH side go?
Regards,
Ray L.
My bad. I have the lever in the photo upside down . . or backwards, depending on the perspective. It works exactly as you surmised. The 'beak' if you will, pokes thru the hole at the top and interferes with the ball holder descending into the body, this preventing the tool release. The 'tail' reaches down to the bottom of the body and is spring loaded. The spring is not in the photo because those parts are not here yet. The spring goes in the bottom of the pivot slot just below the pivot pin (also not in the photo).
The claw engages the slot on the adapter and then in the last 1/4" or so of travel, it pushed in the tail of the lever, thereby removing the interlock and allowing the air cylinder to release the tool.
You are uber observant, so I will presume your next question will be why is there a 'tooth' on the 'beak'. The 'beak' is retracted, the plunger is then pushed in (down) by a small air cylinder. The plunger not only 'releases' the ball grip, it also physically ejects the adapter about .060". The next action in the sequence if to the arm to descend with the toolholder in-hand, so to speak. At this pint in the sequence there are two options:
1) let the plunger up and lower the arm simultaneously and the plunger will raise all the way to the snap ring stop. It will then need to be depressed again by the air cylinder in order to return the tool
2) lower the arm first and then let the plunger up. When the claw is lowered from the pod, the lever re-opens via its own return spring and the 'beak' pokes back into the chamber. Letting the plunger up at this point results in the 'tooth' catching the top edge of the plunger and holding the gripper in the 'open' position.
In the second configuration, the tool is returned simply by the claw raising the holder into the pod. The claw re-compresses the 'tail' as it raises the last 1/4" or so, thereby releasing the plunger and locking the holder into the pod.
Hopefully that all made sense.
Steve,
Makes perfect sense, and very clever. Thanks!
Regards,
Ray L.
-
Better late than never . . .
-
(;-) Are yous guys building tool changers or PITCHING machines.
(;-) TP
-
(;-) Are yous guys building tool changers or PITCHING machines.
There is definately a security breach around here somewhere . . . .
-
Feedback is that people really like to see the raw materials and not just the finished parts. Also, mock ups seem to be popular. It would be nice if there was more participation in this thread instead of my e-mail, but in any case, here are some mock ups of the mechanisms for the arm and the carousel. I am most likely going to use the same gear box for both, although the carousel will be driven by a smaller BLDC servo and a Coplet Accelnet drive.
To Hood; incidentally, the Copley does have the configurable roll-over feature. I have not had time to check out the rest of the related feature set that you described, but it looks like it may be the hot setup for the carousel. The Mitsu J3 has a special anti vibration capability that should make it a good choice for the uber accell/decell of the arm.
http://www.youtube.com/watch?v=8JkFqzwlyiI
The Mitsu motor can blast over 6K RPM momentarily so with the 50:1 worm gear drive the swing should happen in about 1/4 of a second give or take. Any of the 6 tools should be reachable in less than 1/2 second with the little BLDC servo motor. I *might* be able to eliminated the pneumatic release of the tool from the carousel, so I think I am on target for sub 5 second tool changes and be well back from the maximum speeds.
-
Nice work!.
It is also amazing to me that people will send you an email or PM but they won't post in the forum so everybody can participate.
Mike
-
Look forward to hear what the Copley's can do :)
Hood
-
Nice work!.
It is also amazing to me that people will send you an email or PM but they won't post in the forum so everybody can participate.
Forums are the place where the small can be giants and the weak can be bullies. People have a bad experience with some a-hole who ridicules their ideas or comments and they don't post any more. With 3,400 reads and the same handful of people participating, I have to believe there are some great ideas out there that we will never hear about. It is foolish and short sighted to trade a moment of bravado for losing a resource, but a lot of people do it. Great ideas often come from unlikely places.
If somebody posted a comment or idea in this thread and got slammed for their trouble, I'm afraid I would have to go postal on the slammer.
-
Look forward to hear what the Copley's can do :)
Hood
Turns out the Mitsu has it also. Not a big surprise there. I purchased a second identical gearbox and the Mitsu 400 fits it so perfectly! I would like to get another Mitsu for the carousel, but they are a bit pricey for that application, methinks. Maybe I could justify it by getting a 750 watt Mitsu for my Z axis and recycle that 400 watt guy to carousel duty.
A good friend of mine had an interesting strategy that so far has not worked for me. If he wanted it, he bought it and his universal justification was 'I needed it to make me happy'. Must be nice . . ::)
-
Nice work!.
It is also amazing to me that people will send you an email or PM but they won't post in the forum so everybody can participate.
Forums are the place where the small can be giants and the weak can be bullies. People have a bad experience with some a-hole who ridicules their ideas or comments and they don't post any more. With 3,400 reads and the same handful of people participating, I have to believe there are some great ideas out there that we will never hear about. It is foolish and short sighted to trade a moment of bravado for losing a resource, but a lot of people do it. Great ideas often come from unlikely places.
If somebody posted a comment or idea in this thread and got slammed for their trouble, I'm afraid I would have to go postal on the slammer.
Very well said Simpson. I hope no one was put off too much by previous activity. Your doing some great work and we appreciate you sharing!
-
Are yall talking about me again ? I try to be nice(;-).
(;-) TP
-
Nice work!.
It is also amazing to me that people will send you an email or PM but they won't post in the forum so everybody can participate.
Forums are the place where the small can be giants and the weak can be bullies. People have a bad experience with some a-hole who ridicules their ideas or comments and they don't post any more. With 3,400 reads and the same handful of people participating, I have to believe there are some great ideas out there that we will never hear about. It is foolish and short sighted to trade a moment of bravado for losing a resource, but a lot of people do it. Great ideas often come from unlikely places.
If somebody posted a comment or idea in this thread and got slammed for their trouble, I'm afraid I would have to go postal on the slammer.
Very well said Simpson. I hope no one was put off too much by previous activity. Your doing some great work and we appreciate you sharing!
My problem with inputting any ideas is that I'm just an idiot,... and after watching the creativity shown here (and your other threads), I believe that's why you're not seeing a lot of ideas or comments.
We're all on the sidelines drooling thinking "wow, I hope to be that smart someday" ;D
Dave
-
I'm another one of those on the sidelines, but I'm thinking, he knows what he's doing and appears to of covered all the details/problems I can think of, so not much point contributing!
-
I'm another one of those on the sidelines, but I'm thinking, he knows what he's doing and appears to of covered all the details/problems I can think of, so not much point contributing!
I may not contribute anything worthwhile but I will at least cheer the person on.
When I was actively building my mill and was doing a thread on the 'Zone it was very similar, thousands of views and only a handle ful of comments. It makes you feel lonely sometimes. :'(
Mike
-
Actually I thought of something I always wanted know about these type of interchangeable setups. How does the individual toolholder keep from spinning in the socket when under extreme loads? Is it simply the spring/air pressure from the contacts of the balls?
I have a little toolchanger called the Rapid Changer, and they first had issues with the toolholders spinning so they went with a heavier spring. Then, that didn't "completely" work so the went to toolholders that have radii cut out to mesh with the clamping balls, and while that worked, it was harder to get the toolholder to seat in the toolchanger properly.
Some pics for you.
Dave
-
Hi Dave,
The BT holders have two flats at 180 degrees which engage with keys on the spindle. See photos earlier in this thread. In this one you can see the seats for the keys on the spindle:
http://www.machsupport.com/forum/index.php/topic,22080.msg155852.html#msg155852
And this photos you can see the f;ats on the holder:
http://www.machsupport.com/forum/index.php/topic,22080.msg154764.html#msg154764
Dan
-
I can't believe that Rapid Changer has adequate drawbar force if it needed those cuts to mesh with the gripper balls to keep from spinning. While the BT30 does have drive dogs as mentioned by Dan, there are plenty of taper specifications that don't. Friction in the taper itself should prevent the spinning. Just one more reason they need so much force.
Best,
BW
-
Yes, I have an Emco F1 bench top mill and it uses BT30 holders without the flats and the spring force is sufficient to hold the holders not allowing them to slip.
Dan
-
all the denford atc triacs after 1990 use either bt35 or bt30 toolholders without the drive dogs, so they rely solely on the taper for drive.
-
Interesting, fellas, thanks for chiming in.
So how much force are we talking about?
Thanks,
Dave
-
Depends on the taper, but quite a lot. A number of folks tried with R8's and it wound up taking more than they expected. Life seems to begin at 1000 lbs or more. Best numbers I've seen on it come with the docs for drawbar force gages if you want to Google around. Some are 3000 lbs or more.
Best,
BW
-
Interesting, fellas, thanks for chiming in.
So how much force are we talking about?
Thanks,
Dave
R8 requires upwards of 2500 # drawbar tension for max retention. And, BTW, the set-screw in the slot on the side of an R8 has nothing to do with preventing the tool from slipping, except when the drawbar is being tightened. The friction created in the taper FAR exceeds the feeble ability of that screw to prevent slippage. Spec for 30-taper is about 1300#, 40-taper (IIRC) is about 1800#, 50-taper I'd guess upwards of 2000#. Keys are not necessary for small machines (up to 3-5 HP), above that, the keys seem, if not necessary, at least worthwhile.
Regards,
Ray L.
-
Thanks, fellas, and yes, Ray, most guys I know usually yank that set screw/key out as soon as they get a mill. More of a nuisance if anything.
So Steve, how much force will be used with this unit?
Dave
-
It would be interesting to know the formula used to come up with the draw bar pull force required. It undoubtedly uses surface area/vs tool load. Another good thing to point out here I think is the importance of keeping the spindle taper and tool holders clean and free of surface rust and dings. Nothing will undermine this like an unseated tool spinning inside the spindle.............. that makes for a bad day. >:( I have seen a lot of routers in particular that seem to be abused in this way.
Brett
-
That's a good point, and I like when I see the air blast to clean the tooling during toolchanges, however, sometimes it ends up blowing chips/dust right into the critical surfaces.
Dave
-
Brett I think Most of the values were derived from calculations AND trail and error. I have seen values from 900# up to almost 5000# 30-50 taper.
(;-) TP
-
Drawbar tension has been argued ad-nauseum, so instead, how about some physics and 'why everyone can be correct even with vastly different numbers'. First; there is no linear relationship between drawbar tention and the friction generated. The way tapers work is a very large topic, but in a nutshell, the force on the taper (pull on the drawbar in this case) is generating a press fit. If you want to calculate the torque transfer, you would use similar formulae. The inner part of the interface (the adapter) will be considered incompressible and thrown out for the purpose of this discussion. So the amount of interference fit depends primarily on the material of the outer part of the interface (in this case, the spindle itself) and the amount of drawbar force. That may seem contradictory, but the same drawbar force will create different interference with different materials because the material stretch is vastly differnet from one material to the next.
This is where the 'trial and error' would come into play as was mentioned in the last couple of posts. By way of example; my current mill has an X3 spindle which is just soft mild steel. May have been Toyota frames in a past life, who knows. As a guess, I would (generously) give it 50k - 70k yield. My new Mega-Duty 4th axis spindle and the BT30 spindle are 110k yield. A deeply hardened 4000 series steel spindle could easily hit 190k. It is going to take a lot more tension to spread the taper on that last spindle than on my crappy X3 spindle. It would not surprise me if the drawbar tension for a VMC spindle would literally exceed the yield strength of an X3 spindle and permanently deform the taper. A big hole in the nose and the adapter would be up inside the spindle somewhere . . . .LOL!
To address another question: Surface area of the taper is a consideration, of course, but the gorilla is the taper angle.
So before anyone argues that the drawbar alone is retaining the adapter, consider that if the taper angle is very low (i.e. MT, JT, etc) the leverage is great (lots of stretch for not so much force) and the grip is tremendous for that reason. Note that these tapers do NOT need continuous drawbar tension once the initial force has caused the interference fit. It would be a decidedly bad idea to remove the drawbar from R8 for example after the adapter/collet was seated, but the effect of the pseudo press fit is evident when you have to hammer on the holder (thru the drawbar) to free it from the spindle.
In a later post, I will comment on some other topics and show the completed iteration #2 ATC claw . . working perfectly and will be the final design.
-
Surface FINISH also comes into play. The drawbar force IS important to maintain force on the taper it is NOT like a morse #2 where the taper is so long it has a lot of holding power from the force exerted on the taper by the pin(tool holder). Remove the tension from the BT30 holder and it WILL fall out under cutting loads. Let it get light and it will chatter the toolholder in the spindle under heavy loads
I do not KNOW of a STANDARD value or calculation to estimate the force needed. Each Manf I have dealt with has their OWN NUMBER working range they want you to set the drawbar tension to. They know the basic values needed then add in a fudge factor from in field experinces over the years.
Most are similar for each taper. THICKER spindle shafts may need more, harder spindle shafts may need more.
Call a spindle grinder or repair shop and you will get the genaric values used by most OR they tell you to call the manf for the correct values for the machine.
The standard taper used for all the sizes is not used by accident it it a compromise of torque transfer AND being able to release it from the spindle
Just a thought, (;-) TP
-
Steve,
Is that really correct? I would think tapers are purely a matter of friction. If you know the material and surface finish of the inner and outer tapers, you know (or can measure) the coefficient of static friction (assuming a completely clean, dry fit, as it SHOULD always be). Knowing the drawbar tension and taper angle, you can easily calculate the normal force. Multiply the two, and you know the total frictional force. Integrate that over the entire taper, and you can calculate the total torque that can be generated before the taper slips. I can see how stretch might be factor on low-angle tapers like Jacobs, but have a hard time believing it's a factor on CAT/ISO tapers, especially given the relatively low drawbar tensions, relative to the cross-sectional areas of the toolholder tapers.
Regards,
Ray L.
-
Very good points and a lot to digest.
Brett
-
just an example of the different levels of drawbar tension, an emco f1 (boxford 190) has a hand operated crank to release the toolholder which is relatively easy to use, but a denford triac uses a lot more drawbar tension to hold the same size bt30 toolholder, operated from an air cylinder.
The reason is the spindle power, the triac is about 2-3 times more powerfull than the emco f1.
-
If your driving it with lugs then isn't it just a calculation of how much downward pull a tool will exert on the Belleviles.
Derek
-
something i've just remembered, many years ago i was operating a 40 int mill and i checked to see if the lugs were actually doing anything. After the job the gap between the drive lugs and the toolholder was still there, so this job didn't cause the toolholder to use the drive lugs.
-
Surface FINISH also comes into play. The drawbar force IS important to maintain force on the taper it is NOT like a morse #2 where the taper is so long it has a lot of holding power from the force exerted on the taper by the pin(tool holder). Remove the tension from the BT30 holder and it WILL fall out under cutting loads. Let it get light and it will chatter the toolholder in the spindle under heavy loads
I don't recall saying that the drawbar was unimportant. Certainly I did not suggest running a BT30 without the drawbar.
I do not KNOW of a STANDARD value or calculation to estimate the force needed. Each Manf I have dealt with has their OWN NUMBER working range they want you to set the drawbar tension to. They know the basic values needed then add in a fudge factor from in field experinces over the years.
See Ray's post
Most are similar for each taper. THICKER spindle shafts may need more, harder spindle shafts may need more.
Agreed. Now, why do you suppose that is?
The standard taper used for all the sizes is not used by accident it it a compromise of torque transfer AND being able to release it from the spindle
Agreed. Remember I started out by sating that tapers is a large subject. The same forces are at work in all tapers. An MT is no different that a CAT40 as far as the physics are concerned. They behave very differently as you correctly point out, but that is because of the shape. For example a 1 degree taper and an 89 degree taper are going to have very different properties and every angle in between will follow suit, but still follow the same rules.
You have long pointed out that there is a range of tensions quoted and that each manuf has their own numbers and there is 'trial and error' and in this post you mention 'fudge factor'. All true. What I offered is an explanation of some of that factors that make it true.
-
Steve,
Is that really correct? I would think tapers are purely a matter of friction. If you know the material and surface finish of the inner and outer tapers, you know (or can measure) the coefficient of static friction (assuming a completely clean, dry fit, as it SHOULD always be). Knowing the drawbar tension and taper angle, you can easily calculate the normal force. Multiply the two, and you know the total frictional force. Integrate that over the entire taper, and you can calculate the total torque that can be generated before the taper slips. I can see how stretch might be factor on low-angle tapers like Jacobs, but have a hard time believing it's a factor on CAT/ISO tapers, especially given the relatively low drawbar tensions, relative to the cross-sectional areas of the toolholder tapers.
Yes, I mentioned that the formulae are similar, but there is no threshold that I am aware of where the rules change after a certain angle. As you pointed out, the force pressing the surfaces together is what creates the friction. That force can come from many sources. In the taper, the force is not applied perpendicular to the surface but rather, as you correctly point out, thru an angle. If you consider each surface as non moveable, then the calculation is not complicated. But all materials are elastic. It is only a question of degree.
Bolts are a lot easier for people to comprehend. While the calculations are again straightforward, they are not directly concerned with the amount the bolt stretches, but rather the tension generated. The stretch is definitely there in any case. Take the practical example of putting 100ft lbs of torque on an ungraded 'hardware store' 3/8" bolt. It is unlikely the bolt will ever reach the 100lb spec. It would just stretch to failure. A grade 8 bolt would generate an extreme amount of tension. The difference is in the material. If manuf #1 us using bolts from Home Depot, he may well have a different spec than the manuf who use using graded fasteners.
My intention was to point out some of the non-so-obvious, but still very much present factors that might provide some insight into why one manuf would have different specs than another manuf for the same CAT40 taper and why both might be correct in their application. I think it would be safe to say that my X3 spindle would never survive a 4k or 5k pull on a taper. Considering the material that spindle is made from, we would call that a swaging operation . . . LOL!
I think we all do the same thing, Ray. 1) We run the numbers and then 2) build a prototype and then 3) try the prototype and see if the calculator lied or not. Some folks skip step 1.
-
Actually I thought of something I always wanted know about these type of interchangeable setups. How does the individual toolholder keep from spinning in the socket when under extreme loads? Is it simply the spring/air pressure from the contacts of the balls?
Transferring torque thru balls is not unheard of. Motorcycle transmissions use this method. In your toolholders it seems like something of a 'band-aid' for a stubborn problem, but without seeing the internals on the parts, it would just be speculation.
-
Thanks, fellas, and yes, Ray, most guys I know usually yank that set screw/key out as soon as they get a mill. More of a nuisance if anything.
I can vary the force on the spring pack by turning the threaded cap on the stack, so anywhere from zero to about 2,000lbs with the current springs. A different spring is available in the same size (ID/OD) the would take it into the 3k neighborhood, but I don't anticipate needing that.
A little over .200" is needed for release, to that limits the amount the stack can be compressed and still release the tools. I would need to look that up again before I quoted, but as it stands now I have it set to 1,000lbs and release is approx 1,400 lbs. All numbers are calculated as I have no way of directly measuring these forces.
-
That's a good point, and I like when I see the air blast to clean the tooling during toolchanges, however, sometimes it ends up blowing chips/dust right into the critical surfaces.
It has been said that you can easily train a cat to do whatever the cat wants to do. This applies to swarf as well, I would imagine. Having reviewed a lot of different ATC designs, I elected to set a goal of completely encasing the adapters in 'pods' to keep the tapers clean. I even have a soft o-ring at the bottom of the pod . . . OK, maybe a bit of overkill there. :-\
-
Yes, I have an Emco F1 bench top mill and it uses BT30 holders without the flats and the spring force is sufficient to hold the holders not allowing them to slip.
It would be helpful to know the application. There is a lot of distance between driving 1/4 roughers thru aluminum and pushing a big face mill across an interrupted cut on stainless.
-
all the denford atc triacs after 1990 use either bt35 or bt30 toolholders without the drive dogs, so they rely solely on the taper for drive.
That could be because someone determined that the dogs were not needed (say if the mills are all low HP spindles), . . . . -or- . . . it could be that they were unable or unwilling to provide the required indexing for the spindles in order to use the drive dogs.
-
something i've just remembered, many years ago i was operating a 40 int mill and i checked to see if the lugs were actually doing anything. After the job the gap between the drive lugs and the toolholder was still there, so this job didn't cause the toolholder to use the drive lugs.
Let me first say that I am merely speculating when I comment that it occurs to me that in a perfect world the friction would always be sufficient for the job. There are those extreme cases where perhaps a tool spins or even dislodges from the spindle, and then the momentary slip where the friction was overcome momentarily.
I suspect that the latter is far more common and would include things like tool crashes. The damage to the spindle taper could be significant if the holder were allowed to spin, so perhaps the dogs have always been intended as 'arrestors' rather than 'drivers'.
-
The dogs may be overkill for this but they are used as an orientation device when you need to orient the tooling such as when back boring.
-
Yes, I have an Emco F1 bench top mill and it uses BT30 holders without the flats and the spring force is sufficient to hold the holders not allowing them to slip.
It would be helpful to know the application. There is a lot of distance between driving 1/4 roughers thru aluminum and pushing a big face mill across an interrupted cut on stainless.
8mm roughing mill full diameter 4mm deep in aluminium. I am sure I could do twice that depth and the tool wouldn't slip. Don't use face mills on that small machine. Do use slitting saws and had a few times it jamming, but again it's not really a reference as it was stalling the undersized spindle motor (around 5Nm at the spindle).
Dan
-
Interesting, fellas, thanks for chiming in.
So how much force are we talking about?
Thanks,
Dave
R8 requires upwards of 2500 # drawbar tension for max retention. And, BTW, the set-screw in the slot on the side of an R8 has nothing to do with preventing the tool from slipping, except when the drawbar is being tightened. The friction created in the taper FAR exceeds the feeble ability of that screw to prevent slippage.
I will deny I ever said this, but many, many moons ago, we used the pin and a loose adapter as a red neck floating tap holder. For emergency use only, of course . . . . :-[
Spec for 30-taper is about 1300#, 40-taper (IIRC) is about 1800#, 50-taper I'd guess upwards of 2000#. Keys are not necessary for small machines (up to 3-5 HP), above that, the keys seem, if not necessary, at least worthwhile.
It seems logical to me that with only a couple of steps in adapter size and a wide range of spindle HP, that both the drive dogs and the drawbar tension would be specified for the machine and not as a set number per taper.
Does anyone know if there are published HP ranges for each taper?
-
Yes, I have an Emco F1 bench top mill and it uses BT30 holders without the flats and the spring force is sufficient to hold the holders not allowing them to slip.
It would be helpful to know the application. There is a lot of distance between driving 1/4 roughers thru aluminum and pushing a big face mill across an interrupted cut on stainless.
8mm roughing mill full diameter 4mm deep in aluminium. I am sure I could do twice that depth and the tool wouldn't slip.
CNCcookbook has a speed/feed calcuator that includes (IIRC) the HP required and the MRR as part of the data for each calculation. This would be good info to have available for machines that are short on power.
Currently, I have 7HP peak and I have not slipped an R8 . . that I know of . . . , but then I am changing tools by hand and cranking on the drawbar nut with a wrench. I would guess that I am easily putting over 2,000 lbs tension on the drawbar. The spindle is small and skinny and not made from very good material so it gets a killer monkey grip on some of the tooling and needs serious persuasion sometimes to release the tools.
-
Let me first say that I am merely speculating when I comment that it occurs to me that in a perfect world the friction would always be sufficient for the job. There are those extreme cases where perhaps a tool spins or even dislodges from the spindle, and then the momentary slip where the friction was overcome momentarily.
I suspect that the latter is far more common and would include things like tool crashes. The damage to the spindle taper could be significant if the holder were allowed to spin, so perhaps the dogs have always been intended as 'arrestors' rather than 'drivers'
The dogs may be overkill for this but they are used as an orientation device when you need to orient the tooling such as when back boring.
I actually think them being arrestors (cheap insurance) were more an added benefit rather than a feature by design. Most I have seen use tiny bolts to attach them to the face of the spindle. All though they are in a shallow groove an empty tool holder allowed to spin at spindle start up could easily shear them off. Totallyrc hit the nail on the head as to why most I have seen use keys. I hate to mention these last 2 because I really don't like either one but......... I have seen them use keys for a fine orientation adjustment, an alignment tool. These were on a belt driven spindle and I'm just guessing this was their motovation. The outside edges were ground on an angle. This would only allow them a tiny bit of misalignment but between the tool holder rack and some other low end features it needed it. They did however run for many years like this with good results............. most of the time. ;D I think my least favorite use of a key on a spindle was to actuate a release on the tool change arm.
Brett
-
If you were looking for grip/slip value per spindle type you would use the available TORQUE value not HP.
Just a thought, (;-) TP
-
Bret that key with tiny bolts will take a LOT more load than you think the key would be in shear mode in the slot.
Cat holders use offset keys to locate the tool holder to the spindle in ONLY one spot. I have seen a few LOAD the tools backwards with not so good results when it came time to toolchange. Some machines would actually LOAD the tool holder jammed offset in the spindle. You can guess the results.
(;-) TP
-
For those that have the R8 and a ton of tooling you may want to consider the MACH1 tool system. It is as close to an auto system as you will find for the knee mills. AND note it does NOT use beveled washers (;-) IT also uses your standard R8 tooling with a simple mod.
Probably could be used with CNC as an auto changer IF your spindle axis drive was strong enough.
(;-) TP
-
I'd like to offer one other variable that may play into this equation, and that is the thermal effect.
Loading a chilly holder into a toasty spindle.
The expansion / shrinkage of the mating materials immediately following the tensioning of the draw bar would vary the final values ... would it not ?
I've certainly seen this effect with the #5 Morse in a radial drill. Granted, the shank surface finish wasn't pristine but required substantially more effort to remove.
Discard if irrelevant,
Russ
-
They call that shrinkfit. I have seen one spindle that came close to being scrapped because the OP could NOT get the tool holder back out. I had to wrap a heating strip around the end of the spindle/toolholder let it get hot THEN pack the tool holder with dry ice to shrink it and it THEN popped out of the spindle with a bit of persuasion. Lucky it was an NMTB40 spindle with a through hole draw bar.
(;-) TP
-
If you were looking for grip/slip value per spindle type you would use the available TORQUE value not HP.
Torque and HP are essentially the same thing. HP is just torque over time, and the relationship is linear.
-
I'd like to offer one other variable that may play into this equation, and that is the thermal effect.
Loading a chilly holder into a toasty spindle.
The expansion / shrinkage of the mating materials immediately following the tensioning of the draw bar would vary the final values ... would it not ?
I've certainly seen this effect with the #5 Morse in a radial drill. Granted, the shank surface finish wasn't pristine but required substantially more effort to remove.
Absolutely. The adapter will expand and the spindle contract as the temp equalizes, creating (or in this case increasing) the interference fit.
In fact, this is the most common method for assembling and disassembling press fit parts.
-
Torque and HP are essentially the same thing. HP is just torque over time, and the relationship is linear.
Not exactly. Power is torque times rotational speed. And saying that the relationship is linear is a bit confusing as the speed (RPM) is not a constant itself. Take a stepper motor for instance - peak torque is at zero RPM, where the power is zero, while peak power is at some mid point RPM where torque is substantially less than its peak value.
Hence, the way Terry put it, is the correct way. You are essentially transferring torque. Power has absolutely no meaning here. I could have a 6kW high speed spindle putting out the rated power at 30,000RPM, which means only 2Nm torque, whilst a tiny NEMA 34, 25 watt stepper would have twice that torque at zero speed. Power is the amount of work you can do over time, so obviously the 6kW spindle would do more work in a given time at its lower torque than the high torque stepper.
Dan
-
YEP Dan has it right. Torque is the measure of FORCE. HP is a measure of work over time. HP is a calculation and torque is a true measurment.
Another part of the spindle equation is RPM and balance. The inertial value of the imbalance squares as the rpm doubles making balance very important for high speed spindles. It can add to the drawbar pressure equation.
(;-) TP
-
Torque and HP are essentially the same thing. HP is just torque over time, and the relationship is linear.
Not exactly. Power is torque times rotational speed.
That is correct, Dan. However, consider what rotational speed is. It is revolutions per MINUTE. It is time in the HP equation.
And saying that the relationship is linear is a bit confusing as the speed (RPM) is not a constant itself.
Sorry to be confusing. I make the mistake of assuming everyone is on the same page with the terminology. A 'linear' relationship means that the calculation is the same at every point in the curve. It does not mean that the power output of any specific device is linear. For example, the formula for HP is the same if the stepper is running at 2PRM, 20PRM or 200 RPM.
Put another way, the 'torque curve' may be dropping like a rock with the stepper motor, but the relationship between HP, torque and speed is the same no matter where you measure or calculate it.
-
A 'linear' relationship means that the calculation is the same at every point in the curve.
Actually.... conventionally, a "linear relationship" means a graph of the relationship would yield a straight line. For example, Y = 3 * X is a linear relationship, while Y = 1 / X is not. Both *are* "linear equations", since neither contains any exponentials, logs, trig functions, or or non-linear terms.
Regards,
Ray L.
-
YEP Dan has it right. Torque is the measure of FORCE. HP is a measure of work over time. HP is a calculation and torque is a true measurment.
This is semantics. You are just saying the same thing in a different way. I don't know how you define a 'true' measurment. If you put a motor on a dyno, you are measuring toque and you are also measuring speed (time in the equation).
However I will add that torque is also a static force. I doubt you have a way of directly measuring torque in a practical way with a machine spindle, However, you CAN use the motors HP power curve to calculate the torque at any point. Since torque is what generates the stress on the mechanism (which I think was your point), then you can assess the effect, but you would get the torque out of the HP calculation . . which is linear.
If the manuf provides HP curve and also the torque curve, then there is no need to know the formula and the relationship is irrelevant, but inquiring minds may still want to know how it all fits together.
Another part of the spindle equation is RPM and balance. The inertial value of the imbalance squares as the rpm doubles making balance very important for high speed spindles. It can add to the drawbar pressure equation.
Agreed. Certainly the lateral forces rise exponentially (as opposed to linear) as the speed rises. Recall that I found two separate drawbar specs for BT30 differentiated by RPM. Incidentally I would point out that the drawbar tension is force and not pressure, but that would be nit picking, don't you think? ;)
-
A 'linear' relationship means that the calculation is the same at every point in the curve.
Actually.... conventionally, a "linear relationship" means a graph of the relationship would yield a straight line. For example, Y = 3 * X is a linear relationship, while Y = 1 / X is not. Both *are* "linear equations", since neither contains any exponentials, logs, trig functions, or or non-linear terms.
If Y is always 1/X and there are no other terms that change this relationship, the the relationship is linear . . . not meaning a straight line, but rather unchanging . . . like a staight line . . . only different. ??? On topic, the relationship between torque, time and horepower does not change with torque, time or horsepower, so what shall we call that relationship?
Easy to see there is bordome brewing. I will post some photos of the finished Rev1 claw. Then the red pencils can come out and we will resume the fun.
-
Next you can add in the lateral cutting forces involved with the machine. (;-) Light cutting or heavy hogging . 6 flute cutting or 1 flute cutting
You asked for participation(;-) Be Careful what you ask for. LOL
(;-) TP
-
You asked for participation(;-) Be Careful what you ask for. LOL
Your participation is; Usually interesting. Occasionally entertaining. Always welcome and appreciated.
-
Time for new stuff. Two topics the carousel pod and the claw. NOTE the photo attaching did not work again, so I am using the previous method
First photo is a test rig for the prototype # 2 gripper type Carousel pod. It works perfectly, holds the adapter tight against the o-ring seal and is just cool looking. So after much contemplation, Ive decided against doing the hybrid prototype and will stick with the gripper pod. It is a little bit more challenging to make and more $$, but not when compared against doing an entire third prototype. I wanted to avoid torquing the carousel with the release cylinder above the pods, but the force needed to yank the adapter out of the hybrid pod might well be more then the cylinder exerts, so that became something of a moot point.
(http://www.thecubestudio.com/BT30SpindleProject/PodTestRigWEB.jpg)
Enough about the pod. On to the CLAW . . . dun duuuun <- scary sound effect ::)
OK, the claw has a rocker in place of the finger on the previous . . hand . . (it was not strong enough to enjoy true claw status). The rocker snaps smartly to its open and closed positions by way of an over-center spring plunger arrangement. The force is quite high and it requires a very enthusiastic deceleration to dislodge the adapter.
Which brings us to the safety interlock. When the claw rocker is in the closed position, the spring loaded interlock plunger moves between the rocker and the claw body, interfering with the claw rotation and preventing the release of the adapter. The plunger is retracted by the solenoid which allows the rocker to . . rock . . and release the tool. The rocker remains in the open position under spring pressue until it is again slapped around an adapter. I have messed with it a bit and it seems to be working as intended so this will be the final design for the Claw. The pod will get some minor improvements to make the interlock a little bit more bulletproof and that wraps up the tool transfer mechanisms for the ATC.
Last order of business will be the actuators for the arm and the carousel, but that will have to wait as I have higher priorities ahead of this.
I hope everyone has enjoyed the thread so far. My objective here was not to bring cool finished parts for show and tell, but rather to show the development process that results in the cool parts . . including the fails which are part of the normal flow. You hope the design will pass all testing and be a done deal (like the first spindle gripper), or perhaps have some typical adjustments and minor changes and still move into the light. Really you don't like to abandon finished prototypes altogether, but in engineering we have a saying; 'you can't polish a turd'. Mo better go back to the drawing board than to break out the brasso and have at it. The first claw was a clear fail . . which is bad. The first pod, while functional, was supplanted by a superior idea . . which is not so bad.
(http://www.thecubestudio.com/BT30SpindleProject/ClawRev2topWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/ClawRev2OpenWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/ClawRev2Top01WEB.jpg)
-
Are you really Tony Stark?
Beautiful work as usual. Question on the finish of the gripper. Looks anodized. Did you glass bead it first? Soda blast? When I anodize my parts I get a good finish but the finish on the gripper is really nice.
Thanks
Derek
-
Are you really Tony Stark?
Yes
.
-
Are you really Tony Stark?
Beautiful work as usual. Question on the finish of the gripper. Looks anodized. Did you glass bead it first? Soda blast? When I anodize my parts I get a good finish but the finish on the gripper is really nice.
Thanks
Derek
;D
-
WTF is he anyway?
Hood
-
WTF is he anyway?
Hood
Not what, WHO. I had to google him. It is the Iron Man character. I was looking for the connection when my wife said that a claw was a prominent prop in the movie. I don't remember that, but then I don't remember what I had of breakfast . . . or if I even had breakfast. Know what I mean Fred?
-
I was inferring the skill level not a specific prop:) You guys need to get out and see a movie now and again!
Derek
-
With 2 young grandsons next door I have watched IRON MAN about 6,348,976 times SO FAR. They seemed to be fascinated that someone could build such a thing,. I do hope the curiosity continues .
And yes they have ask me to build one and NO I have not (;-)
YET, (;-) TP
-
HIYA Steve have you seen this model tool changer ?
http://www.youtube.com/watch?v=8opxZLSAvN8
(;-) TP
-
I was inferring the skill level not a specific prop:) You guys need to get out and see a movie now and again!
OK, I suppose you have suffered long enough. The aluminum . . . . CLAW <dun duuuuhn!> has been sand blasted with fine sand at about 80lbs. Being puuuurdy is a side effect. The porpoise is to deburr the complex shape without making a career out of it. Damn thing had 60 ways go get me.
Having said that, since it was puuuurdy, may as well keep it that way. The satin clear coat is from Eastwood and it is called Satin Diamond Clear For Bare Metal . . . or something like that. Sticks to bare aluminum . . tough . . handles heat very well . . . amazing stuff.
-
The satin clear coat is from Eastwood and it is called Satin Diamond Clear For Bare Metal . . . or something like that. Sticks to bare aluminum . . tough . . handles heat very well . . . amazing stuff.
Even polished aluminum?
Nice work, Steve-
Dave
-
You guys need to get out and see a movie now and again!
Derek
What is derek on about?
Quote from: Hood on October 09, 2012, 07:54:17 AM
WTF is he anyway?
Hood
Not what, WHO. I had to google him. It is the Iron Man character. I was looking for the connection when my wife said that a claw was a prominent prop in the movie. I don't remember that, but then I don't remember what I had of breakfast . . . or if I even had breakfast. Know what I mean Fred?
What is breakfast and what did Fred do this time? I'm just utterly befuddled about all this. Only thing I'm sure of is those are some fine looking pieces Clint Eastwood made for you Steve! ;D
-
The satin clear coat is from Eastwood and it is called Satin Diamond Clear For Bare Metal . . . or something like that. Sticks to bare aluminum . . tough . . handles heat very well . . . amazing stuff.
Even polished aluminum?
Yes, but polish has wax or silicon in it, so you need to make sure the surface is super clean. Ditzler 'wax and grease remover' is my favorite, but Dupont 'PrepSol' is also good. You need to follow the directions to a 'T' with these products or you will just be moving the contamination around and not removing it.
Any paint that sticks to aluminum or is described as 'self-etching' (aluminum primers) have acid in them. This is respirator ONLY stuff. Don't even think about breathing even a small amount of it. If you can smell it, you are breathing it.
Nice work, Steve-
Dave
THX.
-
What is breakfast and what did Fred do this time? I'm just utterly befuddled about all this. Only thing I'm sure of is those are some fine looking pieces Clint Eastwood made for you Steve! ;D
Well, you have to ask yourself this question; 'Do I feel lucky' . . . . well do ya . . . FRED?
-
HIYA Steve have you seen this model tool changer ?
Two things are really impressive. 1) the . . . CLAW <dun duuuhn!> actually rotates 180 degrees during the arm sweep. and 2) the invisible tooling.
-
I almost bought one of these changers detailed in TP's post, it was on a Matchmaker that my friend has (or has just sold to mc) which was in reality a rebadged Shizuoka. It works well but the problem is there was no way for me to use it as I required long pullstuds for my mill and the changer required you just used a normal drawbar. If the pullstids on my mill had been a bit shorter I probably could have got away with it but the rise and fall wasnt enough with the long studs.
Heres a pic of it. http://www.machsupport.com/forum/index.php/topic,10331.msg74181.html#msg74181
Hood
-
On this side of the pond it was known as the one armed bandit.
(;-) TP
-
Here is another interesting design.
http://www.edingcnc.com/index.php?pagina=detail_galerij&galerijid=46
(;-) TP
-
Yes a lot of them seemed to have the Bandits, not sure what was originally on the Matchmaker but it had a Heidenhain (think 155) on it when Mark bought it. That is actually what he bought it for as he has quite a few machines with that controller and he wanted spares. I was thinking of buying the Matchmaker from him as it is a solid machine, every bit as stout if not better than the Beaver I have but I really wanted a bedmill so decided against it.
Hood
-
I found a pdf copy of the installation/user manual for the Dana/Summit tool changer on the Matchmaker, and all the wiring/programming examples are for a Bandit control. Plus it's all designed to pretty much bolt straight onto Shizuoka mills, right down to the LH spindle thread.
-
Guys,
Quick video of my toolchanger. This is being manually sequenced, so it's kinda slow. I used my pendant to trigger the PDB, had a manually operated pneumatic valve to operate the "lift", and a microswitch to turn on the Geneva motor.
http://www.youtube.com/watch?v=QvSzfLZGH9Q
Regards,
Ray L.
-
Ray, Very nice mechanism. Smooth and to the point , Great Job.
(;-) TP
-
Ray, you mentioned having a pin and rod arrangement for align the tools and the spindle, IIRC. Is that on the to-do list or am I just not seeing it?
Nice job overall. Where did you get that swing arm actuator . . . ;)
-
Ray, you mentioned having a pin and rod arrangement for align the tools and the spindle, IIRC. Is that on the to-do list or am I just not seeing it?
Nice job overall. Where did you get that swing arm actuator . . . ;)
Steve,
The alignment pin is half there - the pin is on the other side of the head. The socket on the arm is not yet there - I just got the bushing today, so it's about an hours work to make the holder and mount it. That will accomodate any minor "sag" that might occur on the arm and it's mount.
You weren't supposed to notice that actuator.... :-)
Regards,
Ray L.
-
Ray,
You better get a move on, buddy. You have been at this ATC project waaay longer than me so it will be :-[ if I beat you to the punch.
I wrapped up some incidental loose end stuff; intelock pin and solenoid for the drawbar actuator, beefed up the interlock lever on the pod, etc, got the shafts made for the swing arm and the carousel, and ordered the steel to make the mounting brackets for everything . . . . . before putting the whole shebang on a shelf.
You're getting a 60 day window now to keep that tortoise moving while the hare is off on a rabbit trail doing other stuff. After that, I figure about a week and I'll be changing tools.
Contest rules: everything does not have to be powder coated or otherwise 'spiffy', but the automatic tool change has to be . . well . . automatic, and under its own power. No manual operation and no . . uh . . temporary actuators . . .
However, other . . 'favored contestants' will be allowed to demonstrate the completed ATC on a totally non functional mock up of the milling machine head . . . because there is no actual mill yet . . and because he is making the rules. ;D
Contest winner is responsible for any and all federal, state and local taxes on prizes, which may include a case of the contestant's favorite beer . . and bragging rights.
Tick, tick, tick . . . .
-
Here is another interesting design.
http://www.edingcnc.com/index.php?pagina=detail_galerij&galerijid=46
THX! Every example has some interesting feature.
Good: moves with the head (apparently), tapers of stored tools are completely covered and protected, the arm extending out and retracting is just totally 8)
Not-So-Good: Slow, only a few tools, tools hang over the table.
-
Ray,
You better get a move on, buddy. You have been at this ATC project waaay longer than me so it will be :-[ if I beat you to the punch.
I wrapped up some incidental loose end stuff; intelock pin and solenoid for the drawbar actuator, beefed up the interlock lever on the pod, etc, got the shafts made for the swing arm and the carousel, and ordered the steel to make the mounting brackets for everything . . . . . before putting the whole shebang on a shelf.
You're getting a 60 day window now to keep that tortoise moving while the hare is off on a rabbit trail doing other stuff. After that, I figure about a week and I'll be changing tools.
Contest rules: everything does not have to be powder coated or otherwise 'spiffy', but the automatic tool change has to be . . well . . automatic, and under its own power. No manual operation and no . . uh . . temporary actuators . . .
However, other . . 'favored contestants' will be allowed to demonstrate the completed ATC on a totally non functional mock up of the milling machine head . . . because there is no actual mill yet . . and because he is making the rules. ;D
Contest winner is responsible for any and all federal, state and local taxes on prizes, which may include a case of the contestant's favorite beer . . and bragging rights.
Tick, tick, tick . . . .
Steve,
You've had the considerable advantage of actually being able to work on yours.... Except for about 30 minutes to install the flow control valves and make that video, I've been too busy with "real" work to touch mine in weeks. Hopefully later this week I'll be abve to get back on it. I think I can finish it off in about a week, once I have the time.
Regards,
Ray L.
-
Like the old saying goes, I could finish that up in a couple of hours IF I had all week to work on it.
(;-) TP
-
You've had the considerable advantage of actually being able to work on yours.... Except for about 30 minutes to install the flow control valves and make that video, I've been too busy with "real" work to touch mine in weeks. Hopefully later this week I'll be abve to get back on it. I think I can finish it off in about a week, once I have the time.
Most of us are in that boat, I suspect. I did schedule time for the project, so I could work on it *almost* full time for a few weeks, but I did not get it done completely due to pesky interruptions like food and sleep. Now I will be off line for two months with the next scheduled project, so that levels the playing field and makes for a fair contest, methinks.
As for the prize . . I like Cave Creek Chili Beer . . . . Just sayin' :)
-
Like the old saying goes, I could finish that up in a couple of hours IF I had all week to work on it.
And from the same book:
Torque spec: Tighten till bolt strips, then back of 1/2 turn.
Installation instructions: Cut to fit, paint to match.
Fabrication: Cut once, then measure twice.
Troubleshooting: Machine making loud noise; turn off the machine -or- turn up the radio.
Maintenance: just another word for Duct Tape.
-
Steve,
You forgot one: Measure with a micrometer, mark with chalk, cut with an axe.
Regards,
Ray L.
-
I didnt realise the workshop practice manual I had written was available over your side of the pond :)
Hood
-
Most of us are in that boat, I suspect. I did schedule time for the project, so I could work on it *almost* full time for a few weeks, but I did not get it done completely due to pesky interruptions like food and sleep. Now I will be off line for two months with the next scheduled project, so that levels the playing field and makes for a fair contest, methinks.
As for the prize . . I like Cave Creek Chili Beer . . . . Just sayin' :)
Well, you could be in some trouble then. I expect to be back on mine in the next day or two. The only thing holding me back will be I'm missing some pneumatic doo-dads, which are due to show up any day now....
Regards,
Ray L.
-
I didnt realise the workshop practice manual I had written was available over your side of the pond :)
Hood
Oh yeah, we got it and you have been learning me some stuff. ;D
Steve,
You forgot one: Measure with a micrometer, mark with chalk, cut with an axe.
Regards,
Ray L.
Can I use this Ray? ::)
-
Got the pivot arm actuator done. It's a little "sticky" at the quill end. It's difficult to make this work with a simple mechanism as the arm needs to rotate 140 degrees, and actually goes slightly over-center at the quill end. It seems to work OK, but I will smooth it out, to get rid of the sudden acceleration when it leaves the quill. Worst case, I can add a small "kicker" cylinder to get it started. Other than that, it seems to work nicely.
http://www.youtube.com/watch?v=whtDRNmo8Sk&feature=youtu.be
Now if only the rest of the pneumatic parts would show up, I could start working on the software side....
Regards,
Ray L.
-
Ray it is all in the geometry of the cylinder and arm. I also work in Gate controllers/operators. I have seen installers get the geometry wrong and ripp out gearboxes on the operator.
Look at the pull angle you have when you go to retract the arm, the cylinder is almost directly inline to the arm AND the arm is almost directly inline with the pivot point of the arm. IT has very little leverage left to pull back the arm . IF you move your cylinder end mount out a small amount you will have greater leverage on the arm to get it started and it will retract smoother.
Just a thought, (;-)TP
-
Terry,
Believe me, I am well aware of the geometry. The arm needs to rotate 140 degrees, which is a LOT. Basically, both ends of the travel end up being near a critical point, where it is at the very limit of usable range. If I move the cylinder further out, then the ATC can no longer retract far enough to clear the table. Right now, it's going right up to the edge at the quill end of the travel, but it DOES work reliably. It takes a fair amount of tension in the cylinder to pull it off the quill, hence the "jump", but once it's moved the first 3/4" or so, it's fine. One simple fix is to just add a very small (<1/2") air cylinder that is fed by the same air line that feeds the big cylinder to retract the ATC. This small cylinder will pop out, and "push" the arm away from the head to get it started, which will completely resolve this small problem.
I could, of course, also go with a more complex linkage, but I really don't want to go there unless I have no choice. This simple linkage is close enough, I have no doubt I can make it work perfectly with just a few minor tweaks.
Regards,
Ray L.
-
I found the alignment pin was dragging on the receiver (the receiver is a crude prototype, and needs to be re-done). I removed the alignment pin, and added a small, spring-loaded pin on the arm, to help push it away from the head. It's now working nice and smooth, with very positive location at both ends of the travel.
Well.... Back to trying to "will" the pneumatic parts to show up, so I can finish this bad boy.... I cannot wait to see it working on its own.
Regards,
Ray L.
-
I got my pneumatic parts today, so tomorrow I'll be working on software to sequence the ATC. I've the air cylinders plumbed and wired, and under computer control. I just need to wire up the carousel motor, and start writing code! No "sensors" yet, but I can easily make it operate without them for now.
Regards,
Ray L.
-
Here it is doing a toolchange under computer control for the first time. It's running *really* slow, because I have long delays between the steps, so I can catch it if something goes wrong. It's about 20 seconds in the video, and I should easily be able to get that under 10 seconds when it's all done.
P1060662 - YouTube (http://youtu.be/xPSgeAMuFC4)
Regards,
Ray L.
-
Getting close to miller time(;-)
(;-) TP
-
Ray did you give any thoughts to doing a MOTOR drive to swing the arm instead of air. I worry about the constant BANGING at the end of stroke with the air.
Just curious, (;-) TP
-
Another option might be air over oil to more precisely control the cylinder....just a thought.
Russ
-
A motor drive with a link could give you a slow start /stop zone. It would start slow then speed up through the swing to MID swing then slow down as it came to a stop. The link creates this feature with the motor running at constant speed NO motor speed controls needed.
(;-) TP
-
Ray did you give any thoughts to doing a MOTOR drive to swing the arm instead of air. I worry about the constant BANGING at the end of stroke with the air.
Just curious, (;-) TP
Terry,
Yes, but it creates other problems. It's really not as rough as it looks in the video - I had not done much in the way of "tuning" before I took that. Now that I've had some time to tweak it, it's a lot smoother. I will also be adding dampers at the ends of travel, to make the stops more gentle.
I did run the thing for some time yesterday, just swinging out, doing a random toolchange, then swinging back. I did dozens of them with no "Oopsies". Once I get all the sensing switches in place, it should work really nicely.
Regards,
Ray L.
-
Let us know WHEN you get it slinging chips AND changing tools realtime. We can take up a collection and send the case of beer(;-)
(;-) TP
-
Just spent almost two hours running constant, random toolchanges without a single problem! Now that the mechanics are getting broken in, it's running much smoother, and I see no need to cushion or damp the motion. I'm very encouraged so far!
Regards,
Ray L.
-
Here's a somewhat better video: http://youtu.be/aDEgRS6PI60
Regards,
Ray L.
-
HIYA RAY I noticed that something is PUSHING your spindle up and down. You can see movement of the actual spindle during the change.
Just a thought, (;-) TP
-
HIYA RAY I noticed that something is PUSHING your spindle up and down. You can see movement of the actual spindle during the change.
Just a thought, (;-) TP
Terry,
No, not pushing, the quill is being moved intentionally as part of the process. When loading a tool, I move the quill down 0.050", to make certain the TTS holder is pushed fully into the spindle, so it will seat properly against the spindle nose when the drawbar is tightened. After the drawbar is tight, I move the quill back to home position. There is enough compliance designed into the system to accomodate this small motion.
Regards,
Ray L.
-
GOTCHA, (;-) Lookin GOOD , (;-) TP
-
I've been making progress on the ATC - albeit slowly, due to interruptions. The basic mechanism is proving quite reliable. I've let it run random toolchanges for hours on end, running hundreds of toolchanges along the way, and it has never once made a mistake. And that's running completely open loop! Next week, I hope to get the sensors mounted, so the control program will be able to confirm that each commanded motion has actually fully completed. With the sensor feedback, this thing should be pretty darned bullet-proof.
I have two other project I have to complete before I can actually put the ATC into service. First, I need to complete the new telescoping Y/Z way covers. The current ones are too wide, and prevent the ATC from going fully into its out-of-the-way park position. The second is finally installing a spindle sensor, as I need to be able to ensure the spindle has fully stopped before doing a toolchange. The sheet metal parts for the way covers are all cut and bent, so it's just a matter of about a days work to assemble them, and get them mounted on the machine.
In the meantime, I've started making the "final" ATC parts, to replace some of the existing ones, which were made specifically as disposable prototypes. I've re-made the pivot arm, and air cylinder attachments, and they came out beautiful. The remaining work on the ATC is to re-make a few of the minor parts, mount the "skirt", connect up the door operating cable, and mount the sensors. That will all take some time, but nothing really hard there. I expect when it's all done, I'll either powder coat the whole thing, or perhaps just polish the aluminum.
Regards,
Ray L.
-
Progress on the ATC has, once again, been interrupted by higher priority work, but it is nearing completion. I've just finished getting the "sensors" working, so all that's left is getting the door mechanism going, putting on the "skirt", and doing some minor clean-up. The mechanism is working flawlessly. In fact, it's been running continuous toolchanges since early this AM - nearing 1000 just today. So far, no problems whatsoever. And, now that the sensors are in place, I believe it will correctly detect any problems, and halt operation. I've tried everything I could think of to make it fail, and it has correctly detected every failure I've thrown at it, and simply stopped what it was doing, until I manually cleared the error condition.
On a related note, the functionality of my machine has been expanding for years, and has gotten to the point that the wiring around the KFlop has become something of a rats nest, and I'm starting to run short on I/Os. A friend and I started designing a KFlop-specific breakout board, and in the process of defining what functionality to include on it, I came up with what I think is a great idea to greatly reduce the overall amount of wiring, and the amount of required buffering of signals.
The signals connecting to the KFlop basically fall into a couple of very broad categories. First is the special function and time-critical signals, like axis
drive outputs, Home/Limit/Probe/E-Stop inputs, and a very few others, which MUST be connected directly to the KFlop. These make up about 30-40% of the total signal count. The remainder are not at all time-critical, nor do they really NEED to connect directly to the KFlop. This group includes spindle and coolant controls, power drawbar and tool-changer controls, etc.
A while back, I bought a VistaCNC USB pendant, which I proceeded to rip apart, completely re-writing the firmware for the embedded PIC processor, and converting it from USB to RS232, so it could communicate directly with the KFlop. The code in both the pendant and the KFlop is dead-simple, and the pendant has worked absolutely flawlessly since the day I first put it into service months ago. It uses a very simple purpose-built serial protocol. My
new hair-brained idea is to expand that protocol to allow additional external processors to be connected to that same RS232 link, by either a daisy-chained or star connection, so that all non-critical signals can be moved completely off the KFlop to any number of external processors, each of which can then be located near the device it controls, connected by only a serial link, rather than a mass of wires. Each processor will be separately addressed, so it will accept messages targeted to it, and ignore all others. All processors will be able to send messages to any or all other processors, including the KFlop. Each will be able to accept commands from the KFlop, or other processors, and to send status message (e.g. - current spindle speed) back to the KFlop.
I will re-configure my system so that there will be one processor in the pendant (it's there now, and will just require a few minor tweaks to support the new protocol), another in the E-box for several basic machine functions (spindle controls, coolant controls, etc.), one for the power drawbar and toolchanger, and one for the control panel (this is, essentially the pendant from hell - a console with MPG, and a bunch of dedicated-function buttons...), so I can operate the machine without a mouse.
So, for example, if the user want to release the drawbar, the KFlop will send a message to the PDB/ATC processor telling it to release the drawbar. Once the operation is complete, the PDB/ATC processor will send a message back to the KFlop either confirming the operation completed successfully, or return an error code, indicating why it failed. Similarly, to turn on the spindle, the KFlop will send a message to the spindle processor, telling it what speed to set the spindle to. The spindle processor will set the spindle PWM, turn on the spindle, and send periodic spindle speed updates until the spindle is turned off.
I plan to use off-the-shelf Arduino processor boards, because they're dirt cheap (starting at about $10), highly functional, and available in a broad range of capabilities, with a wide range of plug-in I/O expansion boards. But, ANY processor could be used, as long as it supports RS232. Since each will be
controlling only "local" devices, there is little need to any extra buffering or isolation, further simplifying the wiring. This will reduce the current mass of
cables going from the E-box to the machine down to just the motor/home/limit/probe wiring, and a single RS232 cable, daisy-chained to the
processor boards on the machine.
Regards,
Ray L.
-
Finally had time to make a new Y/Z telescoping way cover (you can see it in the video) that clears the ATC, and all is installed and working perefectly:
http://www.youtube.com/watch?v=54s0aoAT37o
I've also got the new serial network working. It's going to significantly reduce the amount of wiring on my machine, and especially reduce the number of connections back to the E-box.
All that remains on the ATC is installing the "skirt" and door, making a cover for the "works" on top of the carousel, and cleaning up the wireing and plumbing. Functionally, it's already working 100%.
Regards,
Ray L.
-
Nice work Ray! Looks like it is working very well.
-
Wondering if you guys might clear up a thing or two for me...
A few pages back, their was a link to an ebay auction for a new BT30 spindle. I understand about how the bellville washers are compressed, and this releases the tool because the grippers release when the bellvilles are compressed. What I dont get is- unlike a turret mill (standard BP), the BT30 spindle appears to be fixed as it relates to Z. it appears that the spindle itself does not travel, but the spindle housing with all associated parts and air cyl must travel together? am I understanding this correctly?? VMC= travelling spindle housing, BP= travelling spindle (in) housing that is stationary?
So, it appears to me that other than creating grippers, a taper and a place for the spring stack, that the biggect issue is having an air cyl that will travel with the spindle on a BP, so that the relationship between spindle and air cyl is maintained fixed like it is on a VMC?
Following is an example of difficulty of not having air cyl/spindle relationship fixed-
Tools are held on table in carrier, BP spindle must descend to tool carrier height. this means that the air cyl must be able to either travel down significantly to actuate the release of the tool or be a fixed in its relationship with spindle like on VMC.
I do understand that having the air cyl travel up and down like an air power drawabar can(has) been done, but would it not be easier (less sensors/fixed hard stops) to just have the air cyl travel with the spindle and thus maintain the fixed relationship?
Also, the BT30 spindle in the ebay auction looks as if the air cyl actuation puts a load on the spindle bearings when it is actuating? is this correct?
Thanks,
Nate.
-
I would say the best solution is to have the cylinder travel with the quill. Few drawbacks to it though, one is its more complex and another is you will need more headroom.
The Beaver mill I have has a travelling Hyd cylinder and it works well, I can clamp/Unclamp at any point of the quill travel.
Haven't looked at the eBay auction mentioned as I couldn't see it in the last few pages and didnt search further. On my mill the cylinder outer screws to the spindle and thus there is no force exerted on the spindle bearings.
Hood
-
"Real" machines don't have quills - the whole head moves up and down. If you have a knee mill, you move the knee up and down to swap tools from a table-mounted wine rack, and you can define that tools can only be released at the top of the quill travel. You'll want the knee automated anyway, for applying tool length offsets. A tool changer without support for tool length offsets would be pretty useless. Making a power drawbar that would work with the quill at anything but the full-up position would be a living nightmare. Doing it on a BP would be darned near impossible, because there is so little space inside the pulleys where it would have to fit. Besides, I really can't see any advantage to making it work that way. Making one that works in the full-up position is quite simple. With BT/ISO tapers, it's not a problem applying the drawbar load to the spindle bearings, as the load is relatively small (about #1300 pounds max for 30-taper), and the bearings can easily handle it. But, making a drawbar that does not apply any load to the bearings is also quite easy.
Regards,
Ray L.
-
HimyKabibble-
If I get you correctly, You are saying that if you have the wine rack style tool holder on the table, and you use the quill for descending to pick up the tools, and then have the knee do the actual z positioning?
or are you saying that the knee is only for z height offsets. and that the quill (once correct offset is established with the Knee) does all of the Z positioning?
Seems to me that one of the biggest shortcomings on the BP as a cnc mill is that the Z travel is not much. After doing my conversion, I think I get a total of 4.5"
When using the Knee for Z offsetting, do you have it set up to unlock the knee, drive up or down and then relock? is the repeatability/accuracy suficcient? I always thought of the knee as a gross positioning device, and never thought that once unlocked, moved and relocked that the XY position would be repeated better than .002"-.005"
A few years back I thought that if the air cyl had to travel with the quill, that this could be effected by having the CNC Z axis on a (converted) BP also carry some sort of outrigger setup on both sides that extended up and then over the top of the head. I also theorized that a clamping assembley could hold the last .750" of the bottom of the quill and this could extend out and then have risers that would extend up and then back over. this would create an external "frame" like structure directly tied to quill movement and could carry the air cyllinder. Thus tool changes coiuld be effected at any Z height.
Another thought I have had was to have a linear tool rack that travels in X just in front of and mounted to the sliding head dovetail on a BP. it could move back and forth to set locations in a 4th or 5th axis in mach3 and would have trays in the assembley that would slide out via air cyl extension to position under the quill in Y. thus X would be moved to position by an additional axis, Y would be by air cyl pushout and retract in Y would be via Spring pullback.
Z travel would still be extremely limited because it would take 2.5/3 inches of its travel to raise up and have the new tool position under it for capture. I suppose the tools could be pushed both out (away from the mill body) AND up, by air cyllinder thus eliminating the Z travel of the quill being used to change tools. The entire Z quill travel could then be used to do machining ops.
-
Additionally- Hood, thanks for chiming in. you have (for years now) always been MOST HELPFUL to me...
-
Hmmmm... just went to the basement and looked at my BP. Why not have the wine rack style tool holder located at a far extension of the table and have IT raise up and meet the quill for toolchange? this would save all quill movement for Z machining ops.
Like this.
Toolchange call.
full retract Z.
XY position to over current tool used empty position in wine rack.
Raise wine rack up to Quill that is fully retracted.
Tool unclamp and release into wine rack.
Winerack descend.
XY position to new tool location.
Winerack raise to quill.
clamp tool.
Winerack lower.
XY position to required program position for new tool.
With a box like construction with linear ways and a stepper and ballscrew, the winerack up/down could be easily controlled. it could also have a coolant/chip shield that moved out of the way when the winerack raised.
Thoughts?
-
Rays mill uses the quill for normal Z moves and he has the knee powered for tool offsetting.
No real need to lock/unlock the knee if you have the gibs adjusted correctly and the ways are good, its just another axis and you dont have to lock/unlock X and Y after all.
Hood
-
HimyKabibble-
If I get you correctly, You are saying that if you have the wine rack style tool holder on the table, and you use the quill for descending to pick up the tools, and then have the knee do the actual z positioning?
or are you saying that the knee is only for z height offsets. and that the quill (once correct offset is established with the Knee) does all of the Z positioning?
Seems to me that one of the biggest shortcomings on the BP as a cnc mill is that the Z travel is not much. After doing my conversion, I think I get a total of 4.5"
When using the Knee for Z offsetting, do you have it set up to unlock the knee, drive up or down and then relock? is the repeatability/accuracy suficcient? I always thought of the knee as a gross positioning device, and never thought that once unlocked, moved and relocked that the XY position would be repeated better than .002"-.005"
A few years back I thought that if the air cyl had to travel with the quill, that this could be effected by having the CNC Z axis on a (converted) BP also carry some sort of outrigger setup on both sides that extended up and then over the top of the head. I also theorized that a clamping assembley could hold the last .750" of the bottom of the quill and this could extend out and then have risers that would extend up and then back over. this would create an external "frame" like structure directly tied to quill movement and could carry the air cyllinder. Thus tool changes coiuld be effected at any Z height.
Another thought I have had was to have a linear tool rack that travels in X just in front of and mounted to the sliding head dovetail on a BP. it could move back and forth to set locations in a 4th or 5th axis in mach3 and would have trays in the assembley that would slide out via air cyl extension to position under the quill in Y. thus X would be moved to position by an additional axis, Y would be by air cyl pushout and retract in Y would be via Spring pullback.
Z travel would still be extremely limited because it would take 2.5/3 inches of its travel to raise up and have the new tool position under it for capture. I suppose the tools could be pushed both out (away from the mill body) AND up, by air cyllinder thus eliminating the Z travel of the quill being used to change tools. The entire Z quill travel could then be used to do machining ops.
No, you've got it backwards. You use the knee for applying tool length offsets, and inserting and removing tools in the spindle from the wine rack. You use the quill for all machining. Unless you have single machining operations that require more than 4.5" of quill travel (I never have!), this works fine. This is how I operate my knee mill. Position accuracy should not be an issue if your knee is properly adjusted, and locking it should be completely unnecessary.
Regards,
Ray L.
-
More thinking out loud-
If a TTS type of tool holder were to be used, then a special drawbar could be made that (instead of screwing down and in, to effect pulling and retaining the R8 collet into the spindles taper) instead has a threaded portion at the bottom to interface with the R8 collet, but is not intended to ever rotate except when initially setup. this drawbar could be additionally long at the top so as to act as a spring holder, springs could be stacked above the termination of the bridgeports splined spindle top, probably with a spacer from the top of the splines on up to where it would have full clearance above the BP's head bearing retaining cap. The extension could be made like this-
Grind out 1.5 inches of ID from the top of the spindle (splined portion) suficiently large such that a spindle extension (spacer) could have a portion of it that slides into the newly enlarged ID protion of the top of the spindle. With a mild shrink fit, and a well made (hardened) spacer, I think that it would act as a part of the spindle and could be concentric enough to allow for it spinning unsupported above the top of the mills head. With the "spacer" extending above the head of the mill, above the portion of the spacer that needs to rermain small enough to pass into and out of the BP head asdsembley, a land could be created at the perifery of the spacer, and any number or combination of die springs held on the spacer with the (moving) drawbar acting as the retaining cap to the spring(s). A threaded cap could effect perloading of the spring(s) and could also be the surface that an air cyl interfaced with to provide the unlocking downward force.
Really, semi-simmilar to the Mach1 design, except no pullstud (TTS type holders) and the spring placement is out and above the head, allowing for more/better/different spring setups.
Obviously, this could get dicey at rpms above 4000, and could be dicey below that if not constructed with strength, concentricity and balance being paid very close attention to...
Downside? one would be that the release would have the force transferred through the bearings of the spindle. However... were the "spacer" discussed above pinned into place at its interface with the top of the spindle, then the "spacer could have a groove cut into it that a sliding u shaped fork could extend into via air cyl that would act as a brace and would keep the release load from being transfered through the bearings. This assumes however that the u shaped "fork" would be coupled to the air cyl assembley. Easy to imagine...tough to explain.
-
You're kind of re-inventing the wheel. Power drawbars for every kind of toolholder imaginable have been done to death, and what you describe would be less than ideal in a number of important areas. I'd suggest you wander over the CNCZone, and see what other people have done for power drawbars and toolchangers. There's little new under the sun....
Regards,
Ray L.
-
Sure, I have a membership at the zone, but I have to disagree, I think the design I propose has quite a few advantages.
Namely-
springs can be altered.
springs can be of various sizes.
Air cyl can be used instead of impact or gearmotor.
Sensors not needed. Only one hard stop for downward travel.
table mounted tools, no swinging arm, carosel etc...
no dependance upon the air ratchet type PDB engaging correctly.
No need to have a gearmotor with its drive, switches sensors etc, not to mention making sure it engages correctly before turning.
-
Sure, I have a membership at the zone, but I have to disagree, I think the design I propose has quite a few advantages.
Namely-
springs can be altered.
springs can be of various sizes.
Air cyl can be used instead of impact or gearmotor.
Sensors not needed. Only one hard stop for downward travel.
table mounted tools, no swinging arm, carosel etc...
no dependance upon the air ratchet type PDB engaging correctly.
No need to have a gearmotor with its drive, switches sensors etc, not to mention making sure it engages correctly before turning.
Pretty much all of those, with the exception of "springs can be of various sizes" can be accomplished with a more conventional design, and you won't have to deal with the VERY considerable issue of keeping a rotating spring stack and housing well enough centered and balanced as to not be a serious safety hazard. Keep in mind that "extension" will rise above the end of the spindle by at LEAST the amount of quill travel, and it will, necessarily, be rather large and heavy, making perfect centering and perfect balance an absolute necessity. MANY power drawbars use air cylinders, hydraulic cylinders, air-over-hydraulic systems, linear actuators or even motors (see mine, earlier in this thread). Keep in mind, too, that if you're talking R8 or TTS, you need a BIG air cylinder - they are typically done with about a 4" diameter triple-stack cylinder to get sufficient force (required drawbar force for TTS is about 2500#, unless you're running a fractional horsepower machine). Sensors are not needed for any of these. And, even if you feel they are, what's the problem with using $3 micro switch or inductive proximity sensor? A table-mounted wine rack tool rack can be made to work with any of these. There are several examples on CNCZone. Making sure a rotating PDB engages before starting rotation is also not necessary - I don't bother to do that on mine, and it works perfectly. If it does engage on the initial downward movement, it will as soon as the socket and drawbar rotate into the proper position, so it's self-engaging. In fact, I don't think I've ever seen a power drawbar that makes any effort to ensure engagement before rotation starts. You're creating solutions in search of problems.
Regards,
Ray L.
-
"You're creating solutions in search of problems."
Might just be... You are not the first person to tell me that... lol...
-
having studied how you did your stepper/planitary gearbox PDB, I will now eat crow and tell you that it is genius. I now get the relay selecting the current limiting resistor and that is really smart.
Does it tighten at 25ftlbs by rotating a set number of degrees/rotations, and if the motor stalls (stall point set by current limiting resistors) then it is tight?
Opening is at 75ftlbs, rotating a set number of degrees/rotations and it does not stall, just rotates a certain amt to effect release?
Am I getting this right?
I use G203V's, and if I were to do this, I would have to make sure the motor did not hard stop, as this would fault the drive (red light) and require a reset. Now I do resets by de-powering, It does not look as if the 203V's have the ability to be reset except for by de-powering...
is the air cyllinders down force suficcient to lock the air cylinder in position on the loosen stage, so that the R8 toolholder is forced into the fully loose position? if not, how did you accomplish the downward hammering motion to effect loosening of the collet?
-
You've got it basically correct. Current limiting in the Gecko determines torque. I limit torque during tightening, and run full torque for loosening. I loosen 1-1/4 turns, and on tightening I go 1-1/2 turns, stop, the go an additional 1/4 turn. In most cases, the first move will stall near the end, and the second is just ensuring full tightening torque is reached reliably.
There is no reason this would not work just fine with a G203V. When the drawbar stops, the Gecko will limit current, and should not fault. It will only fault if the motor outputs are shorted.
Yes, the air cylinders provide enough down-force to pop the collet free of the spindle. Wouldn't be a workable PDB if it didn't. There is no "hammering" motion, it is simply about 300# of down-force.
Regards,
Ray L.
-
I could, of course, also go with a more complex linkage, but I really don't want to go there unless I have no choice.
Oops! I was going to suggest you articulate as that is what I decided I would need to do when I was considering and actuator similar to what you have . . . but . . uh . . never mind.
Hey, is that friggin' thing done or WHAT??!!
My 60 day project is completed. I have another project that will be wrapped up in about a week and I have material on its way to finish my ATC.
Tick Tick Tick . . do I smell Chili Beer? :D
-
Hey Steve, not sure if you saw it or not but heres the Chirons changer working :)
http://www.youtube.com/watch?v=pILjZ-JXCRQ
Hood
-
OK sports fans, I am almost to the point of working on this ATC again.
Here is the feeble sum total of my progress over the last 70 days or so . . a couple of mounts. I included an older shot (second photo) of the arm actuator to show the air cylinder that will drop the arm vertically.
I had some other weling to do to make a new ficxture, so I made up the carousel mounting arm while I had everything dragged out for welding. Then I was doing some surfacing, so I grabbed the arm off the shelf and cut the mounting pad. I was doing some boring, so I grabbed the swing arm transmission mount . . etc etc. Little bit here little bit there.
However, sometime towards the end of next week, I will be spending about half of my time finishing this thing up. I have some material coming to partially build the new mill head so that I can mock up the final assembly and and start testing.
Oh and incidentally, I have a big fat die spring for the next drawbar and I'll be showing that coming up pretty soon. Bye Bye belleville
(http://www.thecubestudio.com/BT30SpindleProject/SwingArmCompleteWithMountWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/ATCswingArmDriveWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/ATCmountingArmWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/CarouselMountingArmBottomWEB.jpg)
-
Hey Steve, not sure if you saw it or not but heres the Chirons changer working :)
http://www.youtube.com/watch?v=pILjZ-JXCRQ
Hood
Yes I saw this thing before so I guess you posted it somewhere. This is the coolest example of mechanical intercourse I have seen in many moons. The tool 'pods' on my carousel are modelled after a setup you showed . . I think it was on a machine you were thinking about aquiring. Does that sound right? This does not appear to be the same machine.
-
Unless you saw it in the last few days that was not my Chiron but just a vid pulled off of youtube, I just got the Chiron changer working from Mach a few days ago, still not ordered the controller yet, that will have to wait until the new year :(
The one you modelled your pots on was a Beaver VC5 that was at the same place I bought the Chiron from.
Hood
-
Oops. my bad . . . I jumped back into this thread where I left of not realizing there was a ton more stuff that happened after that!
Ignore everything I said, RAY . . until I can catch up on the rest of the progress between October and now. Should make an interesting read, but can't do it just now. Busy busy . . .
-
I could, of course, also go with a more complex linkage, but I really don't want to go there unless I have no choice.
Oops! I was going to suggest you articulate as that is what I decided I would need to do when I was considering and actuator similar to what you have . . . but . . uh . . never mind.
Hey, is that friggin' thing done or WHAT??!!
My 60 day project is completed. I have another project that will be wrapped up in about a week and I have material on its way to finish my ATC.
Tick Tick Tick . . do I smell Chili Beer? :D
Steve,
Actually, the ATC is working just fine, I just can't prove it right now.... But it has run literally thousands of toolchanges in testing without a single error. I went out a few days ago to cut the very first parts using the ATC (the little brackets to hold the "skirt" on the carousel), and ran face-first into an unrelated failure, and I now have the spindle out, and almost the entire head disassembled, waiting for some new bearings and a few other parts to show up. I expect the bearings to arrive probably tomorrow, but I'll be out of town for a few days, so won't get to work on it again until at least Tuesday. But, I am hoping that this will *finally* resolve my long-standing "rattly head" problem. Well, the one on the mill anyway. I'm pretty sure the other one is beyond hope at this point....
I'm also thinking seriously about buying a Novakon Torus Pro/Servo bed mill....
Regards,
Ray L.
-
Unless you saw it in the last few days that was not my Chiron but just a vid pulled off of youtube, I just got the Chiron changer working from Mach a few days ago, still not ordered the controller yet, that will have to wait until the new year :(
The one you modelled your pots on was a Beaver VC5 that was at the same place I bought the Chiron from.
Hood
Oh my, that monster is your own personal mechanical Medusa :o That thing is just a freak!
I will have a lot of questions for you on the controller because that will be my next challenge after the mechanics are completed. In the mean time, I'll just ponder where Medusa keeps putting those things she's holding in her 'hands' :P
-
I could, of course, also go with a more complex linkage, but I really don't want to go there unless I have no choice.
Oops! I was going to suggest you articulate as that is what I decided I would need to do when I was considering and actuator similar to what you have . . . but . . uh . . never mind.
Hey, is that friggin' thing done or WHAT??!!
My 60 day project is completed. I have another project that will be wrapped up in about a week and I have material on its way to finish my ATC.
Tick Tick Tick . . do I smell Chili Beer? :D
I expect the bearings to arrive probably tomorrow,
Ray . . you don't really think I'm falling for the old 'the bearing is in the mail' routine , do ya?
Seriosuly, what I saw from october looked very nice. I can hardly wait to find time to catch up on what you've acomplished. I'm very interested to see if and how you managed the actuation force at the ebds of the 140 degree sweep.
I really gotta fly now but I'll be participating here in the next few.
-
I could, of course, also go with a more complex linkage, but I really don't want to go there unless I have no choice.
Oops! I was going to suggest you articulate as that is what I decided I would need to do when I was considering and actuator similar to what you have . . . but . . uh . . never mind.
Hey, is that friggin' thing done or WHAT??!!
My 60 day project is completed. I have another project that will be wrapped up in about a week and I have material on its way to finish my ATC.
Tick Tick Tick . . do I smell Chili Beer? :D
I expect the bearings to arrive probably tomorrow,
Ray . . you don't really think I'm falling for the old 'the bearing is in the mail' routine , do ya?
Seriosuly, what I saw from october looked very nice. I can hardly wait to find time to catch up on what you've acomplished. I'm very interested to see if and how you managed the actuation force at the ebds of the 140 degree sweep.
I really gotta fly now but I'll be participating here in the next few.
Steve,
It was a matter to optimizing the geometry of the arm and air cylinder to the extent I could, using flow control valves on the air cylinder to keep it from moving at 100 MPH, and rubber bump stops at the end. Very little force is required to engage with the tool in the spindle. I may end up adding an air damper at the "parked" end to smooth out the stop, but it's working fine as it is.
Regards,
Ray L.
-
I could, of course, also go with a more complex linkage, but I really don't want to go there unless I have no choice.
Oops! I was going to suggest you articulate as that is what I decided I would need to do when I was considering and actuator similar to what you have . . . but . . uh . . never mind.
Hey, is that friggin' thing done or WHAT??!!
My 60 day project is completed. I have another project that will be wrapped up in about a week and I have material on its way to finish my ATC.
Tick Tick Tick . . do I smell Chili Beer? :D
I expect the bearings to arrive probably tomorrow,
Ray . . you don't really think I'm falling for the old 'the bearing is in the mail' routine , do ya?
Seriosuly, what I saw from october looked very nice. I can hardly wait to find time to catch up on what you've acomplished. I'm very interested to see if and how you managed the actuation force at the ebds of the 140 degree sweep.
I really gotta fly now but I'll be participating here in the next few.
Steve,
It was a matter to optimizing the geometry of the arm and air cylinder to the extent I could, using flow control valves on the air cylinder to keep it from moving at 100 MPH, and rubber bump stops at the end. Very little force is required to engage with the tool in the spindle. I may end up adding an air damper at the "parked" end to smooth out the stop, but it's working fine as it is.
Regards,
Ray L.
My first claw was a fail even though it worked, primarily because it did not have enough spring pressure to retain a heavy tool during extreme decel at the end of travel. Since I am using a servo motor to drive the arm, I will be side stepping all of the challenges of a pneumatic actuator . . albeit at a very high cost.
I noodled over air driven arm for quite a while before abandoning that route. My design was articulated so there was no problem with the swing range or forces, but stopping the thing at exact point along the travel was something I was not able to work out to my satisfaction. :-[
You move the entire carousel, so you don't have an intermediate stop to deal with, but it looks like you did run into the end-of-travel issues that I was concerned with. Kudos to you for getting something working!
FWIW, one method that I considered was an air operated metal band positioner. I would have needed an encoder somewhere for the intermediate stop, but it would be a simple matter for put a caliper on the metal band for a positive stop anywhere in the travel. This might eliminate the throttling and damping needed in a free swinging arm. A rack and pinion arrangement is also doable, but enclosing it is a PIA and they don't like the taste of swarf.
Anyway, I'm going to catch up on this thread now and probably have a bunch more comments and questions.
-
Steve,
At the spindle end of the travel, I provided a "hard" stop, to ensure proper positioning. This consists of just a bolt with a rubber bumper on the end that contacts a small bracket bolted to the side of the head. At the other end of the travel, I just allow the air cylinder to bottom out. I agree a servo or stepper is a more elegant, if more expensive, way to do this, and would allow tuning to get constant-speed motion, or even a motion profile that should negate the need for stops, dampers, etc. I also used prox sensors to ensure each motion fully completes before taking the next step, so if the carousel for any reason doesn't reach the fully engaged position, there;s no harm done, the toolchange simply stops until the jam is cleared manually. Had I not needed such a wide range of travel (140 degrees), I could've certainly made this work better. Were I doing it over again (and I still might....), I'd make the carousel fixed, and use a transfer arm. I have an idea for that which would be quite simple, using a rotating "claw" on the end of the arm, which would articulate such that the claw was pointed towards the pivot when the arm was rotating, allowing the arm to move much faster, with no worries about the tool flying out.
Regards,
Ray L.
-
I just measured the swing on mine. Are you sitting down?
213 degrees from carousel pickup to spindle. Carousel pickup to the arm 'parking pot' is another 37 degrees.
Total swing 250 degrees! :o I doubt I could even articulate that much swing.
140 degrees would be no problem though, if you wanted to add that. Would be very simple and give you close to full power ate each end of the swing.
What processor are you using to control the sequence and read sensors? ( Speaking of processing, I have a lot of questions about your Kflop project. Is there a thread going on that?)
-
. . . I just got the Chiron changer working from Mach a few days ago, still not ordered the controller yet, that will have to wait until the new year :(
Hood
It is working, but you don't have the controller yet? How does this happen? What did I miss ???
-
. . . I just got the Chiron changer working from Mach a few days ago, still not ordered the controller yet, that will have to wait until the new year :(
Hood
It is working, but you don't have the controller yet? How does this happen? What did I miss ???
Mach is talking to my PLC :)
Got it fully working today, few days back I jumpered a wire to the Tool 9 coil so it will now raise and today I got the cylinder/prox's and arm fitted to Tool 10 and arm to tool 7. Only thing I have to do now is make up some doors for a couple of the holders, using tape on one at the moment ;D
Heres a vid of all 12 tools being called, http://youtu.be/oJ8KAuzgSjE
Forgot to screw the damper on the cylinder of tool 10 so it comes down with a clunk but I think I might just manage to get my hand in to damp it, if not I will drop the cylinder and get it done as I dont like it hitting the end so fast.
Hood
-
"DANGER Robin Hood DANGER"
(;-) TP
-
I just measured the swing on mine. Are you sitting down?
213 degrees from carousel pickup to spindle. Carousel pickup to the arm 'parking pot' is another 37 degrees.
Total swing 250 degrees! :o I doubt I could even articulate that much swing.
140 degrees would be no problem though, if you wanted to add that. Would be very simple and give you close to full power ate each end of the swing.
What processor are you using to control the sequence and read sensors? ( Speaking of processing, I have a lot of questions about your Kflop project. Is there a thread going on that?)
Steve,
Nope, you're not going to do 213 degrees without a 4-bar linkage. I considered that approach, but it's too difficult to design to get the desired motion.
Right now, it's running entirely on the KFlop, but I'll be moving it to an Arduino shortly. Actually, I'm moving most of my external devices (coolant, PDB, ATC, pendant, control panel) to a network of Arduinos. It will really cut down on I/O usage on the KFlop, and the amount of wiring in the whole systems.
Ask away on the KFlop. No thread on that topic. Wouldn't really belong here on the ArtSoft forum, given that I'm using my own home-grown CNC controller app, rather than Mach3.
Regards,
Ray L.
-
. . . I just got the Chiron changer working from Mach a few days ago, still not ordered the controller yet, that will have to wait until the new year :(
Hood
It is working, but you don't have the controller yet? How does this happen? What did I miss ???
Mach is talking to my PLC :)
Got it fully working today, few days back I jumpered a wire to the Tool 9 coil so it will now raise and today I got the cylinder/prox's and arm fitted to Tool 10 and arm to tool 7. Only thing I have to do now is make up some doors for a couple of the holders, using tape on one at the moment ;D
Heres a vid of all 12 tools being called, http://youtu.be/oJ8KAuzgSjE
Forgot to screw the damper on the cylinder of tool 10 so it comes down with a clunk but I think I might just manage to get my hand in to damp it, if not I will drop the cylinder and get it done as I dont like it hitting the end so fast.
Hood
Very 8) Now just add remote control and you can keep your tools anywhere . . . .
(http://www.thecubestudio.com/pictures/DroneHoodWEB.jpg)
-
I just measured the swing on mine. Are you sitting down?
213 degrees from carousel pickup to spindle. Carousel pickup to the arm 'parking pot' is another 37 degrees.
Total swing 250 degrees! :o I doubt I could even articulate that much swing.
140 degrees would be no problem though, if you wanted to add that. Would be very simple and give you close to full power ate each end of the swing.
What processor are you using to control the sequence and read sensors? ( Speaking of processing, I have a lot of questions about your Kflop project. Is there a thread going on that?)
Steve,
Nope, you're not going to do 213 degrees without a 4-bar linkage. I considered that approach, but it's too difficult to design to get the desired motion.
I'm using the term 'articulated' generically. We're likely talking about the same thing; common example would be this hoe bucket. Total sweep on the swing arm is actually 250 degrees. I'm not sure if that is doable even with such a linkage. Certainly 140 degrees would be no problem. Early on I had an articulated arm design, but at that point the rotation was something like 190 degrees. Later the sweep increased to 250 degrees total, but by then I had elected to go with a motor and transmission for other reasons, so I never looked into it.
(http://www.thecubestudio.com/pictures/BackHoeWEB.jpg)
Right now, it's running entirely on the KFlop, but I'll be moving it to an Arduino shortly. Actually, I'm moving most of my external devices (coolant, PDB, ATC, pendant, control panel) to a network of Arduinos. It will really cut down on I/O usage on the KFlop, and the amount of wiring in the whole systems.
I will use an Arduino also. I just got a shipment of them in for another project so I just 'threw another shrimp on the barbie'. I use the highest model which has three serial ports built in, and I use all of them. So far as I know, there is no arbitration built into the RS232 standard, so you will have your hands full using it in the way you intend . . at least it seems that way to me. Why did you choose RS232 over USB or Ethernet?
Ask away on the KFlop. No thread on that topic. Wouldn't really belong here on the ArtSoft forum, given that I'm using my own home-grown CNC controller app, rather than Mach3.
Regards,
Ray L.
My questions would be limited (at this time) to using the Kflop with MACH as an alternative to the Smoothstepper using Kflops's MACH plug in. For beter or worse, I am pretty much married to MACH at this point. :'( If you have not used the Kflop in that manner, then I have no further questions, your honor.
-
"DANGER Robin Hood DANGER"
(;-) TP
Looks like Will Robinson sold the robot. I would have kept the robot and sold Dr. Smith . . .
-
Right now, it's running entirely on the KFlop, but I'll be moving it to an Arduino shortly.
Ray L.
OK, then the game is still on. 8)
Contest rules say final form including control and all sensors. No tortoise in sight yet, but I have the scent.
-
Steve,
"Articulated" and "4-bar" are actually different animals. 4-bar refers to a specific type of linkage that is simple (made of, surprisingly, 4 bars each with pivots at each end, and those pivots connecting them all together. Googling "4-bar linkage" will give you many examples. This class of linkage, properly designed, is capable of producing nearly ANY motion profile desired. But, as I said, the design of such a linkage for complex motion profiles can be quite difficult. Introductory 4-bar design was the topic of an entire semester-long class when I was in college. Obviously an articulated arm is a viable option, but an expensive one, due to the extra joints, and extra actuators/sensors required. I ruled that approach out on mine pretty early. I'm sure it makes good sense on yours.
I used RS232 because it's simple, cheap, robust, and is built into virtually every MCU on the planet. USB is too finicky, not good over any distance, and NOT good in a noisy environment. RS232 will handle all of those easily. For what I'm doing, high speed is not required. I created a very simple "network" protocol that allows me to have up to 10 devices in the network, with a combination of "star" and "ring" connections, for maximum flexibility. The KFlop is the network "Master", and all other devices are, essentially, intelligent "slaves". Right now, I've implemented on controller that handles the spindle, coolant, and air blasts, and also acts as the network "hub", since the KFlop has only one serial port. The pendant and control panel devices are on the "ring" side of the network, and the PDB and ATC are on the "star" side. The KFlop can simply send a serial command that says, for example, "Turn on Flood", "Load Tool #5", etc., and each device does what it's told, and reports back status to the KFlop. This makes the only connections to the KFlop for all these peripherals the serial link, and all the I/Os, control logic, relays, sensors, etc. move to the MCU on each peripheral. Since all are running the same "network stack", 90% of the code on each device is identical, with only the part that implements each controllers unique functionality being different for each. The code is very simple, and very robust, and can easily be ported to virtually any MCU.
Re: Mach and KFlop - I did do a complete implementation of Mach3 with KFlop, but abandoned it once I got my own app going back last January. I was able to do this in such a way that 95% of the code was common between my Mach3 implementation, my KMotionCNC implementation, and what I now use for my own app. So, I could, if I wanted, resurrect Mach3 probably in a few hours. I just haven't wanted to.... Anyway, if you have questions or need help getting it going under Mach3, don't hesitate to ask. And, of course, you're welcome to my code if you want it.
Regards,
Ray L.
-
Right now, it's running entirely on the KFlop, but I'll be moving it to an Arduino shortly.
Ray L.
OK, then the game is still on. 8)
Contest rules say final form including control and all sensors. No tortoise in sight yet, but I have the scent.
Hey! No changing the rules mid-game! :-) The dedicated controller was not part of the original design, and won't be completed for a while, as its part of a larger effort to completely re-build my E-Box, which will be timed by availability of a custom BOB being designed by a friend.
Regards,
Ray L.
-
Ray has already WON the competion. I just need an address to send him a beer of his choice (;-)
(;-)TP
-
Right now, it's running entirely on the KFlop, but I'll be moving it to an Arduino shortly.
Ray L.
OK, then the game is still on. 8)
Contest rules say final form including control and all sensors. No tortoise in sight yet, but I have the scent.
Hey! No changing the rules mid-game! :-) The dedicated controller was not part of the original design, and won't be completed for a while, as its part of a larger effort to completely re-build my E-Box, which will be timed by availability of a custom BOB being designed by a friend.
Regards,
Ray L.
Contest rules from October 14th:
Contest rules: everything does not have to be powder coated or otherwise 'spiffy', but the automatic tool change has to be . . well . . automatic, and under its own power. No manual operation and no . . uh . . temporary actuators . . .
However, other . . 'favored contestants' will be allowed to demonstrate the completed ATC on a totally non functional mock up of the milling machine head . . . because there is no actual mill yet . . and because he is making the rules.
Contest winner is responsible for any and all federal, state and local taxes on prizes, which may include a case of the contestant's favorite beer . . and bragging rights.
Therefor I maintain that the game is still on. The A in ATC is for Automatic . . not Almost. Until one device is competed and functioning automatically in final form, no winner.
Besides, I don't want the contest to be over . . this is way too much fun. ;)
-
Steve,
"Articulated" and "4-bar" are actually different animals. 4-bar refers to a specific type of linkage . . . .
Articulated is just another way of saying 'jointed' so I would submit that a 4bar is a specific type of articulation . . although there are numerous types of 4 bar arrangements, so I would then use the term 'specific' only in a general way. How's that for an oxymoron?
I created a very simple "network" protocol . . . .
There is the answer. I have not the time nor talent to create a protocol, so I shall have to be satisfied to huddle with the rest of the masses.
Anyway, if you have questions or need help getting it going under Mach3, don't hesitate to ask. And, of course, you're welcome to my code if you want it.
Regards,
Ray L.
Much appreciated.
I'm looking to get rid of a USB smoothstepper and early reports on the new Ethernet version were all good, but now it is starting to look a bit like the earlier incarnation with problems going unfixed for extended periods. Kflop seems to be on top of things and responsive to questions.
On topic, of particular interest is the Kflop is extensible and may be a better portal for the ATC operational data than trying to communicate with MACH. I like the idea of having Kflop orchestrate separate CPUs which each have their own process to control. Kflops user code interface is where I would seek whatever advice you might be willing to provide, but the prerequisite of course is to have the Kflop.
To that end, initially all I really need to hear is (regarding the MACH plug-in) 'It works' or 'It works except for this bug, that anomaly, this workaround, etc' (in which case I would pass).
Your endorsement would be enough for me to acquire a Kflop and invest some time in it.
-
Programmers ya just gotta like um.
(;-) TP
-
Articulated is just another way of saying 'jointed' so I would submit that a 4bar is a specific type of articulation . . although there are numerous types of 4 bar arrangements, so I would then use the term 'specific' only in a general way. How's that for an oxymoron?
I think the distinguishing characteristic would be the independence of motion of the various arms in the system. In a 4-bar, there is a fixed relationship between them, regardless of position, as the arms are pinned together at the pivot points. If you know the position of any one of the four arms, you know the position of all of them. To me, at least, an articulated arm implies having two or more independently controllable arms connected together, so you need multiple data points to know the position of the business end of the arm. Semantics.....
There is the answer. I have not the time nor talent to create a protocol, so I shall have to be satisfied to huddle with the rest of the masses.
I mean VERY simple, and easily modified/extended. If you'd like to see what it does, I'd be happy to fill you in. The protocol can be used as-is, with the specific "command set" defined by just changing a few very simple files. It's partitioned so the command set can be changed without touching the underlying "routing" code, and the hardware-specific parts of it are isolated in a few files, so the hardware platform can be changed quite easily, to move it to another MCU. I had to do this, since the same base code must be able to run on both the KFlop and the Arduinos. The only downside is this also force me to write it in old-fashioned C, rather than C++ or some other modern OOP language, which would have made the code a lot "cleaner". But, I did it in a pseudo OOP manner, so it's still quite easy to follow. And, again, the only parts you'd ever need to mess with are really dead-simple, and just have to do with validating incoming messages, and actually executing the commands. All the housekeeping of the actual "protocol" is handled by the low-level code, which neither knows nor cares what the platform is, nor what the messages actually contain, or what they actually do - it just makes sure they get where they're supposed to, and that valid responses find their way back to the initiator.
I'm looking to get rid of a USB smoothstepper and early reports on the new Ethernet version were all good, but now it is starting to look a bit like the earlier incarnation with problems going unfixed for extended periods. Kflop seems to be on top of things and responsive to questions.
If you'd like to try one, I have one (and several USB ones) I'd be happy to loan you.
On topic, of particular interest is the Kflop is extensible and may be a better portal for the ATC operational data than trying to communicate with MACH. I like the idea of having Kflop orchestrate separate CPUs which each have their own process to control. Kflops user code interface is where I would seek whatever advice you might be willing to provide, but the prerequisite of course is to have the Kflop.
To that end, initially all I really need to hear is (regarding the MACH plug-in) 'It works' or 'It works except for this bug, that anomaly, this workaround, etc' (in which case I would pass).
Your endorsement would be enough for me to acquire a Kflop and invest some time in it.
On that basis, I would recommend KFlop without hesitation. Tom Kerekes, the designer, is readily available, and extremely responsive. He's already made several changes based on my input, and has always helped come up with a solution to any problem I've thrown at him. I usually get a response within hours. My only hesitation would be that the learning curve on the KFlop can be a bit of a bugger. But, I'd be happy to give you assistance, and get you pointed in the right direction, so you can largely avoid flailing around in the dark as I had to at times.
You could easily have the basic machine up and running within an hour or two after unpacking the KFlop. Adding all the ancillary features can then be done incrementally.
As far as stability, I'm aware of no problems with the KFlop plug-in, and certainly saw none when I did work with it, albeit briefly. I've never seen any problems raised on the Dynomotion Yahoo Group, which it the primary support portal. If you give me information on your machines configuration, I could probably provide sample code that would get you started with only minor modification and debug. Significant parts of the code I have on my machine might well work on yours with only minor modification. My homing and probing code (perhaps the most complex code, other than my pendant "driver"), for example, I suspect you could use as-is.
With the KFlop, you do have to write code for things that you get out of the box with Mach3. For instance, you have to write code to control the spindle speed, do homing, probing, handling limits, etc. Very simple controls like relay outputs for coolant, etc, can be handled without code, but things like a PWM-driven spindel require some simple C code to be written. But, if you're not doing anything exotic, this code will generally be quite simple, in comes cases only a few lines. If you're at all familiar with C, you'll have no problems whatsoever. But, if you want exotic, it can easily handle that as well. I've added some unusual capabilities to mine, but the code is still trivial. For example, I can re-program all axis parameters on the fly from the PC application. This allowed me to store all configuration data in a single XML file, rather than having some in the C code, some in the application, and some hard-coded. You can also do things like have the KFlop processor hand off lines of G-code to be executed by the CP application, which comes in handy at times.
The KFlop is VERY powerful, and almost infinitely flexible. Once you understand it, it's a truly wonderful thing to work with. As an example, the code to get my ATC working took, total, probably two hours to write, counting the several iterations as the hardware progressed. It stands right now at perhaps two pages of dead-simple code with lots of white space.
Regards,
Ray L.
-
Steve better be making progress! My ATC officially went into service yesterday, and it has now run its first three jobs without a single problem. Today I got the "skirt" on the carousel, so all I have left to do is to get to a bicycle shop for the cable to operate the carousel door, and then do a cosmetic clean up the wiring and plumbing. Other than that, I'm done!
I did try to get a video of it in-action, but coolant and the table enclosure make it pretty near impossible to get anything you can actually see, without the camera taking a bath. Since it's a brand-new camera, I'm reluctant to subject it to that...
Regards,
Ray L.
-
And, we now have a door on the carousel. Just need to get into town and pickup a spring (to close the door) and a bicycle brake cable (to open the door), and we're done!
I can taste Steve's beer already.... :-)
Regards,
Ray L.
-
Ray, truth be told, I fully expect you to win. I just want the competition to be at the highest level, and I know you are capable of that. I have a selfish motivation, of course, but it also hopefully benefits the many readers of this thread to observe the process.
I think we agree that the mechanics, complicated as they may be, are not the lion's share of an ATC, but rather the control and sensors. I would speculate that those who can build the mechanism are many, those who can create an autonomous and safe control system are few. That's where I see the real competition. The finish line includes all sensors and 'release candidate' level fully functional control code. I also want to follow what Hood and other 'non-contestants' are doing with regard to sensors and control code.
To consolidate your questions from the last couple posts, let me say that yes, I am familiar with 'C'. There are several thousand lines of it in my InTurn™ motor control system. And less volume but far greated intensity code in my balancer. Although I am a lightweight with electronics, I get by with the 90% learn and 10% do methodology.
I am going to acquire a Kflop board based on your endorsement. Initially, I just want to replace the smoothstepper, but I have similar needs to yours down the road, so the sizeable learning curve is an investment that will pay off, methinks. You mentioned loaning a board, but the way it is worded, I cannot tell if you mean loading an ethernet Smoothstepper or a Kflop board. ???
As to machine complexity, I have 4 axis of Mitsubishi J3 series industrial drives (and motors) and the spindle is a Copley Controls Xenus drive pushing a big DC brush motor. I am starting to talk to these devices in real time to get at some of the pertinent staus info they provide and create a 'reactive' system that can act on the data from the drives to adjust the paramaters on a CNC machine in real time. Another machining center feature to bring down to we mere mortals. Hence my interest in your amazing work with the Kflop. Yeah, I did some homework on that.
Anyway, I threw this machine together quick and dirty to get enough Y travel to do a specific project. I have since boxed and welded up the column and now it has decent performance. My 'permanent' mill will be similar, but larger, much heavier and of course will have the spindle and tool changer from this thread. Since the spindle is completed (except for the die spring) and the ATC is almost done, I am revisiting the pile of parts and the frame design. Today I purchased a 9" x 42" Bridgeport mill table to add to the pile.
There has been no progress on my ATC because I had some unanticipated work to accomplish, however, I expect to start making progress again on the ATC today. I am simultaneously making a new spindle from A6 tool steel, (which will be fully hardened and ground) and ditching the bellevilles. I expect there will be a lot of interest in my big fat die spring because it *probably* can be implemented on an existing spindle which has no way to get the typical disc spring arrangement installed. Next couple days I will post some photos of the ATC assembled on the side of the new head.
I see from the photos you posted that the swing arm has been eating its Wheaties. Very nice piece! 8) I also note the pivot has an upper and lower mounting. Double nice. Significant improvements from the prototype, as one would expect. Q: curiosity, how did you form the smooth curve in the skirt?
-
Ray, truth be told, I fully expect you to win. I just want the competition to be at the highest level, and I know you are capable of that. I have a selfish motivation, of course, but it also hopefully benefits the many readers of this thread to observe the process.
I think we agree that the mechanics, complicated as they may be, are not the lion's share of an ATC, but rather the control and sensors. I would speculate that those who can build the mechanism are many, those who can create an autonomous and safe control system are few. That's where I see the real competition. The finish line includes all sensors and 'release candidate' level fully functional control code. I also want to follow what Hood and other 'non-contestants' are doing with regard to sensors and control code.
Actually, the sensors and code on mine turned out to be quite trivial. The code is barely two pages of dead-simple C that took only maybe two hours to write and debug.
To consolidate your questions from the last couple posts, let me say that yes, I am familiar with 'C'. There are several thousand lines of it in my InTurn™ motor control system. And less volume but far greated intensity code in my balancer. Although I am a lightweight with electronics, I get by with the 90% learn and 10% do methodology.
I am going to acquire a Kflop board based on your endorsement. Initially, I just want to replace the smoothstepper, but I have similar needs to yours down the road, so the sizeable learning curve is an investment that will pay off, methinks. You mentioned loaning a board, but the way it is worded, I cannot tell if you mean loading an ethernet Smoothstepper or a Kflop board. ???
Must've been a SmoothStepper I was referring to. I only have the oneKFlop of my own, and a backup that actually belongs to a friend.
As to machine complexity, I have 4 axis of Mitsubishi J3 series industrial drives (and motors) and the spindle is a Copley Controls Xenus drive pushing a big DC brush motor. I am starting to talk to these devices in real time to get at some of the pertinent staus info they provide and create a 'reactive' system that can act on the data from the drives to adjust the paramaters on a CNC machine in real time. Another machining center feature to bring down to we mere mortals. Hence my interest in your amazing work with the Kflop. Yeah, I did some homework on that.
Anyway, I threw this machine together quick and dirty to get enough Y travel to do a specific project. I have since boxed and welded up the column and now it has decent performance. My 'permanent' mill will be similar, but larger, much heavier and of course will have the spindle and tool changer from this thread. Since the spindle is completed (except for the die spring) and the ATC is almost done, I am revisiting the pile of parts and the frame design. Today I purchased a 9" x 42" Bridgeport mill table to add to the pile.
There has been no progress on my ATC because I had some unanticipated work to accomplish, however, I expect to start making progress again on the ATC today. I am simultaneously making a new spindle from A6 tool steel, (which will be fully hardened and ground) and ditching the bellevilles. I expect there will be a lot of interest in my big fat die spring because it *probably* can be implemented on an existing spindle which has no way to get the typical disc spring arrangement installed. Next couple days I will post some photos of the ATC assembled on the side of the new head.
I see from the photos you posted that the swing arm has been eating its Wheaties. Very nice piece! 8) I also note the pivot has an upper and lower mounting. Double nice. Significant improvements from the prototype, as one would expect. Q: curiosity, how did you form the smooth curve in the skirt?
The pivot arm bracket is actually the prototype. The top support for the pivot was always planned, it just took a while to get it built, and made it much easier to remove the whole assembly from the machine during initial tinkering and debugging. The "skirt" was just rough hand-formed over a piece of 4" diameter aluminum round, then fastened in place with 6-32 screws. I'm very pleased with how nicely it came out. I now have the door operating as well. I used a simple tension spring to pull the door open, and a cable from the door to the mounting arm on the ram to close. The geometry is such that with just a short (18") piece of bicycle brake cable inner wire fastened between those two points, it pulls tight when the arm parks, and goes slack when the arm pivots to the quill, allowing the spring to pull the door open. Could'nt be any simpler or cheaper.
Now that my annoying spindle problems appear to be resolved, I can finally get back to work! Looking forward to seeing what you come up with. Your work always leaves me feeling totally inadequate! :-)
Regards,
Ray L.
-
Here is one final video, making cuts and doing toolchanges.
http://www.youtube.com/watch?v=zjGi1WBTMBE&feature=youtu.be
Regards,
Ray L.
-
Nice Ray!
But with the cover on you cant see the magic!
-
Some progress to report.
Ding, dong, the wicked Bellevilles are dead and gone. Long live the spring!
Only after scrubbing BOTH sides of each belleville by hand to create flat mating surfaces would the springs stay aligned when compressed. However, that solution created another problem by reducing the travel to barely what was needed to release the gripper. It still worked, but it would not take much variation in a stud to have a tool hang up in the spindle due to insufficient travel to fully release. Worse, even after all of that effort, the disk springs still chafed on the shaft as seen in the following photo. Using disk springs specifically made for a machine spindle or 'nesting' much thinner springs to get up to the required force might have solved these issues. Nitriding the drawbar and tube would certainly help, but my patience ran out for the whole spring stack arrangement. There may be some compelling reason to use this method in certain cases, for example, in-line nesting of coil springs increases travel, but not spring force. However, for the BT30 spec, I see no reason not to replace a troublesome $60 stack of problems with a $19 die spring and not have to deal with the the spring stack issues at all.
(http://www.theCUBEstudio.com/BT30SpindleProject/ChafingWEB.jpg)
Jumping ahead to the completed components:
(http://www.theCUBEstudio.com/BT30SpindleProject/DrawbarComponentsSpringTypeWEB.jpg)
Switching over to the coil spring was relatively simple matter of cutting a pocket for the spring and making retainers and guides. The angle in the next photo is intended to show the bottom of the pocket and it makes the snap ring grove look a bit odd, but it is a simple square bottom grove. One of the unsatisfying aspects of the disc spring stack was that with no tool in place, the drawbar would extend all the way up and press against the actuator making the spindle difficult to turn (for belt changes, etc). If the spindle was started in this condition, it would damage the actuator plunger and the top of the drawbar.
In the new arrangement, a snap ring at the top of the spindle makes a positive stop for the spring when there is no tool holder in the spindle. The snap ring is on order so it is not pictured.
(http://www.theCUBEstudio.com/BT30SpindleProject/SpringPocketWEB.jpg)
And here is the spring installed on the drawbar and partially inserted into the spindle. Most likely, this arrangement could be added to an existing spindle with the spring on top of the spindle. This particular spring was selected for BT30 and the rate is 2550lbs. At coil bind you have approx 2,000lbs and for a gripper with a .21" release travel, you get a max hold of about 1,450lbs. Note that this in not adequate for R8, but larger springs are available that for that application. I have the arrangement set up so that any hold tension from 600lbs to the 1,450 max is available by simply adjusting the threaded retainer cap.
(http://www.theCUBEstudio.com/BT30SpindleProject/SpringInPocketWEB.jpg)
-
Steve,
Nice work as always. The time I spent playing with Belleville drawbars a few years ago convinced me I never wanted to mess with them again, and is what led to my stepper-driven drawbar. I wonder why so many VMCs use them? And huge stacks of them at that.
BTW - I sent you a e-mail earlier today re: KFlop. Did you get it?
Regards,
Ray L.
-
This post is just some misc stuff that has been completed for the ATC.
Following photos are of the completed carousel mounting arm. In the next few days, I hope to have this and the swing arm mounted to the side plate of the new mill head.
(http://www.theCUBEstudio.com/BT30SpindleProject/CarouselMountingArmBottomWEB.jpg)
A previously missing detail is the safety interlock for the drawbar actuator. This little guy sports a hardened pin that runs thru a hardened die pin guide bushing and interferes with the actuator movement unless it is retracted. i.e. the drawbar actuator cannot release the tool unless this pin is first retracted. Totalling all of the special parts and the time to design and build this component adds up to a fairly expensive little gadget. However, safety features are immune from cost cutting and/or production considerations.
(http://www.theCUBEstudio.com/BT30SpindleProject/ActuatorSafetyInterlocWEB.jpg)
-
Steve,
Nice work as always. The time I spent playing with Belleville drawbars a few years ago convinced me I never wanted to mess with them again, and is what led to my stepper-driven drawbar. I wonder why so many VMCs use them? And huge stacks of them at that.
I hear numbers like 3k to 5k forces for CAT50. I hasten to say that this is just what I read on forums. I do not know the facts on this. However, if these numbers are accurate, then it would be difficult to achieve with a coil spring. Disc springs can be nested thereby multiplying the force.
BTW - I sent you a e-mail earlier today re: KFlop. Did you get it?
Regards,
Ray L.
Yes, I did. I'm just now getting to my computer chores; e-mail, reading, posting, etc. I'll have a response for you on that e-mail and also your postings here after I have a minute to noodle over them.
-
Actually, the sensors and code on mine turned out to be quite trivial. The code is barely two pages of dead-simple C that took only maybe two hours to write and debug.
This is a surprise. I have a flow chart that runs for a couple of pages, let alone the coding. Even considering that what is 'trivial' to you is anything but trivial to most people, it would truly be amazing (to me) to be able to control an ATC with two pages of code.
This is the list of tasks that I envision need to be performed on my ATC. I broke it down two ways; down and dirty and commercial/industrial (OSHA). All comments from any reader are welcome.
Simplistic: swing the arm 30 degrees
drop 4"
say a prayer
swing the arm another 60degrees
rise 4"
say another prayer
Safe:
Has the carousel reached the correct tool position?
Is there actually a tool in the holder?
( we already know the claw is unoccupied because we confirmed that when it put the previous tool away)
Open the safety interlock on the claw.
Did it open?
Swing the arm 30 degrees
Did the claw close? (if it closed, then it has the tool, otherwise it cannot close)
Is the tool holder safety interlock open?
Release the tool from the toolholder (pneumatic)
Is the tool released?
engage the claw safety interlock
did the claw safety interlock engage?
lower the arm 4"
did the arm lower?
Is the mill spindle rotating?
Is the mill spindle at the correct azimuth for tool change?
lockout the mill spindle operation (can be done a number of ways)
Swing the arm 60 degrees.
Is the arm in position?
Release the drawbar safety interlock
Did the drawbar safety interlock release?
Activate the drawbar actuator
Did the drawbar actuator open?
Raise the arm 4"
Did the arm reach position?
release the drawbar actuator
Did the drawbar actuator release?
Is the drawbar safety interlock reset?
Is the drawbar height correct (i.e does it have the tool gripped correctly)
release the claw safety interlock
did the interlock release?
Swing the are back 40 degrees
has the arm reached position?
lower the arm 5"
did the arm lower?
move the arm to the park position.
allow mill spindle operation
-
Steve
The second list looks better to me :)
I try and interlock things as much as possible on my machines. I am the only person that uses them so its more for machine safety than personal safety that I take that route. The big lathe could rip itself to bits in seconds if something wasnt right and I didnt have a saftety check in to see it wasnt right.
Mills are not so bad in that they wont likely do as much damage to themselves but bent arms or carousels or whatever are still a real possibility so I think your approach is the way to go.
I like doing these things in a PLC, main reason is I find ladder logic, well how can I say it, logical ;D and it comes easy to me. The Chirons ladder if printed out at 100% would take 46 pages of A4 paper, I wouldnt know where to start trying to do all that in VB let alone any other programming language.
Hood
-
Looking good Steve.
I'm not having any of the issues that you are with the Bellevilles but I suspect that it's because mine are external. The coil spring idea In my opinion is the best way to go. As soon as I get the last bits of my tool changer done I'm going to be redoing the drawbar with the coil spring. I see you used a die spring instead of the valve spring.
I have a question for you about Pull stud tools. Is the gripper supposed to provide lateral support as well as upward pressure or is the taper supposed to take care of that?
Derek
-
Looking good Steve.
I'm not having any of the issues that you are with the Bellevilles but I suspect that it's because mine are external. The coil spring idea In my opinion is the best way to go. As soon as I get the last bits of my tool changer done I'm going to be redoing the drawbar with the coil spring. I see you used a die spring instead of the valve spring.
Pretty much all of my problems with the disc springs was binding and galling in the containment. First they were chewing on the ID of the tube so I made a fatter shaft to keep them centered and off the tube ID and they then chafed on the shaft. I solved that (mostly) by tediously sanding flats on the 'mating' surfaces, but in so doing, I lost a good deal of the already barely adequate travel. Only a few thou off each spring, but there were 60 of them so it adds up to a significant number. The experience was expensive, but useful. I would know what to do differently if I ever have to go 'back to well' in the future. However at this point in time, I'd rather be poked in the eye with a sharp stick than touch another Belleville stack :'(
If you are using a 30 taper I will give you the part number and source for the die spring (in a PM). I saw the top of your spindle and it looks like it will fit in there quite nicely. If you are using any other taper, including R8, then you will need to select your own spring. The die spring was considerably cheaper and is a single spring which simplifies things. The damping of the multi spring valve spring is not needed in this application. FWIW; I made the washers above and below the spring fro 416 SS. The top washer is .26" thick (not counting the slight relief at the top to clear the snap ring) but it is backed up by a wide retaining nut that slightly overlaps the spring so the washer does not take the stress in a cantilever. The ball holder is undoubtedly the weak link. I have tested mine to 3,500lbs tension so I have the minimum 2X safety factor.
I have a question for you about Pull stud tools. Is the gripper supposed to provide lateral support as well as upward pressure or is the taper supposed to take care of that?
Derek
I don't know the answer to that, but I would speculate that the taper is going to dominate over any misalignment in the gripper. The consequence of a misaligned (or low precision) gripper is that not all of the balls will be sharing the load. For example, if the holes for two of the balls is even .003" above the other two, you will have only two balls holding all of the force (unless you make the gripper body from 'stretchy' material . . which would be a really bad idea).
Typically I only accomplish high precision or high finishes where they are necessary, Rough finishes and 'close enough' where it doesn't;t matter and 'normal' everywhere else. The gripper body is one of those pieces that requires precision, particularly in concentricity and the elevation and depth of the ball holes (the depth matters because the round shape effectively changes the height). The outer sleeve has to be reasonably concentric, but is not as critical, in my view, as the ball holder. The Ball holder is tricky because you can't 'nibble' at tool steel of it will harden up on you so you have to take your best shot with a good setup and sharp tools and measure the result.
-
Steve,
Here is my ATC code:
#ifndef ATC_H
#define ATC_H
#include "KMotionDef.h"
#include "../Common/CommonDefs.h"
#include "../Common/PDB.h"
int current_atc_pos; // Current ATC position
int current_atc_tool; // Current ATC tool
void LoadTool(int tool)
{
current_atc_tool = (int)persist.UserData[PERSIST_CURRENT_TOOL];
if (tool == current_atc_tool)
return;
RotateToTool(current_atc_tool);
Move(AXIS_Z, PDB_Z_POS * Z_STEPS_PER_INCH);
while (!CheckDone(AXIS_Z))
;
if (current_atc_tool == 1)
LowerLift();
else
RaiseLift();
PivotToQuill();
if (current_atc_tool != 1) {
PDBReleaseTTS();
LowerLift();
}
RotateToTool(tool);
if (tool > 1) {
RaiseLift();
PDBGrabTTS();
}
PivotToPark();
RaiseLift();
}
void RotateToTool(int tool)
{
int target_tool = tool;
int current_tool = (int)persist.UserData[PERSIST_CURRENT_ATC_POS];
if (tool == 1)
target_tool = 2;
if (current_tool == 1)
current_tool = 2;
while (current_tool != target_tool)
{
SetBit(ATC_ROTATE);
Delay_sec(0.5);
ClearBit(ATC_ROTATE);
Delay_sec(0.5);
++current_tool;
if (current_tool > 11)
current_tool = 2;
}
Delay_sec(0.5);
current_atc_pos = tool;
persist.UserData[PERSIST_CURRENT_ATC_POS] = (double)tool;
}
void RaiseLift()
{
if (!ReadBit(ATC_LIFT))
return;
ClearBit(ATC_LIFT);
ConfirmAction(ATC_LIFT_SENSE);
}
void LowerLift()
{
if (ReadBit(ATC_LIFT))
return;
SetBit(ATC_LIFT);
ConfirmAction(ATC_LIFT_SENSE);
}
void PivotToQuill()
{
if (ReadBit(ATC_PIVOT))
return;
SetBit(ATC_PIVOT);
ConfirmAction(ATC_PIVOT_SENSE);
}
void PivotToPark()
{
if (!ReadBit(ATC_PIVOT))
return;
ClearBit(ATC_PIVOT);
ConfirmAction(ATC_PIVOT_SENSE);
}
ConfirmAction(int bit)
{
int i = 0;
Delay_sec(0.25);
while (i < 10) {
if (ReadBit(bit))
i++;
else
i=0;
}
i=0;
while (i < 10) {
if (!ReadBit(bit))
i++;
else
i=0;
Delay_sec(0.01);
}
}
#endif
Regards,
Ray L.
-
Steve
The second list looks better to me :)
I try and interlock things as much as possible on my machines. I am the only person that uses them so its more for machine safety than personal safety that I take that route. The big lathe could rip itself to bits in seconds if something wasnt right and I didnt have a saftety check in to see it wasnt right.
Mills are not so bad in that they wont likely do as much damage to themselves but bent arms or carousels or whatever are still a real possibility so I think your approach is the way to go.
I have gotten a lot of positive responses on the spindle and the ATC including a couple of dealers/manuf who are interested in incorporating them into their machines (along with the InTurn™ 4th axis). I am seeking product liability insurance at this time with the intent to sell the spindles. I am not able to build a proper 10k + RPM spindle just yet, but that is coming. My immediate concern is the same as yours; primarily torn up equipment and down time, but long term, the serious consequence would be a tool holder coming out of the spindle at 15K RPM and taking a brisk stroll across the shop floor. Mach is not exactly a stable program and PC's in general are prone to tantrums now and then. If the power goes off momentarily, it is hard to predict what a PC is going to do. This happened recently here and air valves were going on and off as the PC tried to get its act together. That's why I have an electric solenoid operated positive mechanical interlock on the drawbar actuator. There is no way to accidentally release a tool. However, an equally important task is to make sure the tool is gripped and seated properly before starting the trip to 15k.
Those are the biggies. The rest is, as you pointed out, to keep the machine from chewing its own ankles off.
I like doing these things in a PLC, main reason is I find ladder logic, well how can I say it, logical ;D and it comes easy to me. The Chirons ladder if printed out at 100% would take 46 pages of A4 paper, I wouldnt know where to start trying to do all that in VB let alone any other programming language.
Hood
Probably a good bit of that code is for communicating with the mother ship and I won't have that feature . . . :D
The pair of Mistu drives will help a lot in reducing the coding and the sensor count as they will fault in some of the situations and they have a position reached signal that will serve as the sensor for several of the steps. In taking advantage of that, the controller would not be able to pinpoint a failure quite as accurately, but the operator can snoop around a bit to figure out what stopped the party.
-
Those are the biggies. The rest is, as you pointed out, to keep the machine from chewing its own ankles off.
I would try to design in a failure point to prevent that. On my ATC, the weak point is the attachment of the tool "forks" to the carousel, which are Nylon screws. If a major jam were to occur, those screws would simply shear off, and the tool holder, and folk would fall on the table. The only failure I can see right now with any possibility of breaking anything would be if the PDB tried to engage while the spindle was running in reverse. Since I've never once had to run in reverse, this one doesn't concern me. If the PDB were to engage while the spindle is running forward, the drawbar would simply unscrew, dropping the tool and collet. Not good, but not catastrophic. Even then, the spindle brake *might* be enough to over-load the motor and stall the spindle, though I have no intention of testing that theory.... Once I get a spindle sensor on the machine, these issues also go away, as neither the PDB
nor ATC will be allowed to engage unless the spindle is confirmed stopped.
Regards,
Ray L.
-
Steve,
Here is my ATC code:
#ifndef ATC_H
#define ATC_H
#include "KMotionDef.h"
#include "../Common/CommonDefs.h"
#include "../Common/PDB.h"
Ray L.
Well this explains things a bit. We have counting different beans, methinks.
What you have here is the conductor and the orchestra is elsewhere. You see I would count the libraries as part of the ATC code. Or certainly at least the ones you wrote yourself. I like looking at your code as there is unsally a tasty morsel or two of the 'tips and tricks' variety, but really I was interested in seeing what actual pieces you used for sensing and how those are mounted and while I can see some checks in your code, the routines are not present so it is hard to divine what sensors you are reading and what reaction is called in response to a fail.
For example, I see ATC_PIVOT_SENSE and ATC_LIFT_SENSE but that doesn't explain what you are doing specifically.
You usually have some clever trick up your sleeve and it take some coaxing to get it out of you . . example; the relay switched torque on your drawbar stepper drive. You had to have something slick going on in order for the thing to work the way you described, because there drive does not have that capability . . until you figured out how to add it.
So if you feel so inclined, I for one would love to have some details on exactly what actions you are monitoring and most importantly why you chose those actions and not others and how you accomplish the tasks. Really the whole methodology and thought process is what I'm after if you are willing to share that.
-
Steve,
BTW - I'm also in discussions with a machine manufacturer, to potentially design a PDB/ATC system for several of their machines, loosely based on my current design. There are, of course, a number of things I would do differently on the second go-round.
Regards,
Ray L.
-
Those are the biggies. The rest is, as you pointed out, to keep the machine from chewing its own ankles off.
I would try to design in a failure point to prevent that. On my ATC, the weak point is the attachment of the tool "forks" to the carousel, which are Nylon screws. If a major jam were to occur, those screws would simply shear off, and the tool holder, and folk would fall on the table. The only failure I can see right now with any possibility of breaking anything would be if the PDB tried to engage while the spindle was running in reverse. Since I've never once had to run in reverse, this one doesn't concern me.
Running in reverse does present some additional problems. I have a big face mill that runs in reverse, hard tapping of course involves reversing the spindle and I also run certain grinding cups in reverse depending on the setup.
If the PDB were to engage while the spindle is running forward, the drawbar would simply unscrew, dropping the tool and collet. Not good, but not catastrophic. Even then, the spindle brake *might* be enough to over-load the motor and stall the spindle, though I have no intention of testing that theory....
You have the luxury of designing and building for your own use exclusively. I have to imagine the stupidest thing a person could possibly do and then try to prevent it. Her in the States, everything you buy has twelve pages of warnings and one page of actual operating instructions. First you have to say 'do not use this toaster in the shower' and 'do not eat the insulation' and 'do not put the large hot coffee between you legs and then drive out into traffic'. Each one of those idiotic warnings is part of the settlement from a law suit. It is hard to explain to people on the other side of the pond how rediculous out legal system is here.
Once I get a spindle sensor on the machine, these issues also go away, as neither the PDB
nor ATC will be allowed to engage unless the spindle is confirmed stopped.
Regards,
Ray L.
A HAH! You are soooo busted my friend! You have more sensors to add. I'll just puit that beer back on the shelf a bit longer, thank you very much! >:D
-
You have the luxury of designing and building for your own use exclusively. I have to imagine the stupidest thing a person could possibly do and then try to prevent it. Her in the States, everything you buy has twelve pages of warnings and one page of actual operating instructions. First you have to say 'do not use this toaster in the shower' and 'do not eat the insulation' and 'do not put the large hot coffee between you legs and then drive out into traffic'. It is hard to explain to peopel on the other side of the pond how rediculous out legal system is here.
Steve,
I'm on your side of the pond - Santa Cruz, CA. The best warning I ever saw was for a Toshiba laptop. They warned you not to drive nails through the battery pack! The drawing accompanying the warning was priceless! It was someone pounding what looked like a railroad spike through a battery pack.
BTW - *I* don't have to imagine "the stupidest thing a person could possibly do". I'll usually manage to do it myself at some point.... :-)
Regards,
Ray L.
-
Steve,
BTW - I'm also in discussions with a machine manufacturer, to potentially design a PDB/ATC system for several of their machines, loosely based on my current design. There are, of course, a number of things I would do differently on the second go-round.
Regards,
Ray L.
We should throw in together. I'll do the mechanics and you can do the controls. I don't like that part so much . . .
-
Probably a good bit of that code is for communicating with the mother ship and I won't have that feature . . . :D
Nah the mother ship only sends a signal out and waits for one back to know its done :)
The pair of Mistu drives will help a lot in reducing the coding and the sensor count as they will fault in some of the situations and they have a position reached signal that will serve as the sensor for several of the steps. In taking advantage of that, the controller would not be able to pinpoint a failure quite as accurately, but the operator can snoop around a bit to figure out what stopped the party.
Yes the lathes turret is kind of like that, only have two sensors on it, one to say its clamped and another for un-clamped, the drive handles the rest and will fault out if it doesnt get there which in turn will fault out Mach. There is also a feature in the drive where it can send an analogue voltage to indicate which position the encoder is at but I decided I didnt need that.
The Chiron just has so much I/O and that is why the ladder is so long, there is a prox on the clamp, two on the main up/down cylinders and 2 on each of the tool cylinders. The ladder makes it easy to make sure one step can not be started before the other is complete.
Those are the biggies. The rest is, as you pointed out, to keep the machine from chewing its own ankles off.
I would try to design in a failure point to prevent that.
Sad thing is its not so simple with a lathe as the turret needs to be sturdy to stand up to the cutting forces and that is why it can do so much damage. A 10inch chuck spinning at 2000rpm colliding with the turret will not have a weak point other than a spindle getting bent or a turret getting ripped off or possibly both :(
Hood
-
wait a minute, I did NOT know there was Beer involved in this!!
Here is an ATC prototype I did for Tormach a while back.
http://www.youtube.com/watch?v=vcWUZJ35yL0&feature=youtu.be
Scott
-
wait a minute, I did NOT know there was Beer involved in this!!
Here is an ATC prototype I did for Tormach a while back.
http://www.youtube.com/watch?v=vcWUZJ35yL0&feature=youtu.be
Scott
Interesting. If I read between the lines in your video description, things did not go well with Tormach on this project.
I have it in writing from Tormach's engineering manager that their MACH software is 'locked down' because Tormach does not believe users should be allowed access the 'flakey' MACH software in their own systems for fear they will mess up the 'motor parameters'.
As you know, in MACH parlance, 'locked down' versions have nothing to do with keeping users out and there are no motor parameters in MACH. These statements indicate to me an unfamiliarity with MACH which lead me to speculate that Tormach uses outside sources for some or all of their development. You have now confirmed that.
-
Ray I believe Tormach's stance is to LOCK OUT access to the motor tuning parameters Steps Per, Vel and accel. From support stance I cannot find fault with that.
Also they do give you the option of unlocking it for you BUT they will not support you after that point. Can't argue with that either.
Scott, you were NOT in process when the bet started, DING DING, DQed (;-)
THere ARE others that are in process but are bashfull.
(;-) TP
-
Steve,
BTW - I'm also in discussions with a machine manufacturer, to potentially design a PDB/ATC system for several of their machines, loosely based on my current design. There are, of course, a number of things I would do differently on the second go-round.
Regards,
Ray L.
I am supposed to be semiretired, but most of my time is still consumed by consulting work, mostly under DNA and in such arrangements, by contract, liability transfers to the client. Selling a product that I manufacture directly to an end user is a whole other animal.
If I am able to get the proper insurance, I will make the spindle and ATC available to end users, otherwise only to manuf who will assume liability under their umbrella.
I am aware that you are on the same side of the pond, that's why I figured you can appreciate the paranoia. The insurance industry is as nuts as the legal system. A well know insurance company gave us a TD because . . now get this . . they decided that the 4th axis constitutes 'automation' and they do not underwrite 'automation' because 'automation' *might* be used to make guns or gun parts. WTF ??? Gotta be lawyer prints all over that one.
-
Ray I believe Tormach's stance is to LOCK OUT access to the motor tuning parameters Steps Per, Vel and accel. From support stance I cannot find fault with that.
Also they do give you the option of unlocking it for you BUT they will not support you after that point. Can't argue with that either.
I think you were addressing my post, so I will respond. Incidentally, you forgot to mention that warranty is also effected if an owner signs Tormach's waiver to get access to their MACH software. Do you find fault with that I wonder.
You can believe what you want. As I said, I have Tormach's official response in writing and I have quoted directly from that document. If you want to see the actual e-mail, just ask and I will be happy to post it so that 'interpretation' is not needed.
I don't care to get into a Tormach debate, but I do want to clarify one thing. In my inquiry, did not suggest Tormach change their policy (regardless of how condescending *most* people think it is). I offered to spend MY time reconfiguring MY software to eliminate any possible conflict with their setup, and that no Tormach customer needed to be involved nor inconvenienced nor given access to anything. The process would be one engineer (me) consulting with an other engineer (Tormach's) to provide QC for the benefit of Tormach owners.
Their response (again in writing) was that they are " . . . not excited about helping third parties, particularly those [who's products] compete with Tormach's". They inked me to their 'Duality Lathe' as if that was in any way competitive with or equivalent to the InTurn™. My current controller also accommodates a servo driven spindle with universal hard tapping macros included. Tormach has no equivalent that I am aware of.
The simple solution is to ditch the entire Tormach setup and do a fresh install of MACH and put the InTurn™ stuff over it. Tormach only succeeded in aggravating their own users . . and me. I am asked a couple of time a week what machine I reccommend for use with the InTurn™ 4th axis. Venture a guess?
-
Sorry but I support Tormach's policy. It would be the same if I bought a new car then took it apart to change engine parts to make it more sporty and more HP.
Of course you would not because you want to sell the Owners something. (;-)
AGAIN we should just agree to disagree, (;-) TP
-
Interesting. If I read between the lines in your video description, things did not go well with Tormach on this project.
I have it in writing from Tormach's engineering manager that their MACH software is 'locked down' because Tormach does not believe users should be allowed access the 'flakey' MACH software in their own systems for fear they will mess up the 'motor parameters'.
As you know, in MACH parlance, 'locked down' versions have nothing to do with keeping users out and there are no motor parameters in MACH. These statements indicate to me an unfamiliarity with MACH which lead me to speculate that Tormach uses outside sources for some or all of their development. You have now confirmed that.
You are correct, they took many of my findings, and ideas directly from the CNC Zone thread when I released the ongoing project, then said, "Oh, look what our engineers came up with.." What a crock of Poop. You will also find that they have a "Flash Screen", that Jay did, and guess what no real mention of him as the guy behind it.
I just find it funny that my original prototype, video, was removed somehow from Google Video, and all the links to that video, on even the thread on this site, was renamed, and remapped to their ATC videos..... That would require someone at moderator/admistrator of above (or a hacker), to modify mine and other post.
Noticed some similar tampering with post on the CNC Zone as well.......... Just saying....
Also, just an FYI, there is an ".exe" in the Tormach profile that when you run it, it brings up the standard Mach3 profile, all you need to do on an install if you want to "Lock Out"
users from motor profiles etc. is to use a Hex editor and remove the menu access objects. Not hard at all.
Like you Ray, I don't appreciate how they treat folks like you, and me for that matter.
Having said that, I don't blame them for locking out their profiles/screens from curious buyers due to support issues, but on the other other hand, I don't think
you or anyone else should be talked down to either. There where many, many witnesses to my building my prototype at the CNC zone, along with my ideas for improvements and ect.
Here is the REAL kicker, I was asked to keep the Production ATC under 3K they said no one would pay more than that, so I cut out alot of bells/whistles to keep labor/cost down.
Then Blammo, they come out with there 6K ATC (including the Draw bar device, since that is also part of the complete package)...... Funny how that works isn't it.
Very similar deal happened when I designed the CNC Keys project for another manufacture, not a single mention on their website either, and THAT is why I stopped development on that project. Well enough of my griping, the summery of all of this is this, make sure when your dealing with these companies that you get the ENTIRE deal in writing UPFRONT, any changes or "other directions" also get in writing. My hope is you won't have to go down this road like I have.
Didn't mean to hijack this thread, so Ray, back to your thread...
Terry, I know your a Tormach fan, and that is cool with me, everyone has different experiences, no offense to you or other Tormach lovers on here.
Scott
-
ah, found the topic on the CNC zone, be interesting to see how long it stays there without disappearing as well, since that forum is Tormach sponsored..
http://www.cnczone.com/forums/tormach_pcnc/64889-atc_kit_tomach_about_released.html
Back in 2008.........
Scott
-
Sorry but not a Tormach lover OR hater, I just understand the reason behind their decision to lockout the user from Config.
As to the Company? NEVER have I dealt with them so I could not say one way or the other.
NOW from other dealings and designing machines and gizmos. NEVER work without specific contracts and NEVER put ANYTHING on the WEB. In most cases you LOOSE all rights the moment you do so.
Been there done that as well, KNOW better now, (;-) TP
-
Yeap, what he said............
-
Tormach has made the (perfectly reasonable) decision to get their machines working as well as they can with a single, now very old, version of Mach3. That way, they don't have to deal with the new bugs that get introduced with each new release. They run a slightly modified version of Mach3 that has user access to many critical settings, like the motor tuning parameters, disabled so they don't have to support customers who go in and mess with the settings. They will happily send you the information needed to setup an off-the-shelf copy of Mach3, but they, again perfectly reasonably, will NOT support you once you've done this. Support costs a LOT of money, so I think their position is very reasonable.
Regards,
Ray L.
-
Like you Ray, I don't appreciate how they treat folks like you, and me for that matter.
Having said that, I don't blame them for locking out their profiles/screens from curious buyers due to support issues, but on the other other hand, I don't think
you or anyone else should be talked down to either. There where many, many witnesses to my building my prototype at the CNC zone, along with my ideas for improvements and ect.
Here is the REAL kicker, I was asked to keep the Production ATC under 3K they said no one would pay more than that, so I cut out alot of bells/whistles to keep labor/cost down.
Then Blammo, they come out with there 6K ATC (including the Draw bar device, since that is also part of the complete package)...... Funny how that works isn't it.
Sorry to hear about your experience with Tormach. I remember reading parts of that thread ages ago, but didn't realize the whole story. Just for the record, I've never had any dealing of any kind with Tormach, other than buying some of their TTS tools.
Didn't mean to hijack this thread, so Ray, back to your thread...
That's funny, because it's not *my* thread, it's Steves, and, if anything, I hi-jacked it from him! :-) Now we're having a "who can build the bet ATC the quickest" contest. I think I've won the "quickest", but he will clearly win the "best" with his typically outstanding work. Mine is more from the "quick and dirty, but gets the job done" school.
Regards,
Ray L.
-
well you have BOTH done excellant work!!! I really like the way both of you designed things!
Scott
-
Scott,
Thanks!
I just got done reviewing the whole of your ATC thread. You had some serious trolling going on in there! Yikes!
Sorry to see how it all turned out. I'm curious, if you don't mind my asking - Were you successful in selling your design?
Regards,
Ray L.
-
Yes, sold it to an "Indian" company (the country), they just bought it straight out.
Then they recontacted me about re-doing the programming/screens since they added it to one
of their OEM machines. I had to do the PLC program on a Mitsubishi series PLC, since they could
not get Koyo (ADC), as easy or cheaply. I added a swap arm routine to it as well.
Yes, the Trolling and Shrills where unbelievable there. I did learn a very valuable lesson to never, ever
again, try to show a project on a forum like CNC Zone.
Scott
-
I will catch up on this thread when I have more time. For now, here is a short video of the first test of the partially assembled ATC on the steel plate that will be the side of the head for the new mill.
Couple notes: I only have one Carousel 'pod' so not much point in spinning it around, but there is a motor fitted and the carousel rotates. There is a lot of play in the carousel partially due to the missing motor, but most or all of the play will be removed during final assembly.
I have removed the individual safety interlocks from each pod in favor of a single interlock on the air cylinder that releases the pod.
The sliding shaft is covered by a rubber bellows . . not installed for testing.
The mechanics are running a little over half speed for this first test.
The coil spring replacement for the belleville stack is completed, but I have to reassemble the spindle, test the actuator and then can add it to this mock-up. Testing with the spindle in place is going to require more air valves and relays to drive them so there are some do-dads that will need to be acquired to test all of the pieces together.
http://www.youtube.com/watch?v=DFEfXIAlqCE&feature=youtu.be (http://www.youtube.com/watch?v=DFEfXIAlqCE&feature=youtu.be)
-
Steve,
Smooooooooth as a new baby's you-know-what. Nice! Can't wait to see it running full-speed!
Regards,
Ray L.
-
That is very nice indeed :)
Hood
-
Wow that is super slick. Nice work!
-
Nice motion,
(;-) TP
-
Steve, again a very nice piece of work.
Thanks for sharing your ideas, which inspire me.
-
That is very nice indeed :)
Hood
Hood, can you tell me if there is a low power (say 100 to 200 watt) version of the AB drive you are using for your turret that might have identical features to those that you are taking advantage of?
I am using my spare 400 watt Mitsu J3 for the arm and I purchased a 200 watt Mitsu J2S for the turret. The Mitsu does have programable motion, but it is incredibly convoluted process to set up. From what you describe, the AB is very straightforward.
-
The smallest is 500watt, 2.5 Amps cont 5 Amps peak.
The models to look for are
2098- DSD-005X
2098- DSD-005-SE
2098- DSD-005X-DN
The first is the Indexing drive, second is sercos but if you disable that it turns it into an Indexing drive, the last is the Indexing Device Net version.
There is also
2098- DSD-005
2098- DSD-005-DN
Neither of these support Indexing.
Hood
-
Sorry but I support Tormach's policy. It would be the same if I bought a new car then took it apart to change engine parts to make it more sporty and more HP.
Of course you would not because you want to sell the Owners something. (;-)
AGAIN we should just agree to disagree, (;-) TP
You analogy is far off the mark, I'm afraid. There is certainly no shortage of disagreements between us, but in this case it would be imaginary. We would be disagreeing on whether the number 2 or the color red is the best flavor. It is obvious that you missed the point entirely.
It matters zero to me what your opinion of Tormach's policies are, but it does matter that you present your speculations about me as if they were facts. You may have noted that when I am speculating, I say so plainly. I think that would be a good policy for you to support.
-
The smallest is 500watt, 2.5 Amps cont 5 Amps peak.
The models to look for are
2098- DSD-005X
2098- DSD-005-SE
2098- DSD-005X-DN
The first is the Indexing drive, second is sercos but if you disable that it turns it into an Indexing drive, the last is the Indexing Device Net version.
There is also
2098- DSD-005
2098- DSD-005-DN
Neither of these support Indexing.
Hood
Thanks! I'll check them out.
500 watt is major overkill for my little carousel, but if the price is right, it would probably save enough time to be justified.
-
I think your own words sum it up nicely. You were disappointed that Tormach would not work with YOU to develope 3rd party hardware to run with a standard TORMACH. (;-)
I don't care to get into a Tormach debate, but I do want to clarify one thing. In my inquiry, did not suggest Tormach change their policy (regardless of how condescending *most* people think it is). I offered to spend MY time reconfiguring MY software to eliminate any possible conflict with their setup, and that no Tormach customer needed to be involved nor inconvenienced nor given access to anything. The process would be one engineer (me) consulting with an other engineer (Tormach's) to provide QC for the benefit of Tormach owners.
Their response (again in writing) was that they are " . . . not excited about helping third parties, particularly those [who's products] compete with Tormach's". They inked me to their 'Duality Lathe' as if that was in any way competitive with or equivalent to the InTurn™. My current controller also accommodates a servo driven spindle with universal hard tapping macros included. Tormach has no equivalent that I am aware of.
The simple solution is to ditch the entire Tormach setup and do a fresh install of MACH and put the InTurn™ stuff over it. Tormach only succeeded in aggravating their own users . . and me. I am asked a couple of time a week what machine I reccommend for use with the InTurn™ 4th axis. Venture a guess?
(;-) TP
-
Thanks! I'll check them out.
500 watt is major overkill for my little carousel, but if the price is right, it would probably save enough time to be justified.
I think the list in the UK for a 005-SE is approx £1000 plus VAT, obviously list and what you actually pay is quite a bit different but it gives you an idea of tha ballpark figures.
It often makes me chuckle when I see them listed on eBay second hand, often they are more than you could get one new and paying full list ;D
Hood
-
I think your own words sum it up nicely. You were disappointed that Tormach would not work with YOU to develope 3rd party hardware to run with a standard TORMACH. (;-)
You still don't get it, and you still post your personal speculations about me as if they are facts, so let me help you out with a few ACTUAL facts.
1) The 'development' was already completed long ago, so your latest statement above is also . . false. Tormach is decidedly low-tech. Their 'engineering white paper' is actually just a list of excuses as to why steppers are 'better' than servos. 'Disappointed', you say? I think they would need to go back to school for a while before they could help me with development, frankly, but 'working with ME to develop' anything (as you claim) was definitely not on the menu.
2) The objective was only to QC my MACH customizations with theirs. With the single exception of Tormach, every other vendor has been more than happy to cooperate. If you have ever done this, you would know that it is a simple matter of making sure multiple apps don't use the same outputs and other MACH resources and then merging the screens. This in not rocket science, but is apparently beyond Tormach's abilities. I will SPECULATE that Tormach is not so much unwilling as they are unable to cooperate because their stuff is third party and they don't understand it themselves. Using outdated version of software in order to 'avoid new bugs' is an excuse, in my view. They either do not have the talent in house or they do not want to pay for third party review of bug FIXES.
3) You made the accusation that my motivation was to 'sell something to the Owners'. Way off base once again. Several previous InTurn™ models have gone to Tormach owners. Most of these users already had clean MACH installs and had ditched Tormach's stuff. One of them took on the task of integration himself. Two of the new MEGA-Duty models from the last batch went to Tormach owners. One of them has a clean MACH install and the other was looking for a solution. These were done deals and I did not HAVE to do anything at all to 'sell something to the Owners'. My effort was after the fact, by customer request, and would have been an uncompensated support function.
Now let me explain 'Red Neck Logic'; A red neck farts. Just then it begins to rain. Conclusion: farting causes rain. Proof: 'seen it mahseff'. Once the redneck has this 'fact' lodged in his brain cel, there it no way to change it.
Things are not always what they appear. I remain uninterested in your opinion about Tormach and I am still not going to get into a debate with anyone about their policies, but I do ask that you collect some facts and maybe ask a question or two before you shoot your mouth off about my motivations, intentions, or any other part of my character.
This thread is not about Tormach, or Tormach fanbois. If you have nothing to contribute to the topic, please lurk quietly.
-
Thanks! I'll check them out.
500 watt is major overkill for my little carousel, but if the price is right, it would probably save enough time to be justified.
I think the list in the UK for a 005-SE is approx £1000 plus VAT, obviously list and what you actually pay is quite a bit different but it gives you an idea of tha ballpark figures.
It often makes me chuckle when I see them listed on eBay second hand, often they are more than you could get one new and paying full list ;D
Hood
Ebay is a good resource, no doubt, but you do need to know what you're buying. I've seen the same thing that you've mentioned.
I definitely can't spend that kind of money on a single component for a prototype when I already have a solution in the J2S, but if I sold the J2S setup, that would cover the cost of the AB. The Mitsubishi drives are about ideal for a 4th axis and it seems the AB is ideal for a turret/carousel type application. I guess you just have to pick your poison.
Thanks very much for the info.
-
I've gotten a couple of questions about the speed of the ATC, so I thought I'd expand on that a bit.
The gearboxes are rated for something like 2,400 RPM input continuous and there is no 'instant' or 'momentary' rating. The Mitsu servo motor on the arm has a max instant rating of 6,000RPM and I have tested the gearbox and shaft at that speed without any problems. I suspect the only issue would be heat and in this application it is doubtfull there would be any significant build up, but I'll know more about that when I start doing continuous torture testing.
The reason the speed is held back in the first test video is because the arm is not counterbalanced and the whole machine tries pretty hard to jump off the table if it is run much faster than is shown. I had not considered a counter balance for the arm, but it became an obvious punch list item as soon as I fired the thing up with the arm on it . . LOL!!
Some compromise weight will need to be chosen for a counterbalance since the tool weights will vary quite a bit, but anything will be better than nothing in this case.
Since the control system for the ATC will be so simple :P, I may as well add a speed parameter per tool position. In this way the arm speed can be programmed to be appropriate for the tool. It occurs to me that this may actually net a significant improvement in the overall throughput.
Question for Hood and/or anyone else who has a commercial ATC; is the operating speed of the ATC fixed at a certain rate or can it be altered per tool? If not, then can it be altered overall?
The servo drive is easy enough to change speed, of course, but the air cylinders would be a bit more complicated to speed control. Anyone have some ideas on how to accomplish that?
-
And one last comment on the TB30 spindle;
The spindle is reassembled with the new die spring and initial testing shows it to work perfectly. There is considerably more travel and also more options as far as setting the forces. Currently it is set to a calculated force of approx 1,400lbs and there is an additional .050" travel after release of the tool.
This compares with . . if I remeber right . . about 1,000lbs with the bellevilles and a scant few thousanths of travel after release.
I did not rebalance the spindle, so I can't comment on that aspect, but I don't forsee a problem there. If anything, at a few pennies under 20 bucks, I would be inclided to buy another spring and just balance the spring itself and call it a day. This spindle will run at max 7,500 and I will bench test it at that speed once I get the drive system built for it.
-
Regarding the AB drives, not sure if I made it clear when I said disabling the sercos turned them into Indexing drives. What I maybe should have said was drives capable of Indexing. You can use them as normal servo drives with either position (step/dir, encoder follower etc) or analogue input. You can also use them as Indexing drives, in fact you can use them as both with an input to change between modes. Example an input could switch between Indexing mode and Position mode (Step/Dir or Follower) or between Position and Analogue or Analogue and Indexing etc.
The advantage I have is I am using these drives for myself so I can afford to use second hand ones and I refuse to pay too much for them, that may change is I was really desperate for one but at the moment I have quite a few spares and the most I have paid was under £300 for the 22Kw ones. There are lots of the smaller drives (3Kw and under) for £300 or so but being a poor boy (as I was informed in another thread ;) ) I wont buy at that price.
Your problem with this approach is you are putting a product to market so even though picking up a cheap one for your proto may be an option, you would have to consider the new cost for your product.
Regarding speed, I dont have any commercial changers with the exception of the Chirons and it is air cylinders. They are adjustable like any air system is and have actually been reduced from default. Tool change is supposed to be 0.9 seconds tool to tool but with the Up/Down cage being slowed it is approx 1.8seconds.
The lathe did have a turret on it when I got it but it was hydraulic and worked by rack so really just one speed unless you adjusted the flow rate. The turret I made to replace that has the servo and I have limited the rotation to what I think is suitable and wont put too much stress on things, I could have turned the speed up to probably 5 or more times what I have. Heres a vid showing the speed I settled on. http://www.youtube.com/watch?v=hRTfxuXf6yk
Hood
-
Meant to ask, what is the cost of the Mitsu drives? Say the 400watt ones.
Hood
-
The servo drive is easy enough to change speed, of course, but the air cylinders would be a bit more complicated to speed control. Anyone have some ideas on how to accomplish that?
Steve,
You have to do it be either throttling the air flow, or damping the motion. I would think you could make an adjustable air flow valve by making a needle valve actuated by a small stepper or servo - perhaps an RC servo? You might be able to adapt something from an automotive idle air control valve, or perhaps even a fuel injector and a simple PWM control.
Regards,
Ray L.
-
Steve,
Thinking about it some more, I think a solenoid-controlled needle valve driven by a simple (even software) PWM would be the way to go. Simple to implement, and should give very good control.
Regards,
Ray L.
-
Nice try stevo BUT " your" words proved my point . IF you look back through your posts the main objective was to sell someone on the site your wares. AND it was hard to sell your wares to Tormach users as long as Tormach LOCKED you out of the picture. Then you were pissed because Greg would NOT work with you so you could access his customer base
Then there is the ME syndrome. Basic genetic flaw in a LOT of engineers who think the world cannot revolve without their guidance.
I dought that Greg needs your help with anything concerning his CNC products. He is light years ahead of you already .
But you Have a good one anyway, YA hear
(;-)TP
-
Hey Steve,
So, give me a web site where I can get those Springs I REALLY, REALLY like that idea over Belviells, I guess my concern would be, how would you keep the spring centered so it would not wobble the shaft?
Also: Having built a Prototype ATC, I would NOT try and use "Air Cylinders" for any type of speed control. It is very hard to get the flow valves right, and the instant that your pressure, or humidity changes at the source, then those speeds are not longer valid........ Other than Bang-Bang motion, they are not good velocity or positioning devices. It would probably make MUCH more sense to do a Rack/Pinion or acme screw/nut for linear motion.
Having said the above, another way, but more expensive, is to drive the entire contraption via PLC brain, you could use "High Speed counters" or Stepper/motion cards to drive or read, or drive/read your motors. You can use inexpensive AC or DC motors, and put an Encoder on the motor or on the moving part, your choice, and read its distance via the High speed counter, you can set all kinds of very good positioning that way. The next level up would be stepper (or servos), controlled via a motion card on the PLC (with Encoder feed back). Many of these cards you can select various "Motion Profiles" and set up motion/velocity tables, with set positions that can be called. (this applies to both linear and rotary motion).
IF your real, real hell bent for leather, then you could just use analog cards, with linear rheostat for feed back coupled with a PID loop per motion axis/device...
Many, Many ways to kill that cat, even do PID loops in a plugin if you like the pain....
Scott
-
Scott,
If he were to do what I suggested, using a PWM-driven solenoid valve, if he put some kind of encoder, or even just a potentiometer, on the arm, I bet he could actually control the speed of the air cylinder quite easily, with a very simple, very low bandwidth PID controller.
Regards,
Ray L.
-
You could push Light hydraulic fluid(ATF) instead of air .
Just a thought, (;-) TP
-
You could push Light hydraulic fluid(ATF) instead of air .
Just a thought, (;-) TP
But that would require a hydraulic cylinder, pump, reservoir, etc....
Regards,
Ray L.
-
You could push Light hydraulic fluid(ATF) instead of air .
Just a thought, (;-) TP
But that would require a hydraulic cylinder, pump, reservoir, etc....
Regards,
Ray L.
We use air cylinders with a parallel mounted oil filled cylinder and res. with flow controls to control the movement. (I mentioned this option early on but didn't describe it well). There are cushions built into the "OIL" cylinder which decells and softens the stopping at each end of the stroke. We use Festo mfg'd. units as well as making up our own and they work VERY well. I am in the process of changing one out for a servo and ballscrew on a precision saw slide as we want to vary the feedrate for different materials/conditions.
This method however does not addres the the issue of changing the actual speed with the SW. Will be interesting to see if you folks can come up with a feasible way to incorporate the two. That would be something I could possibly use in the future.
Russ
-
Similar to Ray's approach, there is a board ...
http://www.cnc4pc.com/Store/osc/product_info.php?cPath=25&products_id=179
that could possibly be affixed to a FCV instead of a Pot to control the speeds via SW.
Would be cool to try.
Russ
-
The software is the "Easy(er)" part, if you go with PWM air solinoids, your still having to drive that solinoid with a PWM source, and you would also need
some type of feedback, i.e. analog, pwm, encoder etc. I think this is probably just as expensive to do as using a PLC with motion card (and/or Analog card).
BUT, having said THAT...... Aurturo sells a Mod bus card and Peter Homann sells a Modbus card, both are very cheap, and both can do PWM (spindle control).
They also have multiple analogs on them, so you could also do some kind of feedback..... cheaply.
Hydraulic/hybrid system was mentioned above, air over hydraulic. That adds a large cost, in the hydraulic components, even small units are usually close to or over 1k.
I personally, like the "KISS" principle, which may be just push a ACME screw with a nut, with a cheap stepper and drive (Chinese drives are real cheap), and/or use cheap AC or DC motors to push the screw and use mechnical Micro switches (or proxes if you have a little more cash), to handle your bang-bang limits. You could even put micros before the end micros, that could be used to cut the speed toward the end of travel prior to the stop micro.
Really, for an ATC "Kit" or even OEM targeted device, especially if for smaller machines (ran by primarily hobbyist), cost would be a major issue.
You would have to add in your Warranty and Liability cost as well.
Unless your shop is tooled up to run production on your ATC product, it might be better for you to just sell the IP outright with the understanding of you being clear of all claims and support.
been there, done that......... :)
Another option, that might make you some ongoing cash, is to offer the kit in "Plan" form, or DVD etc. I would probably go with real paper Prints, since it is just to easy to copy/distribute DVD materials. You could offer phone Support "Packages" for some set amount for some hours, as an option. At the end of the day, it's your show and your choice on how you would like to proceed.
But, I really like the changer your doing, and still want a source for those MONDO springs!!
Scott
-
A simple Air/Hydraulic unit consists of a reservoir of oil with air pressure feeding in the top and oil flowing out the bottom. I use an Air solenoid to charge the Reservoir when flow is needed. Control the feedrate from the oil side with an orifice or needle valve. Great for simple single side cylinder charging and spring return OR you can use a double side double feed valve to do double side cylinder control.
This works WELL for low pressure applications such as this.
You can also do the Charge cyylinder trick with air pressure and a closed loop for the Oil side. Single or double acting as well.
There are many low cost ways
Just a thought, (;-) TP
-
Don't want to use air (;-) Use a power steering pump to drive your hydraulics. Low cost self contained reservoir drive it with a small electric motor at a lower RPM.
(;-) TP
-
Scott,
I don't think you quite understood what I was suggesting. For control, the PWM is implemented purely in software, the only hardware being an external pass transistor to drive the solenoid. Then there's just the solenoid valve itself. An automotive idle air control valve would probably do the trick for next to nothing. Feedback is provided by a simple potentiometer, read by a built-in A/D in the MCU. Can't get any cheaper than that. I'd bet the whole thing wouldn't cost $10.
Regards,
Ray L.
-
:)
-
IF you wanted slow zones you can use a rotary actuater with a drawbar link. As the rotary arm rotates from retracted to extended position the link geometry SLOWS the tool arm to a stop as the arm reaches Max+ and Max- travel. No jerks at the extents of travel.
(;-)TP
-
I personally, like the "KISS" principle, which may be just push a ACME screw with a nut, with a cheap stepper and drive (Chinese drives are real cheap), and/or use cheap AC or DC motors to push the screw and use mechnical Micro switches (or proxes if you have a little more cash), to handle your bang-bang limits. You could even put micros before the end micros, that could be used to cut the speed toward the end of travel prior to the stop micro.
:)
-
If anyone's interested,Ive just made power draw bar for my tool changer,used a stack of belvilles,a 20mm spring collet ala tormach.
To power it I used 2 50mm dia 20mm stroke air cylinders in parallel,with a bar joining the 2 piston rods, that in turn operates on a pivoted crank arm giving a 3-1 advantage,its got loads of power I can plough a 18mm dia ripper 10mm deep through ali at 250mm pm no problems no slip, good enough for me and most folks I guess ?
My first attempt was a large stepper with a 1.5mm pitch lead screw operating the 3-1 lever I had micro switches, and a logic circuit controlling it.All i can say is it cost me a fair amount of time and money and never really worked reliably, and even with all that mechanical advantage it never really had the grunt.I would advise from my own experience don't go down that route!
Tony
-
If anyone's interested,Ive just made power draw bar for my tool changer,used a stack of belvilles,a 20mm spring collet ala tormach.
To power it I used 2 50mm dia 20mm stroke air cylinders in parallel,with a bar joining the 2 piston rods, that in turn operates on a pivoted crank arm giving a 3-1 advantage,its got loads of power I can plough a 18mm dia ripper 10mm deep through ali at 250mm pm no problems no slip, good enough for me and most folks I guess ?
My first attempt was a large stepper with a 1.5mm pitch lead screw operating the 3-1 lever I had micro switches, and a logic circuit controlling it.All i can say is it cost me a fair amount of time and money and never really worked reliably, and even with all that mechanical advantage it never really had the grunt.I would advise from my own experience don't go down that route!
Tony
Sounds like a very reasonable solution. By my calculation, you're getting on the order of 1500# drawbar tension (assuming some frictional loss) - not enough for really maximum retention on a TTS holder but enough for many/most machines. A leadscrew I would not expect to work well - the frictional losses in the screw/nut will kill you.
I've got a new PDB design I'm building now that will be REALLY cool, simple, cheap, and tiny - even smaller than the impact wrench PDBs. It's already been proven functionally. But, it's top-secret for the time-being, because I plan to sell them.
Simpson better get cracking on his ATC! I'm already working on my second one, for my new machine!
Regards,
Ray L.
-
Himmy, will it be controlled by a single hi/low signal?
tony
-
The little hydraulic intensifier at the heart of this drawbar is the coolest design I've seen for a Belleville drawbar solution:
http://www.sdmfabricating.com/LOD.html
That intensifier plus one small air cylinder is all you need.
Regards,
Ray L.
-
Well thats nice but OTT, compared to my method,plus Ive a few of my cylinders that I can just add to increase the power.
As far as Bellvilles are concerned Ive used thinner than usual washers and put them in ,in sets of 3 back to back my thinking being that they will be more supple,abit like a leaf spring,seems to work
Tony
-
keep them well greased when you use them stacked, as they tend to rust between each one.
-
there are belleville washers and then there are disc springs which are the Items that you should use, On vehicle leaf springs its advisable not to grease between the leaves as it attracts grit and forms a grinding paste,or so ive been told! probably a good thing in our application to lubricate though!
Tony
-
Meant to ask, what is the cost of the Mitsu drives? Say the 400watt ones.
Hood
Sorry for the late response. Been a busy boy.
New price is a bit under US$1,000 each for the drive and the motor. Ebay has a number of vendors who have good used stuff and usually you can get good clean pair (Drive and motor) for about $800 or thereabout. The 750's usually run close to $1,100 to $1,200 for a set. A caveat with the J3 series is that the cables are stupid expensive . as in . . up to $100 for each cable and they have special unobtainable connectors so anybody shopping for J3 Mistu, make sure you address the cables as they can be a significant part of the cost.
-
Nice try stevo BUT " your" words proved my point . IF you look back through your posts the main objective was to sell someone on the site your wares. AND it was hard to sell your wares to Tormach users as long as Tormach LOCKED you out of the picture. Then you were pissed because Greg would NOT work with you so you could access his customer base
Then there is the ME syndrome. Basic genetic flaw in a LOT of engineers who think the world cannot revolve without their guidance.
I dought that Greg needs your help with anything concerning his CNC products. He is light years ahead of you already .
But you Have a good one anyway, YA hear
(;-)TP
Jeez, don't you get tired of being wrong? ::)
I have no trouble selling to Tormach owners. If I did not sell to Tormach owners, I would not have any reason to work with Tormach on their MACH software mods. You don't have to be a conceited, narcissistic engineer (like myself, apparently) to make that connection. Just have some common sense.
Also, you are just delusional if you think putting together a stepper driven bench mill and using an entire tiny lathe as a 4th axis is 'light years ahead' of anyone. Perhaps those examples of engineering prowess are impressive to you, but honestly I think even the greenest newbie can get to that level in a few months.
Next correction; I was not and am not 'pissed' at Tormach or any individual there. Nor am I 'pissed' at you. You are entitled to make as many inaccurate, incorrect or even idiotic statements as you like. I am entitled be amazed at baseless and idiotic some of your comments are. I am firm believer in the serenity prayer and your Tormach fanboi affliction and your belief that your assumptions are facts is one of those things that I can do nothing about.
And to correct your last incorrect assumption, I am not 'locked out' of anything. Tormach's MACH implementation took about 5 minutes to crack and I did the QC myself in about 5 more minutes. As I said, this stuff is not rocket science. Some people like to pretend that it is so that their knowledge of it seem oh so impressive to newbies. If you simply copy a standard copy of the MACH .exe over Tormach.s, it come up with access to all parameters. Oh my - OOPS,. was that supposed to be a secret? . . Sorry Obi Wan. . .
-
Hey Steve,
So, give me a web site where I can get those Springs I REALLY, REALLY like that idea over Belviells, I guess my concern would be, how would you keep the spring centered so it would not wobble the shaft?
Scott
Sorry for late replay. Your question about the centering will be answered in the next few posts. I'm happy to give you the source and part# for the spring, but since it is a potentially dangerous part, please PM me so I can give you the standard disclaimer blurb along with the info.
I will address the air cylinder throttling separately. Thanks for your comments!
-
Steve,
Thinking about it some more, I think a solenoid-controlled needle valve driven by a simple (even software) PWM would be the way to go. Simple to implement, and should give very good control.
Regards,
Ray L.
Right on the money, as usual. The trick is to find something 'off the shelf'. I'm thinking there must be a computer controllable (PWM, analog, or whatever) orifice or regulator. Hopefully there are some industrial control gurus lurking around here that can provide an example.
-
Hey Steve,
Also: Having built a Prototype ATC, I would NOT try and use "Air Cylinders" for any type of speed control. It is very hard to get the flow valves right, and the instant that your pressure, or humidity changes at the source, then those speeds are not longer valid........ Other than Bang-Bang motion, they are not good velocity or positioning devices. It would probably make MUCH more sense to do a Rack/Pinion or acme screw/nut for linear motion.
Scott
I have to agree with you from the aspect of control. While I was noodling around on ways to power the various actuators, I discarded rack and pinion because of the difficulty in keeping it clean. The environment is difficult because of the flying swarf, coolant, etc. My career was primarily fault analysis on mechanisms in industrial environments, so I am always thinking in terms of 'what would happen if someone shoveled gravel into this machine' . :o
-
Hey Steve,
Having said the above, another way, but more expensive, is to drive the entire contraption via PLC brain, you could use "High Speed counters" or Stepper/motion cards to drive or read, or drive/read your motors. You can use inexpensive AC or DC motors, and put an Encoder on the motor or on the moving part, your choice, and read its distance via the High speed counter, you can set all kinds of very good positioning that way. The next level up would be stepper (or servos), controlled via a motion card on the PLC (with Encoder feed back). Many of these cards you can select various "Motion Profiles" and set up motion/velocity tables, with set positions that can be called. (this applies to both linear and rotary motion).
I have a pair of Mitsu servo motors actuating the carousel and the arm. The whole thing is controlled by a processor running C code. After thinking about Hood's excellent suggestions, I reasoned that writing some code to generate a step stream and run the servo drives would be shorted time-wise than the learning curve for a new type of industrial drive. The AB seems quite a lot more intuitive that the Mitsu, which is extremely complicated to use in that way, but the code to run the drives was relatively simple and I have the carousel routines finished and tested. The arm is the same servo drive (just a newer model) so the same code, slightly modified will do that job as well.
Ultimately I am thinking about following HimyKabibble's lead in implementing a Kflop and connecting 'satellite' task specific processors to it, but that's a big meal that I don'thave time for at the moment, so I elected to pile the ATC code on top of my existing InTurn™ 4th axis motor controller since it is already tied in and talking to MACH. The little processor has it's hands full as it is, but a tool change happens 'in a vacuum' so to speak when everything else is on hold, so it will probably be fine.
IF your real, real hell bent for leather, then you could just use analog cards, with linear rheostat for feed back coupled with a PID loop per motion axis/device...
Many, Many ways to kill that cat, even do PID loops in a plugin if you like the pain....
Scott
Running in the dark . with scissors . . and my hair on fire . . is pretty much normal condition for me . .
-
Simpson better get cracking on his ATC! I'm already working on my second one, for my new machine!
Regards,
Ray L.
Hey, haven't you won this thing yet? I caught you sand bagging one sensor as I recall. Is that done?
-
Steve,
Thinking about it some more, I think a solenoid-controlled needle valve driven by a simple (even software) PWM would be the way to go. Simple to implement, and should give very good control.
Regards,
Ray L.
Right on the money, as usual. The trick is to find something 'off the shelf'. I'm thinking there must be a computer controllable (PWM, analog, or whatever) orifice or regulator. Hopefully there are some industrial control gurus lurking around here that can provide an example.
In hydraulics, you get electrically controlled variable flow control valves, and I'd guess someone will do something similar for air.
However the usual way for controlling cylinder speed with air is by controlling how quickly air escapes from the non-powered side of the cylinder. The flow restrictors are usually plumbed in with a one way valve, so you get full flow into the cylinder, but any flow out goes through the restrictor.
-
OK, I think I have answered all of the questions. Couple general comments;
* Not interested in introducing hydraulics.
* PWM out and analog in and out are standard feature of the current crop of tiny processors so extra boards are not needed.
*Anything that requires operator intervention is contrary to the design goals here.
* cost is always a consideration except where safety is concerned, but it is not high on the menu here as this device is not intended for hobbyists . . per se.
* Bang-bang is not the concern and is easily handled by cushions built into the cylinders. It is a standard option with Fabco, for example.
* fixed speed control of the air cylinders id not a concern as that is easily accomplished with available adjustable orifice valves specifically for that purpose. I have one of those on the drawbar actuator in the 'grip' direction to reduce the shock load on the gripper. Shock load is beneficial and perhaps even needed on the release cycle, so there is a small acceleration gap and no restricts.
* the challenge is to be able to computer control the speed of the air cylinder that raises and lowers the arm. Given that for any specific part, the overall thoughput can be significantly effected by optimizing each operation, it makes sense to run the ATC as fast as possible with small low mass tools and slow down for very heavy or large tools. I have read volumes of concerns over drawbar release pressue being taken by the spindle bearings, but in most cases that is really a needless worry as the static load on the bearings is typically far above the drawbar pressure. However, taking a very heavy tool and slamming it into the spindle taper is not a good plan. Note that industrial robotic loader/unloader mechanisms almost always decel just off the target and gently deliver the load.
-
Steve,
Thinking about it some more, I think a solenoid-controlled needle valve driven by a simple (even software) PWM would be the way to go. Simple to implement, and should give very good control.
Regards,
Ray L.
Right on the money, as usual. The trick is to find something 'off the shelf'. I'm thinking there must be a computer controllable (PWM, analog, or whatever) orifice or regulator. Hopefully there are some industrial control gurus lurking around here that can provide an example.
In hydraulics, you get electrically controlled variable flow control valves, and I'd guess someone will do something similar for air.
However the usual way for controlling cylinder speed with air is by controlling how quickly air escapes from the non-powered side of the cylinder. The flow restrictors are usually plumbed in with a one way valve, so you get full flow into the cylinder, but any flow out goes through the restrictor.
Exactly correct, however, the objective is to find a computer controlled restrictor. Must be something out there. I'm just being lazy really . . don't want to do the research :-[
-
OK, here we go . . finally.
Air cylinder speed is a topic, so I'll start with some mods to the arm raise/lower actuator cylinder. When I removed the restrictor, I found it actually did not move as fast as I wanted, so I had to take it apart and drill the air passages a bit larger and use 1/4" tubing. Fortunately there are 1/4" tube x 1/8" NPT fittings so I di not have to drill and re-tap the ports. It's plenty perky now.
Not so exiting photo here (don't worry it gets a lot better in the next couple posts):
(http://www.thecubestudio.com/BT30SpindleProject/Arm_cylinder_ModWEB.jpg)
-
I'm going to back track a bit now. I don't think I covered this well, and there are questions so here is the story on the 'big spring'. The photos are self explanatory for the most part. The spring is held at center by the washers above and below the spring. The smooth bore for the spring pocket is to allow the spring retention washers to slide smoothly. The spring itself does not rely on the bore for centering. In fact is does not touch the bore at all:
Belleville chafing problem is evident on the first photo. The bar gets an initial compression and is retained by a snap ring in the top of the spring pocket. This has a side benefit in preventing the bar from extending up and pressing against the actuator when there is not tool int eh spindle nose.
(http://www.thecubestudio.com/BT30SpindleProject/Drawbar_Spring_Assembled01WEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/Spindle_DrawbarComponentsWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/SpringPocketWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/Drawbar_ExplodedViewWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/Spindle_Drawbar_InSpringPocketWEB.jpg)
-
A detail I don't think I covered previously. If I did, then sorry for repetition here:
What I consider to be THE most important part of the whole device is this little guy. The pin is hardened steel and rides in a hardened die bushing in the drawbar actuator. This pin protrudes from the bushing and provided a positive mechanical interference with the actuator's operation. In other words, the spindle CANNOT release a tool unless this solenoid is first activated removing the interference. The actuator has some serious Kahunas, so the hardened components used here are so that the interlock can get more than one use. :P
(http://www.thecubestudio.com/BT30SpindleProject/Actuator_InterlockPlunger_CloseWEB.jpg)
-
I think that to run higher than about 8k to 10k, it would be advantageous to have an extremely rigid spindle. This will be much more resistant to harmonics. I made a spindle using A6 tool steel. Properly hardened and tempered and fully ground, it will undergo torture testing in my InTurn™ 4th axis. Obviously, I will not be running the 4th axis at 15K RPM, but it is a brute and quite capable of testing the strength and rigidity of the new spindle. Once I am confident of its performance, I will make a similar one for the upcoming high speed BT30.
(http://www.thecubestudio.com/BT30SpindleProject/Spindle_InTurn_GroundWEB.jpg)
-
OK, moving on to some chunky parts;
The bottom of the new head is a 1" thick chunk of pre-ground mild steel (probably A36). Initially, the new head will be bolted together and I may take it down and weld it up at a later time after all adjustment, mods, tweaks, etc are completed. The sides are also steel plate with the ATC side being 1/2" thick and the InTurn™ side being 1" thick. The entire head and ATC will be assembled on a test stand that I made up in order to work on the thing safely until the new mill column is ready. I am having a very tough time deciding what to use for the column, but that's another story.
Hanging at the top of the test stand is the head's upper mount. The lower mount is the base in the first photo. The soda can is just fore scale . . I did not make that part. ;)
(http://www.thecubestudio.com/BT30SpindleProject/Head_BasePlateWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/TestStandWEB.jpg)
-
Steve for the column you may want to consider 6" Square X .500 wall 4130 tubing.
(;-) TP
-
Moving on the everyone's favorite, the <dun...duuuun> CLAW.
Hope you are paying attention Ray, 'cause here come the sensors.
There are a bunch of things to keep tabs on with this crazy thing and some relief comes in the form of 'position reached' signals available from the servo drives. All positioning of the arm and the carousel will rely on those signal from he drives and will serve as 'sensors' for that part of the sequence.
Things that are monitored in the following photos are:
1) is the CLAW open?
2) is the CLAW closed? (same sensor, but a different event)
3) is the interlock in place?
4) is the interlock retracted? (same deal)
For continuity, I'll mention these as well although they will be in the next post along with the new spindle lock
5) is the spindle lock engaged?
6) is the spindle lock released? (seems redundant, but some events need two separate sensors to monitor each state)
The first photo is the overview. The second is self explanatory and shows the switch that 'watches' the claw closure.
The last two show the switch that looks at the solenoid. This was a little more complicated in that I had to make the specialty shaped plunger from scratch and add a small rod that extends outside the solenoid case and activates the switch. These particular solenoids do not come with a spring extender so I have to add that internally to both this solenoid and the drawbar interlock solenoid.
The solenoids and the switches are sealed, but I think that I shall still put a cover over the whole works on the arm to keep junk from interfering with the workings. There are a million nooks and crannies that all need to stay relatively clean.
(http://www.thecubestudio.com/BT30SpindleProject/Arm_Sensors_AllWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/Claw_Sensor_Close02WEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/Arm_Sensors_CloseWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/Claw_Sensor_Open_CloseWEB.jpg)
-
Steve for the column you may want to consider 6" Square X .500 wall 4130 tubing.
(;-) TP
I am considering 8" square x .5" wall or 6" x 10" rectangular x .5" wall. While it does not seem so intuitively, the tortion characteristics are the same. Either of these sizes can ship UPS in a three foot length. I really want 12" square, but the shipping is truck and just stupid expensive. A local steel yard here can get is without hitting me with the shipping, but they require that I purchase the entire 21 foot piece at something well over $1,000. They will cut it up for me though, so that's in the plus column.
Really I want cast iron. Grey cast iron has something like 30 times (from memory) the damping of steel. An important consideration. Note that only Grey iron has this characteristic, not ALL cast iron. I am thinking about using the column from one of those small el-cheapo off shore knee mills. They are all over the place and cheap. There is one locally on Craigs list for $1,000 and I think I could get it a lot cheaper than that even. The irony of it would be if they don't use Grey iron. Its for sure they do not use mehanite. It might actually be better to use the steel tubing and fill it. I just have not made up my mind on that yet.
One thing I have decided on is the table. I have a 42" x 9" Bridgeport mill table on its way. 'Previously Owned' of course, and hopefully useable as is. Paid $500 with ship, but if I have to have it ground or otherwise tinkered with , that cost could easily double.
-
Steve,
"The soda can is just fore scale . . I did not make that part." - I was REALLY impressed, until I read that.... :-)
You've been busy! All looking really good!
On the airflow control - look at automotive idle air control valves. I'm not sure what the airflow range is, but the functionality is exactly what you need, and they're pretty cheap. I suspect they could be modified for lower flow, if necessary.
Those are nice little solenoids you're using. Where do you get those? What do they cost? I'll probably need some on the new 12-tool fixed carousel ATC I'm designing for my new machine. All the ones I've found so far are stupid expensive.
Regards,
Ray L.
-
Another detail and then on to more cool stuff.
The seemingly obvious requirement of a counterbalance for the swing arm did not occur to me until the whole mechanism tried to jump off the bench during the initial testing.
So the counter balance is bolted on for now. The smaller round weight is for adjustment and I made up a few in different sizes from scrap. Should help considerably. I may make up an entire new arm once the prototyping is completed, of I may just weld the counterbalance to the arm . . . or not . . :-\
(http://www.thecubestudio.com/BT30SpindleProject/Arm_CounterBalanceWEB.jpg)
-
Steve,
"The soda can is just fore scale . . I did not make that part." - I was REALLY impressed, until I read that.... :-)
Alright then, I retract me previous statement. I made the soda can . . oh, and I invented the internet. That Gore guy just wants to steal my thunder.
On the airflow control - look at automotive idle air control valves. I'm not sure what the airflow range is, but the functionality is exactly what you need, and they're pretty cheap. I suspect they could be modified for lower flow, if necessary.
You DO know that IAC are steppers don't you? I do HATE steppers, you know that too right?
OK seriously that is an excellent suggestion and I have handfuls of those things around here from some Crossfire Fuel injection work I used to do. For car buffs with too much time on their hands, this may be an interesting read:
http://www.digitalcorvettes.com/forums/showthread.php?t=101038 (http://www.digitalcorvettes.com/forums/showthread.php?t=101038)
IAC are driven directly off the ECM so it should be very doable and they are easy to mount, so I don't see much of a problem using those. Great suggestion . . . but I still invented the internet, so there.
Those are nice little solenoids you're using. Where do you get those? What do they cost? I'll probably need some on the new 12-tool fixed carousel ATC I'm designing for my new machine. All the ones I've found so far are stupid expensive.
Regards,
Ray L.
http://www.mcmaster.com/#solenoids/=l6bf8u (http://www.mcmaster.com/#solenoids/=l6bf8u) scroll down to sealed
-
OK last post for a while. I have another project hat cropped up so I'll be away from this project for about another 30 days.
Saving the best for last, here is the spindle lock for the BT30. For my current 'Franken mill', I just grabbed one of the calipers for my 4th axis and stuck it on the mill head. It works great even though it is a bit ghetto.
I had planned to use one for this project also, but alas there was no way to get it to coexist with the large diameter spindle cartridge. So back to the drawgin board for a one-off custom caliper for the BT30 project. Basically it is just a modification to the InTurn™ articulated pneumatic caliper with a longer arm to move the cylinder out away from the cartridge tube . . which required a longer stroke cylinder, a thicker body . . and so on down the line. It is a very cool part, I think, but I just stole if from the InTurn™, so it's not really 'new' new.
What is new if the sensor in the form of yet another little sealed limit switch that bears on the caliper arm. This will report that the caliper is either active or that is is open. The cylinder is removed in the last two photos.
(http://www.thecubestudio.com/BT30SpindleProject/SpindleLock_Assembled_BackWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/Spindle_RotorAndCaliperWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/Spindle_Caliper_SensorWEB.jpg)
(http://www.thecubestudio.com/BT30SpindleProject/Spindle_Sensor_CloseWEB.jpg)
-
The real problem with the Coke can is it should have been a PEPSI throwback instead of COKE ZERO. You cannot expect a high powered INTURN to run on unleaded. It needs a can of HIGHTEST PEPSI.
(;-) TP
-
I think Steve and I need to have another contest, to see if he can finish his *first* ATC before I finish my *second* one! :-) I've got the design done, just waiting on the new machine to arrive so I can get the final mounting dimensions. This time I've gone with a very compact. fixed 12-tool carousel (vs 10 for my first ATC), mounted alongside the column - well out of the way of the user (never like having the carousel hanging next to the head....). A high-speed transfer arm will move the tools back and forth quickly (not nearly as quickly as Steve's, but much faster than my current one). I have an active "gripper" for TTS tools on the transfer arm, so I can rotate the arm pretty much as fast as I want, with no worries about the tool flying off. The arm will be driven by a servo, so smooth, rapid motion should be very easy to achieve. I'm keeping the Geneva mechanism, but re-designed it to internal, rather than external, Geneva for much smoother motion. All told, only two small motors, and three very small air cylinders to perform all the motion. One side-benefit is this design will be trivial to adapt to virtually any milling machine from an X2 to a full-size VMC, with only two very simple design changes - the bracket that attaches the whole thing to the column, and the length of the transfer arm. Nothing else would have to change. It would even possible to make a "universal" kit that could be adapted to almost any machine without any machining, other than drilling some mounting holes on the column.
Regards,
Ray L.
-
Nice work Ray.
It sounds like you and Steve are very busy.
Mike
-
I am currently working on the sensors and control software. The last sensors are for the arm up/down and I could find no way to get those on the shaft that satisfied me. It is more complicated than it looks because the shaft will be covered by a rubber corrugated bellows below the gearbox and by either the same bellows or just a cover tube between the air cyl and the gearbox . . and of course the whole shaft rotates. Also I'm not crazy about the air cyl being mounted separately from the gearbox. So I had to go on hold with that until I get a new cylinder with position sensors on the cylinder itself. I have the cylinder on the way and I plan to mount it on a tube supported by the gearbox.
My secondary task was working on the software, but I am building on top of the InTurn™ motor controller and in a moment of apparent masochism, I reasoned that this was a good time to do a major upgrade to the controller and launched into that. Among other things, I changed the interface from serial modbus to plug-in modbus . . which took more than a few minutes. :P THEN, after a few brewskies, it seemed sensible to move my own development box to TCP modbus. After all the setup on MACH's side is similar to the plug-in serial, right? Well, in my setup, the "PLC" runs the modbus slave, so moving from 'serial anything' to TCP was a rewrite of a significant part of the code.
Those are my excuses. If I come up with better ones, I shall post them separately. The work is completed and working now, but apparently too late as it seems the tortoise has crossed the line.
So, Ray, if you are declaring your 'first' ATC completed including all sensors and controls, then all I can say is CONGRATULATIONS and where do I send the beer :)
I accept your challenge to finish my first before you finish your second. However, I counter challenge you to see who is the first to have a completed and sold version operating in the field. I believe I have the drop on you there. :-X
Anyway, at the time you took the checkered flag, I was on this lap: Note that the carousel gearbox can be rotated in 90 degree increments and the drive motor will be tucked in at final assembly. It is flying out there now to make the thing easier to work on and observe the actions.
(http://www.thecubestudio.com/BT30SpindleProject/HeadOnTestStand01WEB.jpg)
-
I am currently working on the sensors and control software. The last sensors are for the arm up/down and I could find no way to get those on the shaft that satisfied me. It is more complicated than it looks because the shaft will be covered by a rubber corrugated bellows below the gearbox and by either the same bellows or just a cover tube between the air cyl and the gearbox . . and of course the whole shaft rotates. Also I'm not crazy about the air cyl being mounted separately from the gearbox. So I had to go on hold with that until I get a new cylinder with position sensors on the cylinder itself. I have the cylinder on the way and I plan to mount it on a tube supported by the gearbox.
My secondary task was working on the software, but I am building on top of the InTurn™ motor controller and in a moment of apparent masochism, I reasoned that this was a good time to do a major upgrade to the controller and launched into that. Among other things, I changed the interface from serial modbus to plug-in modbus . . which took more than a few minutes. :P THEN, after a few brewskies, it seemed sensible to move my own development box to TCP modbus. After all the setup on MACH's side is similar to the plug-in serial, right? Well, in my setup, the "PLC" runs the modbus slave, so moving from 'serial anything' to TCP was a rewrite of a significant part of the code.
Those are my excuses. If I come up with better ones, I shall post them separately. The work is completed and working now, but apparently too late as it seems the tortoise has crossed the line.
So, Ray, if you are declaring your 'first' ATC completed including all sensors and controls, then all I can say is CONGRATULATIONS and where do I send the beer :)
I accept your challenge to finish my first before you finish your second. However, I counter challenge you to see who is the first to have a completed and sold version operating in the field. I believe I have the drop on you there. :-X
Anyway, at the time you took the checkered flag, I was on this lap: Note that the carousel gearbox can be rotated in 90 degree increments and the drive motor will be tucked in at final assembly. It is flying out there now to make the thing easier to work on and observe the actions.
Steve,
Wow! That is an impressive bit of machinery. I don't understand it.... But I'm sure it'll make sense when you post the first video! :-)
I declared by first ATC "done" some time ago, and it's been in-service for a month or more, with no problems so far. The only "missing" piece, briefly, was spindle "stopped" sensing, which is now provided by the VFD.
My second ATC (still TTS, but could very easily accommodate 30-taper with pretty minor changes) is now completely designed, and I'm busy sourcing components, and doing CAM. If you beat me to the first "sold" unit, it won't be by very much! I'm actually designing this one for a machine manufacturer, and designed it specifically to be almost trivial to adapt for virtually ANY machine by changing only the mounting brackets and the length of the tool transfer arm, which can easily be made adjustable length. Not hardly as impressive as yours, but obviously aimed at a very different market. I actually think it's about the best work I've ever done - a number of things I've never seen done the way I did them, which is the secret to the very low cost. It can easily sell for FAR less than any ATC of comparable capacity and performance I've ever seen. And, it does not really sacrifice performance either - I expect complete toolchanges to be under 10 seconds, from spindle stop to spindle start, compared to 20 seconds on my current one. The motor-driven PDB is *tiny* - only a 2" x 3.5" footprint, 9" tall, but capable of well over 30+ ft-lbs drawbar torque (actually, it's capable of generating enough torque to just twist the drawbar right off...), but MUCH faster than my current one. The ATC is 12-tools, with a fixed carousel, only 10" in diameter, and a high-speed tool transfer arm (relatively speaking - not in the same league as yours!). The PDB, carousel, and transfer arm are all servo-driven, and there are sensors on *everything*, so virtually any fault will be detectable. Only one air cylinder in the whole thing, a small one at that, to operate the "lift" that inserts/removes tools to/from the spindle. Sensors include carousel position, transfer arm position, transfer arm tool "claw" lock sensor, transfer arm "lift" position, voltage and current for all the servo motors, and more. With luck, the prototype will be done sometime in March. Can't start fabrication until I complete the current build (biggest to date) of one of my other products.
When do we get video?
Regards,
Ray L.
-
NIce design Steve, Might want to check your camera though it seems to have a splotch of grey paint on the lense (;-)
(;-) TP
-
NIce design Steve, Might want to check your camera though it seems to have a splotch of grey paint on the lense (;-)
(;-) TP
Too funny! Thanks for the heads up. I'll have to get that lense checked out. ;)
-
I declared by first ATC "done" some time ago, and it's been in-service for a month or more, with no problems so far. The only "missing" piece, briefly, was spindle "stopped" sensing, which is now provided by the VFD.
OK that 'missing sensor' was the last status I knew of. I hereby concede defeat. Just need to know where to send the beer.
My second ATC (still TTS, but could very easily accommodate 30-taper with pretty minor changes)
This will be quite a challenge, I would think. Your TTS based design does not seem capable of gripping a stud. You must have something up your sleeve again.
If you beat me to the first "sold" unit, it won't be by very much!
Too late! My first and second BT30 spindle prototypes are both sold and the ATC is sold, including the entire new head. Pretty much the whole setup as you see on the test stand will be installed on a new mill as part of a larger project.
I'm actually designing this one for a machine manufacturer, and designed it specifically to be almost trivial to adapt for virtually ANY machine by changing only the mounting brackets and the length of the tool transfer arm, which can easily be made adjustable length. Not hardly as impressive as yours, but obviously aimed at a very different market. I actually think it's about the best work I've ever done - a number of things I've never seen done the way I did them, which is the secret to the very low cost. It can easily sell for FAR less than any ATC of comparable capacity and performance I've ever seen. And, it does not really sacrifice performance either - I expect complete toolchanges to be under 10 seconds, from spindle stop to spindle start, compared to 20 seconds on my current one. The motor-driven PDB is *tiny* - only a 2" x 3.5" footprint, 9" tall, but capable of well over 30+ ft-lbs drawbar torque (actually, it's capable of generating enough torque to just twist the drawbar right off...), but MUCH faster than my current one. The ATC is 12-tools, with a fixed carousel, only 10" in diameter, and a high-speed tool transfer arm (relatively speaking - not in the same league as yours!). The PDB, carousel, and transfer arm are all servo-driven, and there are sensors on *everything*, so virtually any fault will be detectable. Only one air cylinder in the whole thing, a small one at that, to operate the "lift" that inserts/removes tools to/from the spindle. Sensors include carousel position, transfer arm position, transfer arm tool "claw" lock sensor, transfer arm "lift" position, voltage and current for all the servo motors, and more. With luck, the prototype will be done sometime in March. Can't start fabrication until I complete the current build (biggest to date) of one of my other products.
Speaking of 'air speed' I did some research and came up with a number of digitally controlled presure regulators, so I will be going that way instead of messing with an automotice IAC. A very clever solution though, in any case.
I do hope you plan to stick around and share this with the audience. Have you noticed the number of reads this thread has amassed? Besides, I need you good advice as well as others. I'm not so clever really. I just listen when smart people talk. I have ordered a Kflop, so you're not getting rid of me any time soon.
When do we get video?
Regards,
Ray L.
I'm also finishing up a project, and there is another small one right behind it, so it will be a couple of weeks before I can get the rest of the ATC software done and make a video of the testing. As I mentioned, I am changing the arrangement of the air cylinder that actuates the vertical arm movement, so I will probably be covering that with photos before I do a video. Finishing the ATC is actually now a part of the upcoming large project, so while it was going to be on hold for a couple of months, now it will get lots of attention.
-
My second ATC (still TTS, but could very easily accommodate 30-taper with pretty minor changes)
This will be quite a challenge, I would think. Your TTS based design does not seem capable of gripping a stud. You must have something up your sleeve again.
Obviously the drawbar would have to be changed, but once that's done it should work. Although, I could forego pullstuds and just use a screwed-in drawbar, as it it now. I'd grab the toolholders by the groove, basically as I do now with TTS. The carousel disk and spring "forks" would have to change, as would the fork on the transfer arm, but those are all simple changes. I think it could be done by changing only about a half-dozen parts, and perhaps changing the "lift" travel.
I'm up to my eyeballs in KFlop stuff at the moment. I've made quite a few improvements to my controller app recently, and I'm now doing some re-wiring in my E-Box, to clean things up.
Regards,
Ray L.
-
To Anyone:
- How important would it be to have a manual (pendant or equiv) method of operating the ATC? In practical terms, the carousel has to be loaded with tools and so on. My thinking is that inputting a tool change into the MDI is adequate, but it would be good to know if there are any 'wish list' methods that people would like to have, and/or to know what 'real' ATCs provide in terms of manual operation.
- How important would it be to have the ATC perform some type of automatic tool touch off? It would be good to get a discussion going on how this works on machines that have it, and how it might be implemented. I saw one that appeared to use lasers.
To Ray:
- You originally mentioned that you were keeping the Geneva for the carousel in your new design, but your latest description says 'servo' powered. Are you using that term generically, as in 'motors' (i.e. steppers) or have you come over to the dark side?
- In *easily converting your new ATC to work with 30 tapers, how do you accommodate the drive dogs?
*you always say everything is 'easy'. Flying is easy for a duck. It is a bit of a challenge for a donkey. Depends on how you're equipped, I would say . . . ;)
-
On My ATC the carousel is the C axis. I jog the axis and load the tools by hand and then do MDI sending the C to 0 which is tool 1. I started doing it this way until I could do a custom screen set but found it was easy to do so I'll probably continue to use this method.
I don't have a tool setter on the mill. I have one on the router and it's quite accurate. On the mill it's no big deal for me to touch off manually in one way or another so I'll probably stick with that. The BT30 seems to have good repeatability so I don't envision me ever touching off after every tool change.
Just finished a run of 380 pieces. More than 400 tool changes over a 6 day period without a single miss. Without a doubt the ATC has been some of the best time and money I've spent on the mill since the original conversion. The newness has worn off and I don't even pay any attention to it anymore. It just chugs along making me money!
Derek
-
Question on rotating mill head::
I have noted that Bridgeport's CNC mill has a non rotating head as does the Smithy CNC bench mill and others. Years ago when I had a shop with manual mills it was not uncommon to have the heads turned this way and that, but I have yet to ever rotate the head on any of my CNC mills.
Unless the rotation is actually an axis under CNC control, I have some difficulty seeing a justification for it. While I think it would be beneficial to build in some kind of tramming 'assist', I am leaning toward having the head NOT rotate, per se. :-\
Thoughts on this?
Thanks in advance,
-
On My ATC the carousel is the C axis. I jog the axis and load the tools by hand and then do MDI sending the C to 0 which is tool 1. I started doing it this way until I could do a custom screen set but found it was easy to do so I'll probably continue to use this method.
I don't have a tool setter on the mill. I have one on the router and it's quite accurate. On the mill it's no big deal for me to touch off manually in one way or another so I'll probably stick with that. The BT30 seems to have good repeatability so I don't envision me ever touching off after every tool change.
Just finished a run of 380 pieces. More than 400 tool changes over a 6 day period without a single miss. Without a doubt the ATC has been some of the best time and money I've spent on the mill since the original conversion. The newness has worn off and I don't even pay any attention to it anymore. It just chugs along making me money!
Derek
Good info. Thanks, Derek!
-
The only time I rotate my head is to set the tram. There have been times when I wish it was fixed.
-
To Anyone:
- How important would it be to have a manual (pendant or equiv) method of operating the ATC? In practical terms, the carousel has to be loaded with tools and so on. My thinking is that inputting a tool change into the MDI is adequate, but it would be good to know if there are any 'wish list' methods that people would like to have, and/or to know what 'real' ATCs provide in terms of manual operation.
- How important would it be to have the ATC perform some type of automatic tool touch off? It would be good to get a discussion going on how this works on machines that have it, and how it might be implemented. I saw one that appeared to use lasers.
To Ray:
- You originally mentioned that you were keeping the Geneva for the carousel in your new design, but your latest description says 'servo' powered. Are you using that term generically, as in 'motors' (i.e. steppers) or have you come over to the dark side?
- In *easily converting your new ATC to work with 30 tapers, how do you accommodate the drive dogs?
*you always say everything is 'easy'. Flying is easy for a duck. It is a bit of a challenge for a donkey. Depends on how you're equipped, I would say . . . ;)
Steve,
FWIW - On mine, I will provide a macro or modified screenset to "assist" the user in loading tools. It will sequence through the slots, bringing the transfer arm to the forward position so the user can put the tool into it, then it will put it in the carousel. On my current one, I added buttons to rotate and home the carousel to facilitate manual loading/unloading of tools.
I also plan to have a macro that will automatically touch off all loaded tools (I can tell which slots are occupied, and which are not), and load the tool table automatically. I think that will be a nice convenience, and time-saver.
I kept the Geneva, though I switched from an external Geneva to an internal, for a much smoother motion profile. It is driven by a real servo motor, with encoder, mostly just because it was convenient, and only a very few $ more expensive (<$5 more) than the plain gearmotor I used on the first one.
When/If I do 30-taper, the tool "forks" on the carousel and transfer arm will have locating tabs for the drive dogs, to maintain proper orientation. The spindle will be either servo-driven, or I will provide some simple indexing mechanism. Indexing is already designed into the PDB for the target machines with 3-phase spindles.
The one rule that I have used my entire career is "If it's complicated, you haven't done it right". I never start building something until I've gone through enough iterations of the design that I am convinced it is about as simple as it can be. This approach has *always* paid off, and has allowed me to consistently do things in short order others told me I'd never be able to do at all. I spend much more time thinking about designs, and much less time actually building them. And, in most cases, they work almost perfectly on the first try. The more common approach I've seen far too often is to rush into construction, then spend months putting on "Band-Aids" for all the things that were not well thought out up-front, and you end up with something more complex, and expensive, than it needed to be. ANYTHING can be boiled down to very simple pieces if you think about it long enough. Just yesterday I figured out how to eliminate one whole motor from the new ATC design, with no loss of functionality or performance.
A few of my favorite quotes:
"Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." Antoine de Saint-Exupery (1900 - 1944)
Everything should be made as simple as possible, but not simpler !
--Albert Einstein
Any intelligent fool can make things bigger and more complex... It takes a touch of genius --- and a lot of courage to move in the opposite direction.
--Albert Einstein
Regards,
Ray L.
-
If you decide to go with the BT30 be careful with your fork design. I made mine based off of Lyndex un balanced tools. When I purchased my tools from Mari they have balance holes in the fork groove. I had to rework the fork tips to slide over the balance drill points.
Derek
-
Steve not all CNC bridgeports are rigid head.
(;-)TP
-
Steve not all CNC bridgeports are rigid head.
(;-)TP
I'm aware. But it seems to me that the machines with fixed heads were designed from the git-go to be CNC and the others are in the same population as the masses of other 'converted' manual knee mills.
In my view, a typical manual knee mill is just fundamentally the wrong arrangement for CNC. The fact that you can buy lots of differnet ones in CNC trim doesn't mean it is a good idea. So the question remains, for those mils that HAVE rotatable, nodable, or otherwise adjustable heads, how much are these features actually used on a CNC mill . . if at all.
-
Steve,
I've never once had the need to rotate the head on my knee mill, and I really wish it was rigid. I would expect anything you can do with a tilting head, you can do just as well by fixturing the workpiece differently, or using a 4th axis.
I agree completely about CNC-ing knee mills. If I had it to do over again, it's not where I would start - too many compromises. My new machine is a bed mill. I think most of the reason there are so many CNC'd knee mills is just that there are so many knee mills out there, and often for very cheap. Bridgeport probably just took the path of least resistance on their early machines.
Regards,
Ray L.
-
I also plan to have a macro that will automatically touch off all loaded tools (I can tell which slots are occupied, and which are not), and load the tool table automatically. I think that will be a nice convenience, and time-saver.
This was my conclusion also, but it does not seem so easy to do, at least on BT30 where there are holders with very long 'noses' that could then have very long cutters in them. Conceivably the delta could be 6 or 8 inches from a short holder+short tool to an extended holder+long tool. Seems like moving a touch-off pad (or laser) with a LOT of travel would be the only way to do it.
I spend much more time thinking about designs, and much less time actually building them. And, in most cases, they work almost perfectly on the first try.
Well, my observation is that you built an entire ATC over quite a long period of time with prototype components that were later replaced with final design parts. Now you have designed an altogether different one which, as described, should be quite an improvement on the first. That process seems pretty normal to me . . . . . Just sayin' ;)
The more common approach I've seen far too often is to rush into construction, then spend months putting on "Band-Aids" for all the things that were not well thought out up-front, and you end up with something more complex, and expensive, than it needed to be.
Seems you have discovered my secret method. Except that after about three band-aids, I scrap the whole idea and start over . . :D
Favorite saying;
"Trip of thousand miles begin with one step"
- Asian probably - don't know who - circa: long time ago
"Evrybody is incopetent, just at differnent things"
- Will Rogers maybe? cira: some time before he died
"I cut it off three times and it is STILL too short"
- my mentor . . . . explains a lot, huh :)
-
My new machine is a bed mill.
OK, I can't take it any more!
WHAT machine are you getting? Tell me or kill me . . . either way . . stop the pain, . . . please.
-
I also plan to have a macro that will automatically touch off all loaded tools (I can tell which slots are occupied, and which are not), and load the tool table automatically. I think that will be a nice convenience, and time-saver.
This was my conclusion also, but it does not seem so easy to do, at least on BT30 where there are holders with very long 'noses' that could then have very long cutters in them. Conceivably the delta could be 6 or 8 inches from a short holder+short tool to an extended holder+long tool. Seems like moving a touch-off pad (or laser) with a LOT of travel would be the only way to do it.
I spend much more time thinking about designs, and much less time actually building them. And, in most cases, they work almost perfectly on the first try.
Well, my observation is that you built an entire ATC over quite a long period of time with prototype components that were later replaced with final design parts. Now you have designed an altogether different one which, as described, should be quite an improvement on the first. That process seems pretty normal to me . . . . . Just sayin' ;)
The more common approach I've seen far too often is to rush into construction, then spend months putting on "Band-Aids" for all the things that were not well thought out up-front, and you end up with something more complex, and expensive, than it needed to be.
Seems you have discovered my secret method. Except that after about three band-aids, I scrap the whole idea and start over . . :D
Favorite saying;
"Trip of thousand miles begin with one step"
- Asian probably - don't know who - circa: long time ago
"Evrybody is incopetent, just at differnent things"
- Will Rogers maybe? cira: some time before he died
"I cut it off three times and it is STILL too short"
- my mentor . . . . explains a lot, huh :)
Steve,
My approach to the touch-off macro is to have a "compliant" touch-plate - one you can approach at high speed without risk of damage to the tool. So, you do an initial high-speed probe, to figure out the rough length of the tool. Then, back off a short distance, and do a second, slow, probe to measure the length.
If I'd had an uninterrupted period of time to do the ATC, the total time would've been down to a few weeks. The parts I re-made were more to refine dimensions that I could not measure accurately, rather than fundamental design changes. The initial design is VERY close to as first drawn. The new design has a completely different set of constraints, and the MAJOR advantage of being aimed at production, rather than a one-off. That makes many things practical that were not in a one-off. The basic carousel design, BTW, is nearly identical, except for the number of tools, reducing the maximum allowable tool diameter (both per the customers request), and the change to internal Geneva (this helped make it more compact). So, a lot of the initial design has been carried over. The major change was making the carousel fixed, and adding the transfer arm, which both helped make the whole system more easily adapted to other machines - an important consideration for a "product" I hope to make lotsa money from. This added complexity, but n0ot much. This is where all the major design changes are.
Regards,
Ray L.
-
Ray,
Are you going to spill the beans about your new milling machine voluntarily, or do I have to send Carmine over to see you? >:(
-
Steve,
No secret - I ordered a Novakon Torus Pro Servo bed mill - 25"X x 15"Y x 12"Z travels, 2HP 3-Phase, AC servos, 500 IPM Rapids, ground ballscrews, linear rails on Z. Best of all, it's EASY to fully enclose, so first project will be fully enclosing it with 8020 and Plexi, and finally get to use the high-pressure (60 PSI) coolant system I've had sitting in the corner since I built it last year - WAY too messy on the knee mill. I will, very shortly, rip out the BOB and 86 Mach3, and put a KFlop in it so I can run my CNC controller app. This is the biggest machine I could find at a reasonable price, and the guys I've talked to that have them are VERY happy. It seems to be quite rigid, and very accurate. I hope to also add linear scales, fed into the KFlop, to get a bit more accuracy out of it by compensating for thermal errors.
Regards,
Ray L.
-
Nice looking machine! Had not seen this guy before.
The Kflop is here and I *might* have some minutes to play with it today. Initially, I just want to get it running with MACH and get rid of the smoothstepper, but I had in mind to run the ATC on it, or to it, or around it or whatever. But now Arduino has just come out with a new board with a much higher end processor that is almost a direct swap out for the MEGA board that I use in the InTurn™ motor controller, so I may go with that. The new processor is also Atmel and it similar to the processors on Atmel's industrial development boards, of which I have a couple. Up from 16 bit to 32 bit and up from 16mhz to 84mhz, DMA, native USB, blah blah blah. Couple comparos I've seen show 4x to 5x improvement. Just available in the last couple months. I have one on its way and I'll probably put my balancer software on it first, then the InTurn™ controller with the ATC code. Calculating the command params for the Digital Signal Synthesizer requires floating point math which really bogs down the MEGA.
An IH bench mill is being shipped to me today. It gets a conversion with ground ball screws and the new head along with ball slides on the Z. Seems like a pretty nice machine. Owners like them. Very difficult to actually get one though, especially a manual version.
I wanted to mention that the new head will NOT rotate. Feedback I got here and elsewhere was all the same; nobody uses the rotating head and several people said it was actually a problem occasionally. Most likely the mount will be 12" x 12" x 1.25" surface ground plate with the slide bearings bolted to the back and the head to the front.
Finally, I have some data on a filled column vs cast iron, so I may be able to compare the two and satisfy myself that a steel tube full of epoxy or concrete is equal to (or better by some accounts) than grey cast iron. 'Course, now I have to start all over and build another head . . .
-
Steve,
Feel free to call or e-mail if you need any help with the KFlop. I'd also be happy to send you all my code as a starting point/educational tool. I"m sure there's a lot there you could use, with straight-forward modifications.
This really is the golden age of cheap, powerful little MCU boards, isn't it? Seems like every couple of weeks there's a new board that's better, stronger, faster, cheaper than the best of the previous week. My PDB and toolchanger will be built around a lowly Arduino ProMini. At about $6 each, it's hard to pass up. I've been wanting to try the Propellor. Not so much because I need 8 processors on a chip, but I'm a multi-processor guy from way back - I was doing HyperCube research at CalTech/JPL back in the '80s, and helped build what was, briefly, the worlds fastest computer, the CalTech/JPL Mark III Hypercube.
I think you've made the right choice re: rotating the head. But how are you going to drive the Z axis if you fill the column? Where will the ballscrew go?
Regards,
Ray L.
-
I think you've made the right choice re: rotating the head. But how are you going to drive the Z axis if you fill the column? Where will the ballscrew go?
I am not a fan of slotted columns. As an engineer, I'm sure you know what it does to the torsional rigidity of the structure.
When you have ball slides, the overall depth of the slide is (generally speaking) the space you have between the back of the head and the front of the column. The screw is in that space.
On this IH machine I will be mutilating, until I have my hands on it in a week or so, I won't know what is doable, but I can tell you for certain, the screw will not be inside the column, so if the owner wants to fill the column with something, he will have that option. At this moment in time, it looks like it will be treated to a Mitsu 350 series drive and motor for the spindle so filling the column might be a requirement in order to keep the head in the same county as the rest of the machine :o
-
Feel free to call or e-mail if you need any help with the KFlop. I'd also be happy to send you all my code as a starting point/educational tool. I"m sure there's a lot there you could use, with straight-forward modifications.
THX, Ray. Much appreciated, really.
I have the Kflop working with MACH . . well . . . on my desk anyway. Took a bit to figure out how to get it to spit out a TTL step, but it's clicking along now. I though it had differential signals for some reason, but it does not seem to. No matter, I'll just throw line drivers on the outputs. There are some odd quirks that are probably caused by operator error :-[ but I'll send you an e-mail about that. Not really germane to this thread.
Surprisingly, it seems to get along OK with 64bit Windoze 7 Pro . . . so far. It is going to live on XP and Win7 32 bit, but for now it's in my office.
-
Progress is creeping along on finishing up the ATC. The last tasks are to put the sensors on the various parts that I want to monitor. Sensing the vertical position of
the arm proved to be a vexing problem, but the solution to that dilemma also provided improvements on some other aspects of the actuator that I was not completely satisfied with.
The photo below shows the new cylinder mount. Previously the cylinder mounted to the head on a heavy bracket. This required very close alignment to prevent the shaft from binding as is slides thru the gearbox. This new arrangement eliminates that separate mount by attaching directly to the gearbox. This not only guarantees alignment, it also provided for a single attachment point to the head and completely encloses the shaft. The cylinder has a magnetic piston and the position is readable by external adjustable sensors as shown in the photo. There will be a second sensor at the top of the cylinder to read the 'UP' condition. The shaft below the gearbox will be covered by a corrugated rubber bellows.
The cylinder pictured was intended to be only temporary and for feasibility and testing, but the performance seems perfectly adequate, so at this point I am inclined to leave it in place.
There was a question earlier about the shaft 'slamming' at the end of its travel. This cylinder has internal rubber bumpers at each end of the stroke that cushion the sudden stop. The air inlet and outlet fittings on the cylinder have needle valve adjustment for flow, so the speed (although not the force) can be adjusted.
I am very well pleased with this arrangement and will be duplicating the magnetic sensing method on the pod tool release cylinder as well. It is an expensive solution, but one that will prove reliable in a difficult environment.
There remains only one interlock and associated sensors for the tool pod release function. . . . and make 4 more tool pods for the carousel. Most likely I will use a solenoid actuator and optical sensor in this case.
I will post a picture of that last interlock when it is finished and that will complete the mechanics of the ATC, leaving the most complicated task of the whole project; the control software . . which I would say is about half done. I'll make another video of the testing once these last tasks are completed . . no ETA on that.
(http://www.thecubestudio.com/BT30SpindleProject/ATCSwingArmActuatorCommentedWEB.jpg)
-
Thought I'd toss out a "teaser" video of my new teeny, tiny power drawbar. Works like a charm - capable of >30 ft-lbs torque on the drawbar. Servo-controlled.
http://www.youtube.com/watch?v=QkkYabNx3Kg
In the video, it's mounted on my knee mill, for testing purposes. It will very soon go on my new CNC bed mill, which I just received on Thursday. The bed mill version includes a built-in spindle lock, and spindle sensor, so the PDB can't be engaged if the spindle is powered.
Regards,
Ray L.
-
I have been working on the control side of things so there is not much to report on the hardware. As I mentioned earlier, the control scheme is really the more difficult part of the ATC project. The prototypes in this thread are being incorporated into an all new CNC converted IH milling machine and the controls will have the ATC fully integrated. I will only be posting here the portions of that project that are directly related to the BT30 spindle and the ATC.
To that end, I have just completed the computer, motion control setup and interfacing for the project. On Ray's recommendation, I went with the K-flop board over the smoothstepper. While the Kflop board is far from the 'plug and play' that I was hoping for, eventually (with very prompt, accurate and well targeted support from Dynomotion on their Yahoo group) I did get it going. A major caveat with the K-flop is that 'most' of it is 3.3V only. I got caught with my pants down on this because I only read the K-flop spec to the point where it said '5V tolerant', however, when I went to integrate the board and read the entire spec, I discovered that only a few of the pins are 5V tolerant and the rest are 3.3V ONLY, which was bad enough, but they also are limited to a tiny 10mA current. I now have 7 different boards and many of them had to be redesigned to work with the K-flop, and I had to make up a special batch of boards for this project . . a rather large task that I had not anticipated. :'(
Although there is essentially zero cutting time on the system, so far it has not exhibited the USB smoothstepper's bad manners, so I am encouraged that it will be a good solution for the initial MACH3 support as well as hosting some or all of the ATC control in the future. The K-flop is running a total of 7 axis; 3 linear, 3 rotary and the servo powered spindle.
At this time, my InTurn™ 4th axis motor controller is hosting the ATC controls. The motor controller has a multi task board (swapaxis, digital signal synthesizer, line driver) that is mounted in the CNC computer so for the sake of simplifying the external cabling, I ran all of the controllers cables to the CNC computer. From that central point, all of the cabling runs to two large enclosures which house the EIGHT Mitsubishi AC servo drives and a bank of 12 relays. From the enclosures, one set of cables goes to the head and include the spindle servo cables, the ATC cables and all sensor cables. The second enclosure has a set of cables that run to the mill base and contain the X,Y, and Z axis, the InTurn™ 4th axis, 5th axis servo cabling as well as relay and sensor cables.
The first step in the overall task is now completed. I fabricated an aluminium 'mezzanine' plate to mount the motion control board and all of the supporting interface boards. This is mounted in a large server case with the dozens of wires collected into SIX connectors on the back of the case. In the planned arrangement, the computer will be on or near the floor with approx 6 feet and 10 feet cable runs to the two wall mounted drive enclosures. Cable runs from the enclosures to the mill will be 8 to 10 feet. All of the servo cables are Mitsubishi factory made parts. Pneumatic valves and mechanical solenoids run at 12V and the optical sensors and limit switches run at 24V to help with noise immunity.
I am very happy to have the electronics portion (which I do not enjoy) completed so that I can now move on the the mechanics of the conversion (which is the fun stuff).
This is the mezzanine with all of the goodies installed and wired up, but without the external connectors
(http://www.theCUBEstudio.com/theROCKproject/theROCKMezzInstalledNoConnectorsWEB.jpg)
Next is the completed setup final install and running
(http://www.theCUBEstudio.com/theROCKproject/theROCKMezzRunningWEB.jpg)
This is the back of the server case showing the 6 DB25 connectors needed to get all of the wires where they are going.
(http://www.theCUBEstudio.com/theROCKproject/theROCKCaseConnectorsWEB.jpg)
-
Steve,
Looks good, but needs more wires! :-) You know Craig has his Bobs ready? I got my two a few days ago - they look really nice.
What connectors did you use to connect to the KFlop? I've never seen any like that.
Regards,
Ray L.
-
Steve,
Looks good, but needs more wires! :-) You know Craig has his Bobs ready? I got my two a few days ago - they look really nice.
I was already pretty far along with the wiring diagram by the time the BOB was ready. I had made a career out of this task already and was well into modifying the interface boards (to not instantly fry the K-flop's 3.3V 10mA pins) so the last thing I wanted was to rewind to add a new piece to the puzzle.
Craig had to make some decisions about I vs O and I would need to really compare his choices to my needs and see if there was enough of each flavor. On this project, there is a LOT of I/O and having a bunch of them pre-defined *might* present a problem, but I'll peek at that issue for the next project. I also need differential signals and relays that can drive coils and 24V isolated stuff and on and on blah blah blah, so if I decide to stay with the K-flop, most likely I will make my own BOB with all that stuff on it already. The incomprehensible tangle of individual wires is definitely a huge attraction, but alas we must occasionally yield to practicality ::)
Certainly the BOB would be a big help to anybody doing a *typical* MACH install with 3 axis of single ended signals.
What connectors did you use to connect to the KFlop? I've never seen any like that.
The connectors are just the mating parts for what is on the K-flop. I chose to use crimp pins for flexibility (read: fixing screw-ups). I prefer to use color codes and heavier wire than is available in ribbon.
-
Steve,
I've never seen IDC connectors that easily allowed using individual wires - just ribbon cable. Where did you find those?
Regards,
Ray L.
-
Sure has been quiet here for a long time....
Not sure if you guts are up to speed on what I've been doing. I bought a Novakon Torus Pro mill a few months back, and LOVE it! And, I'm designing a PDB and ATC to fit all the Novakon machines. The PDB is a few weeks from beta test, and I hope to get the ATC there by end of the year. The PDB is a vastly improved version of my motorized drawbar, with a dedicated microcontroller to control the (rather complex) sequencing. Everything fits very nicely inside the sheet-metal cover on the head, with the Grab/Release buttons and an LCD display mounted to the door. This version has sensors so the firmware can verify nearly all operations, and it displays an error message on the LCD if anything goes wrong. Like my other PDBs, this one also works equally well with TTS or straighy R8 tools. Average time to grab or release a TTS tool is under two seconds. It torques the drawbar to a very consistent 30 ft-lbs, so TTS pull-out is a thing of the past
Below is the final "production prototype" mounted on my machine.
Regards,
Ray L.
-
Typically I do not do 'conversions' but I have been busy building a CNC mill using the castings from a new IH mill for one of my 4th axis buyers who also purchased the second prototype BT30 spindle, Drawbar actuator and the prototype ATC. This machine is 5 axis and has 8 servo motors.
One of the reasons I agreed to do this project was that it seemed feasible to come out the other side with a CNC retrofit 'kit' for this largest of the 'off-shore' square column bench mills. However, shoe-horning large precision ground pre-loaded (large ball nuts) ball screws into this machine (even though it it the largest of its type) required so much machining of the castings that it became obvious that this is not a candidate for a DIY project. The IH column is open in the front so it had to be boxed in and linear guides were used in place of the dovetails.
I committed to this project so it will be completed, and it is an excellent platform for the BT30 spindle and ATC, but there will be no retrofit kit for the IH mill and this will be a one-time deal.
There have been no postings here because the progress so far did not involve the BT30 spindle, drawbar actuator or ATC, but that's about to change as the head is now completed and the BT30 spindle detailed in this thread (the second one actually) is now running. The owner chose a huge spindle motor for this machine which caused a lot of problems and required a lot of re-design of the head and drive components. That slowed progress considerably, but it's finally all completed and running.
Here are the videos covering progress on this project, axis by axis. The next video will cover the all new head which contains the BT30 spindle, drawbar actuator and ATC. The spindle motor is a 3,500 watt Mitsubishi AC servo running on three phase power. This next video should be up in a few days. The project wraps up in a few weeks when the owner comes to collect his new 5 axis toy.
http://www.youtube.com/watch?v=20l-HVgJtcc (http://www.youtube.com/watch?v=20l-HVgJtcc)
http://www.youtube.com/watch?v=_sH8nBUxIV8 (http://www.youtube.com/watch?v=_sH8nBUxIV8)
http://www.youtube.com/watch?v=b812wagyBW0 (http://www.youtube.com/watch?v=b812wagyBW0)
-
Steve,
Impressive, as usual. Funny how 500IPM doesn't *look* as fast as it sounds. Though it certainly seems fast when you realize your expensive new endmill is heading straight for that fixture bolt....
I hope to finally be starting on my new ATC build in the next few weeks. It's been designed since about January, but I just haven't had the time to actually build it.
Regards,
Ray L.
-
Here is the head in its current condition. Actuator is not on the head in this shot. The motor pulleys are large diameter because if they were smaller, the tension would exceed the belt spec . . even with 1" wide belts. That's a Big A$$ Motor.
Dual speed ranges are 1:1 and 2:1 MAX spindle continuous is 6,000 RPM with 'instant' speed of 7,000PM.
(http://www.thecubestudio.com/theROCKproject/HeadComplete02WEB.jpg)
-
Nice Job Steve. Have you considered using weldments for the frame instead of buying castings. That way you can have FAST linear ways to build from and design in plenty of room for LARGE screws. Very good idea on the large screws.
(;-)TP
-
You are forgetting that the original intent was to come away with a retrofit kit for the IH mill. That would be difficult to accompish without using an IH mill as the prototype.
In any case, cast iron has between 6 and 10 times the vibration damping of steel, depending on the type of iron. I will be using weldments on my own mill, but my first choice would be cast iron. The welded up parts will be taken to a shop here in Dallas that will stress relieve the parts and then grind the mounting surfaces for the guides.
-
Very nice Simpson36! I really enjoyed the videos.
Brett
-
After a very long stint doing a mill conversion, ironically, one of the very last tasks is completing the ATC.
The positive upward stop on the drawbar was provided by the gripper 's ball cage contacting the bottom of the Belleville spring stack. Eliminating the bellevilles left an empty tube which allowed the ball cage to move higher in the spindle IF there was no tool holder installed (the toolholder itself is typically the drawbar's upward motion stop).
So with no toolholder in the spindle, the drawbar (which rotates with the spindle) would be pressing against the actuator plunger (which does not rotate) with quite a lot of force. If the spindle was inadvertently started at that time . . well . . not good news.
An interlock would remove the possibility of doing damage by preventing the spindle from running with no toolholder in place . . . but there no available place to put such a sensor . . except perhaps a laser looking across the spindle nose. The solution was to take the spindle apart and replace the empty Belleville tube with a solid steel sleeve that would provide the same stop that the belleville stack was providing.
Below are shots of the sleeve and it's relationship to the drawbar. The tapped hole in the top of the sleeve is for a pull stud to get the part out of the spindle if and as needed. disassembly.
(http://www.thecubestudio.com/BT30SpindleProject/DrawbarStopSleeveWEB.jpg)
And finally;
(http://www.thecubestudio.com/BT30SpindleProject/DrawbarAndStopsleeveWEB.jpg)
-
I'll talk a bit on the ATC control scheme and then post a link to a video fragment showing the arm moving a toolholder from the spindle to the carousel and back. The video is not a mock up for testing this time around. It is the actual milling machine that will be compleded and shipped very soon.
Earlier in this thread there was a lot of discussion on the control scheme and programming for the ATC. I started the programming using my InTurn™ 4th axis controller as a serrogate since it already was set up to read sensors, generate a step/dir pulse stream, talk to MACH over ModBuss and the processor could easily handle the tool changing sequences. Connceptually a great idea, but in reality the added ATC code was making the software very much more complicated with a net result of making the InTurn™ part of the code far more difficult to document. maintain and upgrade.
I then made the decisiotn to move the ATC to its own processor, but that introduced a new can of worms in the form of having multiple devices communicating with MACH over mod buss. Getting one device to bi-com with MACH is challenging enough.
The solution, hopefully, is a Kflop motion controller from Dynomotion. The plan is to move the code to the Kflop for a very tightly integrated scheme which relies on the Kflop''s com with MACH thru the plug-in. This was 'supposed' to be a piece of cake since the Kflop was 'plug and play' with MACH and all I would need to do is port the existing C code for the ATC over to the Kflop and then finish it up. Well . . . . I suppose some people would say that catching a Yetti is no problem, all you need is a big enough cage and some Purina Yetti Chow to lure the beast and badda bing . . . Yetti Stew. The reality there is also probably just a little different.
After a LOT of work, I have these guys working with MACH including the homing programs. Homing the spindle is a requisite for a tool changer that deals with any toolholder that has drive dogs. (i.e. pretty much all industrial tapers). The drive dogs MUST be lined up, and lined up accurately, or the toolholder is going to refuse to go into the spindle. The handling mechanism ( a 'claw' in my design), has to accomodate the dogs at the spindle nose.
The last mod I had to make to the ATC was completed just yesterday and it was to add a 'tooth' to the tool storage 'pod' to keep the drive dogs aligned. During testing, the BT30 toolholders would occationally refuse to go into the spindle. It took some time to discover the cause; after delivering the toolholder to the storage pod on the carousel, the claw simply retracts to its 'park' position. As the spring loaded claw pulls off of the toolholder, it sometimes rotated the tooldoler very slightly in the pod. The next time that tool is grabbed, the dogs are out of alignment and there is no joy when it arrives at the spindle nose. A 'tooth' on the pod keeps the toolholders aligned and that problem is history.
Finally I am down to doing the final programming for the ATC on the Kflop. This mill already uses MACH's XYZABC and Spindle axis, but the Kflop has two more axis available and those are what will control the ATC's servo driven arm and carousel.
So here is a brief video showing the ATC moving a toolholder from the spindle to the carousel and back.
www.thecubestudio.com/ATC/ATCtest-480p.mp4 (http://www.thecubestudio.com/ATC/ATCtest-480p.mp4)
-
Steve,
The video is cool! I'm almost to that point with my new ATC for all the Novakon machines - I've had the hardware built for several weeks, but been too busy with other work to make time to do the firmware.
Regards,
Ray L.
-
Looking good Steve, nice smooth, fast operation just like it should be :)
Hood
-
Oh BTW my Chiron uses dogs but they are sprung and the spindle is rotating at 200rpm whilst tool is offered up and thus no need for orientation.
I think the reason they did it this way is because, of the speed of the changer, a tool can still be rotating slightly when its withdrawn from the spindle (bearing on tool holder if you recall) and thus orientation of the holder could be out. To keep orientation they would have had to slow the toolchange sequence down a bit and actually stop the spindle. That would make no difference to me with what I do but as the Chirons were always meant for high volume work, such as the Auto Industry, I think it mattered there.
Hood
-
Steve,
The video is cool! I'm almost to that point with my new ATC for all the Novakon machines - I've had the hardware built for several weeks, but been too busy with other work to make time to do the firmware.
Regards,
Ray L.
Well, as you might agree, the software is the real challenge, especially now that I have completely lost my mind and decided to go forward with porting to the Kflop. I can tell you now that I used your name in vain more than a few times for recommending the Kflop as a 'plug and play' MACH accessory and using the word 'easy'. Note to self: When Ray uses the work 'Easy', be afraid . . be very afraid.
The HimmyKabibble vernacular exchange rates:
"Easy = Doable, like walking on water or spinning straw into gold."
"Difficult = requires intervention and assistance from advanced civilizations"
Eventually, I did get things working. It is quite a powerful system, but they really need to hire you to write some manuals like you did for MACH. I will be starting on that port in a few days so if you hear your name on the wind with colorful adjectives attached . . . that would be coming from the East, most likely . . . ;)
-
Oh BTW my Chiron uses dogs but they are sprung and the spindle is rotating at 200rpm whilst tool is offered up and thus no need for orientation.
I think the reason they did it this way is because, of the speed of the changer, a tool can still be rotating slightly when its withdrawn from the spindle (bearing on tool holder if you recall) and thus orientation of the holder could be out. To keep orientation they would have had to slow the toolchange sequence down a bit and actually stop the spindle. That would make no difference to me with what I do but as the Chirons were always meant for high volume work, such as the Auto Industry, I think it mattered there.
Hood
Interesting. When you say 'spring', do you mean the dogs can move up into pockets in the spindle, or that they move out to the side (in the manner of an old flyweight governor)?
Are there ramps or something like that on the dogs to let the spinning toolholder contact then and push them to . . wherever they go to?
That changer of yours is just an incredible thing to watch. But tool changing at 200 RPM with drive dogs? I can imagine . . with some hesitation and perhaps cringing . . yanking the toolholder at 200 RPM, but getting the toolholder IN at 200RPM just sems like all but impossible.
An impressive accomplishment to be sure. Can you describe the process in some detail?
-
Running a job on it at the moment but will take a couple of vids of it later on to let you see whats happening.
Hood
-
I just looked at the video fragment ans I probably should add a couple of comments:
The arm flexes up and down. This is inentional. The drawbar not only releases the grip, it also physically shoves the toolholder out of the spindle by a small amount. The arm flexes to accomodate that movement
The carousel is a bit loose. That will be tightened up during final assembly, but some play will remain intentionally to allow for minor misalignments.
The arm shaft will eventually be cut off flush with the bottom of the arm.
The arm shaft will have a corrogated rubber bellows covering it.
The arm will have a cover to keep swarf off the sensors.
The arm will move a little bit faster in the final version.
Edit: Oh and one last thing . . . the carousel will rotate . . . so that the machine can use more than one tool . . . :D
When I get some more pods made up, I will do another video showing an actual tool CHANGE and also the spindle homing, which is working, but not shown in the video fragment.
-
Well, as you might agree, the software is the real challenge, especially now that I have completely lost my mind and decided to go forward with porting to the Kflop. I can tell you now that I used your name in vain more than a few times for recommending the Kflop as a 'plug and play' MACH accessory and using the word 'easy'. Note to self: When Ray uses the work 'Easy', be afraid . . be very afraid.
The HimmyKabibble vernacular exchange rates:
"Easy = Doable, like walking on water or spinning straw into gold."
"Difficult = requires intervention and assistance from advanced civilizations"
Eventually, I did get things working. It is quite a powerful system, but they really need to hire you to write some manuals like you did for MACH. I will be starting on that port in a few days so if you hear your name on the wind with colorful adjectives attached . . . that would be coming from the East, most likely . . . ;)
Steve,
That explains the odd burning sensation in my ears lately! But, when did I ever say bringing up a KFlop would be "easy"? :-)
Regards,
Ray L.
-
Ok Steve, here are the vids, first shows me offering a tool up to the spindle so you can see the dog getting pushed up and then dropping in.
http://youtu.be/784P5uA9lro
As you will see the toolholders have the bearings but also a brass disc, this acts as a brake so they dont rotate in the tool arm until the dog engages. When the dog drops in it obviously turns the holder and this would soon cause friction if left in that position. However when the tool is clamped it is pulled up to clear the disc and instead the bearing is then in contact with the tool arm.
You should see all that in this vid.
http://youtu.be/fuxOCikZxU8
Hood
-
Hood,
First, THANK YOU for the vids. Thats a piece out of your busy day and it is much appeciated.
I looked at a lot of attempts at R8 tool changers before I concluded that this was not a road I wanted to go down. Likewise, when I looked at the various tapers, I saw only fixed dogs on the spindle . . and I looked at a lot of those spindles and changers.
The spring loaded dog seems like a brilliant solution. However . . . it appears that each tool holder has its own bearing/brass ring brake assembly as a permanent fearture. Is that correct?
- Steve
-
Yes thats correct, it needs the brake so that the tool will not turn with the friction of the dog pressing on the flange before it rotates round to the slot. Also it is needed so that when the tool is released at 200rpm it will stop spinning very quickly.
The bearing is needed as the tool arms are always holding the tool and if it wasnt there it would either have to be loose in the arm which would present problems with it entering the spindle nose or if the arms were made tighter then you would get rubbing and friction and it would seize in no time.
The toolchange is actually damped down on my machine, it was done by the previous owner and I decided just to leave it that way but original toolchange time was I think 0.9 seconds and something like 2 seconds chip to chip.
Hood
-
Well, I was all excited for a while there, and I will say that the system in ingenious in side stepping the alignment issue . . which is a BIG issue. However, requiring all custom tool holders is not on the menu so while I can be amazed and admire the creativity in the solution, it is not an idea that I can use as it goes contrary to one of the primary deisng goals: use standard commercial/indusrial tool holders.
I looked at some of the 'pseudo' CNC toolholders such as the stuff from Tormach and some other similar stuff, but those are really aimed at the hobby community that does light duty work, can tollerate down time and prioritizes cost over performance (for the most part) and that is not my focus, so all of those type solutions got the boot.
Quite obviously, your setup is not hobby level and is well thought out and proefessionally executed, particularly is one considers the time frame in which is was built. However, while I have to admit that I would love to have one . . because it not only performs, but the cool factor is off the charts, it is not an idea that I can run with becuase of the proprietary nature of the scheme.
Thanks for taking time to show the system and explain the operation. I found it 'fascinating' as Mr Spock would say.
-
You shouldnt need custom holders as your tool arm moves out the way and the only time you would need the spindle rotating would be on loading and thus I am sure you could incorporate a wear disc/brake of sorts into the arm. You would have the spindle stopped when unloading a tool and thus no need for the brake to stop the tool spinning when released from the spindle.
The only reason the bearing and brake are needed on the Chiron is the basket tool changer setup as the arms hold the tools at all times, even when machining, and thus the bearing is needed. Add to that the fact that the whole setup of the basket toolchanger was to make super fast tool changes and have no need to actually stop the spindle then you can see why it was required to have these custom holders.
At the time (late 80's) I think most toolchangers were talking 10 plus seconds per tool, the Chiron must have been amazing at under 1 second, even 25 years later it is still amongst the fastest around. Another big benefit that was/is touted by them is if something happens to a tool arm then no big deal as you still have all the rest you can work with until time can be found to get that repaired, on a standard style changer you are stuck until you get the changer repaired or have to resort to manual tool loading. Again probably no big deal for me or your customers but in high volume production that could save a lot of money.
When I got the Chiron two of the positions were missing and the previous owner had used it that way for 8 or 10 years as the type of work he did often only required 3 or 4 tools, so that kind of shows that Chirons talk of no downtime due to a tool arm failure holds true.
Hood
-
Ah but the million dollar question is...
Have you been hit by an arm yet ?
lol.
-
Nope, only the wifes elbow :)
Hood
-
Hood,
Thanks for the info and ideas. It will take some noodling to figure out how much of it might be incorprated into future iterations of my ATC.
What you are seeing here is REV 0 of the ATC and only the second prototype of the BT30 spindle. There is more to come. I have designs for a dual arm, a column mounted (and much larger) carousel, a link belt tool storage and a number of other goodies. I am also continuing to improve and expand other products. I have a new CNC rotary table (finally) and an all new tail stock design for the InTurn™ 4th axis (that I am building now) and so on, so I'm not inclined to discount any ideas that might be useable in any component.
- Steve
-
The ATC is now operational. Below is a quick video fragment that I grabbed before the machine ships out in a couple days. The fragment will eventually be part of a larger video covering the last stages of the mill conversion, but for now, you can get to it with the link.
This video fragment shows a tool CHANGE . . i.e. grabs and stores one tool and then selects another tool, retrieves it and puts it in the spindle. And no more 'smoke and mirrors' trick using MACH axis to drive the motors. What you see is being run entirely by the all new and completely autonomous ATC controller. This video does not show the spindle homing becuase that operation, while finished, was not linked into the tool change macro (M6). The macro now homes the spindle as the first task, then converts the tool numer to binary and presents that at MACH outputs and then signals the ATC controller that there is a tool change needed.
The ATC controller them retrieves the binary number from the MACH outputs and acknowledges to the macro that it will accomplish the tool change.
The ATC controller then operates independently from MACH, including generating the step.dir pulse stream for the drives, and at the end of the tool change, it the sends a 'Done' signal to MACH (to the waiting macro) .
It sounds a lot more simple that it actually is. There are about 1,200 lines of code so far in the all new ATC controller and it is using the newest (4 to 5 times faster) processor. The new processor is 3.3V so that eases the interface problems with the Kflop, which is also 3.3v . . . mostly.
This scheme is a bit unusual, but it does not rely on Modbus or a plug-in and it will work with any CNC that has 6 available outputs . . . i.e. not tied to MACH. My 4th axis controller is already using the modbus and I decided not to try to have them (the 4th axis and ATC controllers) communicate with each other, nor did I want to attemp multiple simultaneous modbus devices.
In this way, the BT30 spindle and ATC are not tied to the 4th axis.
Also not shown in the video fragment, but now operational, are the two primary safety interlocks. The drawbar cannot release accidentally and the claw cannot release (read 'throw') a tool. Until electric solenoids are retracted, the drawbar actuator cannot move and the claw cannot open. The speed seen in the video fragment is something like 2/3 of the full speed. The claw interlock was not installed yet and I did not want tool holders launched across the shop. But rest assured, both the arm and the carousel can run much faster than shown here.
There are quite a few sensors to integrate (hardware is completed, just a matter of adding the sensors to the software), and of course. This was a particularly difficult project, but it is just the first iteration of the ATC. I have lots of other goodies to add like a dual arm, a much larger fixed carousel and perhaps even a tool belt.
http://youtu.be/QmlAgfIbscQ (http://youtu.be/QmlAgfIbscQ)
-
Looks like its working well :)
I am not a fan of low voltage signalling and is one of the reasons I always shied away from the kFlop, that and also I am no programmer ;D but looks like its working well for you :)
Regarding the programming, thats why I tend to use PLCs for things like this, ladder logic is easy for me, code is not.
Hood
-
Question for spindle experts: Are there any specs on the various tapers which define the power or torque capacity of each size? I have been unable to locate this info.
The time has finally arrived for me to build my own mill. The IH conversion that I had been working on . . . . for a while . . . is completed and delivered and I have only a couple of projects to finish up and then I will be . . at long last . . turn my attenetion to my own project mill.
Previously the spindle and ATC was BT30. Thus far, I have built several spindles but only one ATC. The converted machine is completed and shown in this video changing tools and cutting steel and aluminum with a 3.5K spindle motor (close to 9k peak).
http://www.youtube.com/watch?v=k2F6100RKOE
For my own spindle, I am leaning toward BT40/CAT40 because my new spindle motor peaks at 15,200 watts and I think that is going to be too much for BT30. Since I have not found any specs, I started looking at the power ratings of available spindles and it seems like 5.5kw (unknown peak) or thereabout is the high end for BT30.
In addition, I have in mind to mount a tool carousel in the spindle to rotate tools toward the InTurn™ 4th axis (this will be in addition to the existing gang tool plates) and again I am feeling like BT30 might be a bit light for that application. Overall the new mill should be impressive. Built on a new cast iron frame, it will have massive ball slides and ball screws (all high precision ground and pre-loaded)on each axis with travels of 38" X, 26" Z and 14" Y. Each axis will be powered by a 750watt Mitsubishi AC servo motor and be capable of about 700IPM. ATC of course and like the previous project, it will have 5 axis capability. I am looking for repeatability and accuracy measure in tenths for this new mill.
The spindle will feature precision pair of angular contact bearings and labarynth seals with a positive pressure purge on the spindle cartidge this time around (9,000 + RPM spindle) and each ball screw will also have pairs of ABEC3 angular contact bearings. People have been asking for more detail and for some design data so I am planning to spend more time with the video camera on this build and show the actual design and fabrication of some of the parts and pieces, including the bearing blocks for the ball screws and some other interesting stuff. The IH conversion has a lot of this stuff, but I did not have time to cover any of the details in that build.
So, if anyone has some info on the capacities of the various taper sizes, please post them.
-
Looks like its working well :)
I am not a fan of low voltage signalling and is one of the reasons I always shied away from the kFlop, that and also I am no programmer ;D but looks like its working well for you :)
Regarding the programming, thats why I tend to use PLCs for things like this, ladder logic is easy for me, code is not.
Hood
I can think of a lot of reasons to shy away from the Kflop, but it is a very powerful board and given some programming experience, it can be made to do some impressive stuff. The biggest obstacle with the Kflop is lack of documentation. There is no programmers reference, etc. so the learning curve is practically infinite. I will say this for the Kflop; while it can be challenging to get it to do what you want, once you have it working, it is a rock. It just rolls on day after day without a hitch. Smoothstepper was always flakey and had 'moods', but the worst problem was that if there was an issue, it was pretty much permanent because the user had no way to fix it (regardless of programming knowledge) and the manuf was in no hurry (as in months, years) to correct anything.
To comment on 'low voltage signals' I think pretty much all modern processors are 5V or less, more recently 3.3v or less, so there is always going to be a conversion somewhere between the processor and the 'outside world' where signals need to run hotter to cope with massive intefference generators like high voltage wiring and electric motors and solenoids all over the place. Differential signals are actually low voltage if you compare them to the typical 24V industrial control signals. In any case, once you have the isolation and voltage conversions in place, it really becomes a non-issue and you don't even think about it.
There is still a lot of improvements and feature additions to be done to the ATC controller, but I went with a separate processor for the ATC. The Kflop is primarily doing the motion control duties and the programming for that device is pretty much done . . . thank goodness.
I would liken 'programming' vs 'laddering' to your lathe having tool posts both in front of and behind the spindle. I suspect such an arrangement would be mind boggling to most people, but you have gortten used to the complexity over time and have learned how to make use of the capabilities that come along with that complexity. Likely you would feel quite handicapped if you had to make due with a single toolpost after having two. Programming is the same way . . all that is needed is an inexhaustable supply of tollerance and patience . . . . and a Whisky Sour . . or two . . . :P
-
I couldn't agree more about the Kflop. Unfortunately for me I had to abandon it. Which really disappointed me as it seemed like the possibilities were huge. The C programing was just going to be too heavy of a lift with my current workload. I opted to go with the UC-300 controller which has been working really well for me. In the next couple of weeks I'm going to offload my ATC functions from the Pokeys to the UC-300. It has plenty oh IO for the job.
Derek
-
I was looking for info the last day as I had just made a spindle drawbar force gauge up and was wondering what I should be expecting to see, turns out I am well down on force if the numbers are correct, ie my gauge is right or Chiron chose to keep in line with what is said to be the norm.
Anyway not sure if this is actually what you are looking for but have attached it anyway.
Regarding the 5v, yes most if not all processors etc will be 3.3v but the difference with the CSMIO compared to the ESS/ SS/ kFlop etc is the low voltage is all contained in a nice little box and only 24v for I/O is external.
Yes there are differential signals, whether that be the encoders ins ( or step/dir outs if using IP-S) but they are inherently noise immune.
Thing is however I never used 5v for other I/O, was all 24v so that meant I had to convert when wires came out of the controller and when wires went in, was a PITA.
Regarding the SS, I never had any flakiness with even the USB ones but I do agree on Gregs lack of fixes, I waited probably 3 or 4 years for the step/dir spindle to get fixed and it never happenned and now I have moved on, just wish I had taken the jump a year or so earlier to the CSMIO as it is just in a different class.
Hood
Edit, just realised I read your spindle question wrong.
Anyway i can give you some numbers for the Chirons spindle in a bit once I find the info.
-
Ok here are some numbers
30 taper
Max RPM 6000
Cont torque 18Nm
Peak Torque 80Nm
Hood
-
Ok here are some numbers
30 taper
Max RPM 6000
Cont torque 18Nm
Peak Torque 80Nm
Hood
Are these BT30 maximums or specs on your spindle?
Numbers indicate approx 11.5kw. How does that compare with your spindle motor?
-
That was the original Siemens motor as supplied by Chiron. It works out at just over 11.4Kw.
The motor I have on now is only 3800 rpm but torque is 28.5 cont so that works out at 11.44Kw.
I am in hopes of some time managing to make a gearbox up to increase the rpm but whether I will ever get round to it I have no idea.
Hood
-
I think the challenge will be to find a planetary gear set that is compact enough and can handle that much power.
Speaking of power handling, it is encouraging to know that BT30 can handle the power that Chinon is pushing thru it. Doubling the power handling changes the landscape considerably. I would still like to find some hard engineering data on the capacities, but the Chinon specs are a good start and the drawbar tension chart you provided is priceless . . especially considering the endless debates that took place on that topic a while ago. I run the BT30s at 1,000lbs . . excuse me 1,000lb FORCE to keep the author happy . . ::) so I think that's in the ballpark.
Meanwhile, I have completed the new design for a BT40 spindle and located heavier springs (coil, not belleville) to meet the spec in your chart. I can see now where the bellevilles come into play. It looks as though I would not be able to go any larger than BT40 using an off-the shelf die spring. Good to go with BT/CAT40 though so now I have a quandry as I have already started collecting BT30 tooling.
Its always something . . . :'(
-
Yes that is indeed the problem, I stripped an old van gearbox in the hope I would find some suitable gears but all were too big to fit in the space I have.
Regarding the max power handling capabilities of 30 taper, I would take a guess that the Chiron is close to but within the safety margin of the 30 taper, it is a German machine of very high quality meant for high production (not that I use it for that ;D ) and Germans dont just take guesses, they are anal about these kinds of things ;D
I suppose there is a chance that Chiron decided they did not need the max possible with a 30 taper but as said above with it being meant as a high production machine I would imagine they have made it close to the max.
The force gauge I made up was a good and a bad thing, good thing is I now know the force my drawbar exerts, bad thing is I think it is low and I have no information on my spindle so I am really in a quandry, do I try and find the way to strip down or do I just use as is and just hope for the best ;D
Hood
-
You did not say what your drawbar force measured out to be, but FWIW, in my early research on BT30 (I did not look at any other sizes at that time), what I unearthed was a dual spec for BT30 with significantly different tension specs above and below a certain RPM.
The tension specs were 600lbs and 1,000 lbs (above a certain RPM), but I don't recall the RPM break point. I *think* it was 6,000 RPM, so you might be OK as is.
In any case, one would imagine the drawbar having a threaded collar that bears against the spring pack . If so, then it would simply be a matter ot tightening that collar down a bit. Keep in mind that tightening a collar in this arrangement is also taking out travel, so it would be prudent to make sure (after such an adjustment) that the sping stack is not 'bottoming out' (binding) before the pull stud is completely released.
-
I only have 400 lb-f . I do not have any specs on what it should be but I feel it should be higher.
See pic below, it shows almost 1200 PSI but I deliberately made my piston 1/3 inch square so that the value is trippled and thus the 1200PSI is actually only 400PSI.
The big problem I have is I do not have any drawings available to me to even see how things work. I do not even know if it uses bellevilles or some other means and I do not even know how to take the spindle out to see. In fact taking the spindle out itself will not even show me as everything seems to be internal in the spindle and dismantling that seems to be quite involved.
I have found a few bits and pieces of info regarding spindles in Chirons but they are different models and from different years and all a bit different from one another so not even sure if they are of help.
I know it is air operated as two pipes go to the front of the head casting but other than that I am in the dark. I am in Hopes that Graham Waterworth may manage to get me some info as he has asked a friend who is a CNC Tech and he in turn is going to ask around to see if anyone can tell me. Failing that I will just have to go in blind and hope I dont screw anything up, spindle rebuilds I think will be rather expensive ;D
Hood
-
I only have 400 lb-f . I do not have any specs on what it should be but I feel it should be higher.
See pic below, it shows almost 1200 PSI but I deliberately made my piston 1/3 inch square so that the value is trippled and thus the 1200PSI is actually only 400PSI.
Hood
I do not understand the math. 1/3" square piston is .11 SqIn.
-
Words were arse for tit, should have been square inch not inch square :)
I have just tested the gauge by removing the pullstud and taper section and passing the bar through a hole bored into a lump of steel, I then screwed in some threaded rod with a ring welded to the bottom. I suspended from the press's table and stood in the ring. Gauge showed approx 600 to 650 PSI (was awkward to view when standing on it and looking at it upside down ;) ) I weigh 95 Kg so that is about right, 95 x 2.2 x 3 = 627Lbs so 627 PSI on gauge.
Hood
-
Hood you have to be carefull not to take a big bite at a time You now have more Torque at a lower rpm. You may be on the edge of a 30 taper capacity to be able to hold the tool holder rigid. You will have a LOT more torque to take a bigger bite. Yes it still has the same KW but a very different torque band.
If your guage is correct and it appears to be, you are VERY LOW on pressure for the drawbar. We always used 1000psi for the no go on a 30 taper below that and heavy cuts could cause problems. Bet you have some broken spring washers in there. Best to fix it rather than worry about it "What IF".
Just a thought, (;-)TP
-
Terry, I tend not to take big cuts but I definitely would like to get the drawbar force up to where I think it should be. I am in hopes that I can get some info on how things are constructed and put together but if not then I suppose I will just have to pick a quiet spell and then get stuck in and hope I dont do more harm than good ;D
Hood
-
A scary possibility is that the gripper is unthreading itself from the end of the drawbar (assuming it is threaded on).
If that were the case, then it *might* be possible to screw the gripper back onto the drawbar using a conically shaped tool thru the spindle nose and into the gripper. A 'pinecone' shaped carbide burr with a long (6") shaft (as used in porting work on automotive engines) would do nicely. The gripper should be hardened (either finger or ball type) and if you turned the tool shaft with your fingers, no damage will occur and you might be able to feel whether the gripper was turning or not. Likely you would need to try with the drawbar set On and OFF and ideally somewhere in between if that is doable on your setup. Objective here is to assure that the gripper is not being used as a stop (or in some other way being forced against a surface internally) which would prevent it from turning.
In case you do find the gripper loose and are able to get it screwed back on, a 'Q'-tip saturated in GREEN locktite and touched to the end of the drawbar should get some of the stuff wicked up into the threads. I would place this stuff somewhere between blue and red in holding power.
In any case, good luck with the spindle. My bet would be that (if the spindle is a 'cartidge' arrangement) you can get to the drawbar and spring stack without disturbing the spindle bearings.
And you really don't want to disturb those guys. Here's why;
I have a precision option for the InTurn™ 4th axis now and my next mill spindle will have precision bearigns as well, so I have been spent a lot of time studying spindle designs, perusing white papers on vibration analysis and reading the precision bearing specs and handling proceedures as published by SKF, NTN, TIMKEN, and so on. Typically they say 'In the event you have to remove a bearing, it is best to replace it'. ??? $300 - $400 bearing . . . times 2 . . . JUST replace it? Yikes! :o
Oh yes . . and . . . the replacement bearings are slightly smaller than the OEM, so the spindle will need to be 'reground to achieve the correct fit.' Gee, suddenly the bearing cost seems almost trivial. . . :'(
How's this for a tollerance (EOM bearing supplier for Excello) Quote: ' .0001" tight to .0001" loose ' . . . . bizzarre. Staying within two tenths is doable for me . . barely . . but it takes a lot of time because I have to 'sneak up on it' as my machinist buddies used to say. Here is my special one-off hardened A series tool steel spindle. It is currently cut for a standard press fit to a ABEC3 deep grove ball, so there is meat on there to re-cut it for precision bearings. I just need to grab a couple while I'm out getting bread and milk . . . . .
(http://www.thecubestudio.com/4thAxisSpindleProject/Spindle_InTurn_GroundWEB.jpg)
-
I really think it s more likely the springs are broken or weak as I do know they have never been replaced in at least the last 10 or so years.
Dont really want to have to dismantle bearings etc but I have found a PDF from McBroom Industrial Services that mentions removing bearings before removing drawbar but whether it needs to be done in that order I wont know until I either strip it down or hopefully manage to get a part drawing. Also dont even know if my spindle is like the one in the PDF or not.
Hood
-
I really think it s more likely the springs are broken or weak as I do know they have never been replaced in at least the last 10 or so years.
Bellevilles loose tension over time (sometimes a failrly short time) that is a fact and is one of the disadvantages that I was looking to get away from with the coil. I agree with you and BR that most likely you will need a new stack, but you won't know that untill you get inside. If the gripper is coming off, you can check that easily from the outside. It would be sort of a bitch if you tore the whole thing apart and then found out the gripper had unthreaded.
Dont really want to have to dismantle bearings etc but I have found a PDF from McBroom Industrial Services that mentions removing bearings . . . .
After reading the scary tech stuff from the bearing manuf, I made sure that there is room behind the bearings to get a puller in there and reach the inside ring. Some AC bearings can be separated, but if the OD of the inside ring is sitting against a larger OD, you still can't get a puller on it, so you are left with heat/cold as the remaining option for getting it off undamaged.
Please post back here if you do dig into that spindle so that I will get notified. My new spindle project starts in a few weeks, but I may take it to a bigger forum since there is not very much participation here. In any case, I would want to follow your 're-springing' effort.
-
Please post back here if you do dig into that spindle so that I will get notified. My new spindle project starts in a few weeks, but I may take it to a bigger forum since there is not very much participation here. In any case, I would want to follow your 're-springing' effort.
Sorry Steve, forgot all about you asking this and was only reminded when someone PM'd me asking if I had ever done the springs.
I got hold of some drawings from a German guy that does retrofits ( http://www.cnc-steuerung.com/umbau-einer-chiron-frasmaschine.html ) I had seen he had done a retro on an FZ17 and asked if he had the manuals, he didnt but very kindly drove to the person that owned it and took some photos of the relevant pages. (I have since also recieved some scans of a manual from a guy in South Africa who has a machine that is very close in build number to mine)
Anyway, although it was not identical it was very close including the disc spring sizes and amount, so I was able to see how to haul things apart. It was actually very simple, remove 6 M6 capscrews from the nose and pull out the spindle, pressing the drawbar release more or less pushed it out for me.
I replaced the springs and it was better but still only about 600lb if I recall, this was still way lower than I was expecting but was actually the force expected with the size and amount of springs, so something was strange. I then studied the drawings a bit more and I then saw why the actual force was not too important.
The drawbar is precision made and according to the drawings has a length of 298.7mm -0mm +0.1mm and the 2 ball bearings are also precision, grade 3.
I stripped things back down again and all was in spec with the drawbar which surprised me as it is almost certainly the original, the sleeve and the the drawbar top also seemed fine but the balls had a slight amount of wear so I replaced them.
Anyway if you look at the pic below you can see that the balls effectively lock the tool in place as they are constrained by the outer sleeve and the top of the groove in the drawbar head is tight against the ball. This means that really the spring force is not an issue as all it has to be is sufficient to make sure the tool holder is pulled tight enough into the spindle for the ball to lock everything in place. It also means that the air is sufficient to operate things rather than having to go hydraulic.
Hood
-
Hello Hood.
Thanks for the update. That is quite some detective work you accomplished to dig up all of the info on your spindle.
Using balls as a 'latching' mechanism is clever indeed. I see where the precision would be needed in order to stay within the very narrow dimensional range where the mechanism would work.
I have made a half dozen or so spindles now and the die spring is working well. What does not seem to be working is keeping a spindle for my own use. So far I have been unable to accomplish that . . . :'(
-
Its always the way Steve, well thats if you are making things to sell that is :-D
Just finished a new VDI turret for my lathe but its mine, all mine :D
Hood
-
Its always the way Steve, well thats if you are making things to sell that is :-D
Appropriate season . . . you know the 12 days of Christmas song?;
6 BT30 spindles, 5 Tail Stock II, 4 Pneumatic actuators, 3 InTurn™ Ultra's, 2 ATC's and a Partridge in a pear treeeeeeeeee!
All I have is the stinking tree . . and it's dead.
The Partridge?
. . . . . tastes like chicken :D
Just finished a new VDI turret for my lathe but its mine, all mine :D
Hood
I visualize Vincent price wringing his hands and laughing maniacally. :)
However, I can not visualize what it is that you built.
What does VDI mean?
Did you do a thread or video?
Oh yeah and HAPPY NEW YEAR . . . To ALL
-
Ah chicken is not too bad a taste, can think of worse ;D
VDI is Verein Deutscher Ingenieure, thiink it translates to something like Society of German Engineers ;D
It is a means of clamping tool holders into the turret via teeth. Below is a pic of the setup showing the completed turret, the wedges are exploded on some positions to show their make up.
Have a few vids but not done a write up yet.
This vid is it almost complete although I have since altered the layout to have the gearbox on the turret rather than the motor.
https://www.youtube.com/watch?v=35p74Bf51u4
This one is me making one of the wedges.
https://www.youtube.com/watch?v=8CHhfy1CanM
There are a few others of the face and edge machining.
Hood
-
Have a few vids but not done a write up yet.
This vid is it almost complete although I have since altered the layout to have the gearbox on the turret rather than the motor.
Hood
Hi Hood,
Sorry I missed this post. I have not been on here in a few months. In here now because I'm porting my InTurn™ motor controller to MACH4. That project is completed and I made a widget and played with the screens a bit, but now time to get back to work.
Your tool turret is impressive! I was thinking about making a turret to go on the mill spindle nose to present tools to the 4th axis in this same manner . .
Also, I have upgraded the InTurn™ Tail Stock and it now slides along using dovetails on a fat piece of precision ground cast iron. It has 'precision' ACME lead screw and provision to add a motor for a power lead. The mounting is now robust enough to carry a Tool Turret . . . albeit not nearly as massive as yours.
. . . . so I have a couple of questions about your design, if you don't mind.
How are you extending/retracting the turret and how are you locking it? Can't tell from the video or the 3D model.
You mentioned moving the 'gearbox' to the Turret's frame. Are you using a planetary?
Sorry for so many questions. I'll stop here for now, but I actually have more questions.
-
I am using hydraulics to calmp/unclamp the turret, the lathe has them for the chuck and tailstock and it also originally used the hydraulics to work the turret completely by rack and pinion for rotating.
I am using a Hirth coupling on this one, it is the one that was on the original lathes turret.
Yes it is a planetary box, 25:1 if I recall, just what I had lying around.
Hood
-
I am using hydraulics to calmp/unclamp the turret, the lathe has them for the chuck and tailstock and it also originally used the hydraulics to work the turret completely by rack and pinion for rotating.
I am using a Hirth coupling on this one, it is the one that was on the original lathes turret.
There is going to be a lot of torque on that turret when you start digging in those tools with a lathe that big. Are you pulling against a cone clutch or are there wedge shaped dogs in there somewhere or is there a friction plate?
You mentioned machining wedges, but for some reason I though those were for the tool holders.
I have a power chuck yet to install and it has a big hydraulic actuator, draw tube and the chuck itself is mechanical and operates on internal wedges. The hydraulics on the chuck is only something like 600psi, surprisingly low.
Yes it is a planetary box, 25:1 if I recall, just what I had lying around.
That explains the similar sized pulleys. :) With the box on the motor, it is surprising that you would have that much bounce. Was that coming from the belt? Did moving the box to the turret calm it down?
-
It has a Hirth Coupling, see pic below for half of the Hirth, other half on the turret body.
Yes the wedges are for holding/pulling the VDI tool holders into the turret plate.
600PSI is quite normal I think, if that is the Hyd pressure to the chucking cyclinder, mine runs at 35bar so about 500PSI
With the gearbox on the motor it meant the full torque was being passed via the belt, with it now on the turret the torque through the belt is very low as it just has to turn the input of the gearbox.
-
I watched your video some more time(s) and now I can see the turret rotating very slightly as it cinches up on those teeth. If I knew what a hirth coupling was, it would have been clear from the git-go. :-[ Now that I can see all those choppers, it makes total sense. Thanks for explaining.
You mentioned that the hydraulics were from a 'previous turret'. What was the motivation to build a new one? That is a very significant project to tackle, I would say.
-
Ok kind of a long story.
Originally the lathe came with a 6 position turret at the rear and a 4 position toolpost at the front. Both were totally hydraulic so could only rotate one direction and also had to clamp at each position before unclamping and moving to the next etc etc.
I then acquired a turret from a Bullard lathe, it too was totally hydraulic but I stripped off the hydraulic motor that rotated it and replaced with the motor and gearbox I am using still.
So that meant I now had a lathe with 6 position turrets front and rear, Mach however wasn't that great with front and rear tools so I had to always edit my G Code to suit and add to that I had to be very careful with new code to make sure I hadn't commanded a move where one or other of the turrets could smack into the stock or rather the tools such as drills smack into the stock.
So next bright idea was to remove both turrets and make up a scratch built 12 position turret, it worked well for the most part but it was just a bit too cluttered, I would have preferred a 10 position turret but both the Curvic coupling from the Bullard turret and the Hirth from the original would not allow 10 positions due to the amount of teeth (48 on Hirth, 24 on Curvic) so it was 8 or 12 and I went for 12. That brought problems when using drills, boring bars and normal turning tools, they were all too close and I had to be very careful not to crash a drill or boring bar into the chuck.
So next idea was to go to an 8 position turret but make it a VDI turret so that the tools could be swapped in and out within seconds and be very repeatable, so that is where I am at now.
I also like a challenge and the VDI was defitately a bit more complex than the previous one I made ;)
I used the original turrets Hirth coupling and also the body/cylinder but I had to do quite a lot of modifications including adding thrust bearings internally and I likely would have been quicker and easier making one up from scratch.
Pics below of
1. Original setup (seen from back side of lathe)
2. Bullard turret after I modified it
3. Scratch built turret
4. Latest, and likely the last, turret.
-
Now, that is quite a story. Well done. Very well done.
You may have been able to get 10 tools by having 6 on one 'half' (first 180 degrees) and 4 on the other and still used the 48 tooth coupling, however, The software to run such an arrangement would be complex and error prone and probably require a sensor or two to operate safely. Seems to me that 8 equally spaced tools with quick change holders is the bulletproof solution.
Just one more (compound) question if I may; presumably the turret is supported by a draw tube or shaft. Dies that shaft ride on bearings in the housing or on the hydraulic seals? It is a hardened piece from the original machine or did you design/build that part? Lastly, I note that the pulley does not shift in and out with the Turret. Is the shaft splined, keyed or slotted or is there a special slip joint and again was this recovered from the original machine or new design?
For a minute there, I thought you had posted a photo of a different machine, but it appears you built a full cover and painted it to match the machine. First class. Very nice!
OK, last, last question; there appear to be some kind of trapezoid shaped 'posts' sticking out of the turret surface that the tools attach to. Also I noted holes in the turret edge that look like there are screws or cams of some kind. Is this the quick change mechanism? I can't tell from the photo how it works, but unlike an Aloris type piston, it would seem your arrangement must pull down as I see no tracks or dovetails to locate the toolholder.
Sorry for so many questions. It is a rather complex machine.
-
You may have been able to get 10 tools by having 6 on one 'half' (first 180 degrees) and 4 on the other and still used the 48 tooth coupling,
I tried many designs, to keep things as balanced as possible I had some designs with one position being 4 teeth apart and the next 7, but I was never truly happy with the design.
however, The software to run such an arrangement would be complex and error prone and probably require a sensor or two to operate safely.
It would not really have been a problem as I use the Indexing drive and thus would just have different encoder counts for each position, the actual M6 macro in Mach would have been exactly the same as I have now with the exception of 10 tools possible rather than 8.
Seems to me that 8 equally spaced tools with quick change holders is the bulletproof solution.
Yes seemed to me like the best solution but definitely not the easiest as the VDI setup requires fairly accurate positioning of the holes, especially the wedge ones in relation to the tool holder ones.
Just one more (compound) question if I may; presumably the turret is supported by a draw tube or shaft. Dies that shaft ride on bearings in the housing or on the hydraulic seals?
It is a shaft, see pic below . It has a brass bush at the front (yellowish bit in the drawing) the rear just has a quad seal.
It is a hardened piece from the original machine or did you design/build that part?
It is from the original but is not hardened, just a decent quality alloy steel, probably EN24T or similar.
Lastly, I note that the pulley does not shift in and out with the Turret. Is the shaft splined, keyed or slotted or is there a special slip joint and again was this recovered from the original machine or new design?
I bought a length of splined shaft and a splined bush and fixed the splined shaft to the end of the turrets shaft with a socket capscrew, I also machined a flat on the end of the turret shaft and machined away the end of the splined shaft leaving a stub on one side so that the two shafts mated together and could not rotate.
The original turret was rotated by rack and pinion under hydraulic pressure so it did not need any thrust bearings. I had to recess the front of the piston section to fit a roller thrust and I also fitted one at the rear.
The piston itself is basically just a disc with two piston rings on the periphery and it fits on the shaft and the shaft to piston is sealed with a quad ring.
The piston itself does not rotate, it has a pin that fits into a hole in the turret housing (red circled in drawing) so it is just the shaft that rotates and the piston only slides back and forth.
I have also attached a screenshot of the splined shaft, splined bush and also the part I made up to activate the limits that tell whether it is clamped or unclamped. Also a pic of the back where you can see the limit switches. Prox's probably would have done just as well but I like my mechanical switches :)
Will answer your last question in another posting as I will have to get some screenshots as to how the VDI system works.
Hood
-
OK, last, last question; there appear to be some kind of trapezoid shaped 'posts' sticking out of the turret surface that the tools attach to.
Think you may be meaning the blanking plugs, they are just used to cover the holes left when there is no tool in that position, stops crap getting in :) I was going to make some up out of delrin but WNT had a special on where they were only £5 each so I bought a few. They are just moulded plastic rather than machined but that is all that is needed. See pic below from WNT website showing a similar one.
Also I noted holes in the turret edge that look like there are screws or cams of some kind. Is this the quick change mechanism?
Yes that is the wedge, it is fitted into a hole that is at 20 degrees to the top of the tool holders slot. The wedge itself has a face milled at 20 degrees and then vee'd to mate with the vees in the tool holder. The centre of the hole for the wedge is 30mm in from the turret face and the central vee in the tool holders are 29.7mm from the face that contacts the turret plate. That means when you screw in the wedge it mates with the front of the teeth on the holder and forces it back into the face of the turret.
Normally VDI turrets are round rather than having flats on them. I do not have an accurate enough indexer, or rather I may do but was not willing to trust it's accuracy for this job ;) so I milled the OD edge so that the faces would provide the indexing and also the correct angle for the holes.
Hopefully the pics below will show how it all works.
Hood
-
Hood,
Well, I just transitioned from 'impressed' to 'amazed'. :o The video does not do justice to the complexity and ingenuity that went into this machine. Often the best solutions are evolutionary. The trick is to persevere thru the fails. Assess the performance gaps, re-design and have another go at it.
Plus, you have anticipated and answered my next question already, so I'm all caught up!
However, as a follow up on the topic of 'indexing' drives:
It would not really have been a problem as I use the Indexing drive and thus would just have different encoder counts for each position, the actual M6 macro in Mach would have been exactly the same as I have now with the exception of 10 tools possible rather than 8.
I recall that you are using Allan-Bradley and those guys have both 'universal' input and also positioning in the same drive. Is that correct? With Mitsubishi, you have to buy different drives and Yaskawa has 'motion controller' to pair with their drives. I have one of the Yaskawa controllers that I purchased with the idea of running an ATC carousel or chain with differently spaced 'tool spots', but I have not had time to mess with it. Also, I have in-house at the moment a customer's new Sigma5 Yaskawa motor and drive in 3kw size and it is uber impressive. The motor is about half the size of my 3.6kw Mitsu and it's fast!
As you know, the manuals on these things are just huge and I have not looked at what the new Yaskawas have to offer (beyond what I needed for the task at hand), but I am leaning toward the Allan-Bradley since from your description, it would be more or less 'universal'.
-
The ingenuity of the VDI system has nothing to do with me, it was the German Engineers Society :)
The Allen Bradley DSD drives come in a few versions, there is the standard model which can accept Stp/Dir and/or Analogue commands and then there is the Indexing version which can, in addition to that, do Indexing.
The standard drive would have a designation something like
DSD-020
The Indexing drive would be
DSD-020X
There is also a Sercos version which you can disable the Sercos via the Ultraware Software and doing that turns it into an Indexing drive. The model number for them has a SE on the end, so something like DSD-020SE.
There are also 2 versions that support device net, they too can be used as normal drives but only the Indexing version of it can do Indexing.
How the Indexing works in the AB drives is you assign up to 6 Inputs and depending on which Input combination is active you can have up to 64 different Index positions set. You can configure the RPM, Accel, Decel, Position etc for each Index.
Below is a screenshot showing a few of the Index setups in my Turrets drive.
Hood
-
The ingenuity of the VDI system has nothing to do with me, it was the German Engineers Society :)
The VDI system deserves its own Kudos, no doubt, but then again, a lot of engineering also goes into a harmonic drive, and the designers of that component deserve their kudos in turn, however, it is still a component and can be used in a good design or a bad design. It appears to me that you reclaimed some components and methods from a previous machine, but that monster sitting on your lathe is a Hood design. Integrating all of those pieces together and adding a slip joint and servo control is not a trivial task. Nor was the machining of those tool holders.
I will add that the logic and methodical movement thru the evolution of the machine design is classic prototype engineering. Regardless of where the components came from, the overall machine is an achievement. That's my view. 8)
How the Indexing works in the AB drives is you assign up to 6 Inputs and depending on which Input combination is active you can have up to 64 different Index positions set. You can configure the RPM, Accel, Decel, Position etc for each Index.
Below is a screenshot showing a few of the Index setups in my Turrets drive.
Hood
I don't think it gets any better than that. . Sigma5 has some kind of indexing capability, but I have not had time to dig into that encyclopedia of a manual yet. The Signma5 would be an expensive alternative, in any case, so I'll start poking around for an A/B drive using the data you have provided.
Question: is there a provision for fine adjustment, per tool, to bring the cutting edge to center for say the parting tool??
-
You can get height adjustable holders for the parting tools, see pic.
Theoretically there should be no need for height adjustment as the hole for the holder fitting into should be exactly on centre. The holders themselves should also be exactly on centre for the drill/boring bar holders and the normal tool holders should have their face either 20mm below or 20mm above centre.
That is of course the theory ;) we all know that the square section tools can be off a wee bit so I just pack them up with brass shim. I have made a holder for setting tools up and have it so I can set the height as well as setting a rough length. That way I can command an offset for a known tool and then just take a cut and then I am able to set the offset for the new tool.
Pic of the setting block below.
The parting tool I use is just the normal block/blade type so I can just shim to suit and do not need the type as mentioned earlier.
Hood
-
Oh and I never made the tool holders, bought most from eBay at a good price and only had to buy a couple new, even then the price of them has dropped massively in the last few years.
Hood
-
Oh and I never made the tool holders, bought most from eBay at a good price and only had to buy a couple new, even then the price of them has dropped massively in the last few years.
Hood
Sorry, I meant to say the wedges, which are the tool holder holders . . sorta . .
Thanks for taking the time to explain everything. Very interesting stuff. Nice chatting with you again!
-
Meant to say, just in case you were unaware. The Allen Bradley DSD drives can be used with motors other than Allen Bradley ones as long as they have an acceptable encoder and obviously that the current/voltage ratings fall within the drives specs.
You just need to know certain specs of your motor so that you can make up a custom motor file. Below is a pic of the info needed and the type of encoders accepted.
You can even use an encoder without the commutation signals but if doing so every time you power the drive it will do a self sensing routine which would mean the motor would move approx 1/4 turn.
Hood