Machsupport Forum
General CNC Chat => Show"N"Tell ( Your Machines) => Topic started by: simpson36 on May 05, 2009, 04:24:11 PM
-
Hi everyone,
After much study and work and help from this forum and other resources, I have my servo powered 4th axis up and running. It does everything I wanted it to do.
In the current arrangement, the mill spindle stays active and still functions as normal.
http://www.youtube.com/watch?v=2KNit__LJE4
For now, some interesting programming is needed along with a custom macro to do the thread passes, but in time, I think Mach3 will have features to make this much better.
I don't plan to make aluminum bolts from round stock. The part just demonstrates the various functions I wanted to get working.
-
Hi everyone,
After much study and work and help from this forum and other resources, I have my servo powered 4th axis up and running. It does everything I wanted it to do.
I the current arrangement, the mill spindle stays active and still functions as normal.
http://www.youtube.com/watch?v=2KNit__LJE4
For no, some interesting programming is needed along with a custom machro to to the htread passes, but in time, I think Mach3 will have features to make this much better.
I don't plan to make aluminum bolts from round stock. The part just demonstrates the various functions I wanted to get working.
Excellent demo, congrats on such a great achievement.
ATB
Derek
-
Hi, Simpson36
Very Nice job you've done, Show us more if you get time.
Chip
-
Beautiful work, my friend, thanks for sharing.
Dave
-
NICE JOB, Just a note on the indexer head . If you plan to use it much machine the housing for seals or orings and fill the cavity with oil to lube the shaft and housing. We used those for a few projects and that helps keep them going without much wear.
(;-) TP
-
Thanks for the compliments.
Vmax,
It's an el-cheapo indexer. So far I have machined the ends only for needle thrust bearings so that I can run with zero clearance. There is no internal cavity, as the spindle just rides in the cast iron bore with a few thou clearance. There really isn't enough meat to cut the ends for seals or for ball or tapered roller bearings. This iteration was just for development purposes . . . i.e. "get 'er done" and have something to work with.
I started with a 640oz-in stepper, went to a 900+ stepper, then to a NEMA23 Servo and now to the final NEMA34 Servo. The new 1,800 line encoder for it arrived yesterday (new model from US Digital)
So, now turning back to the mechanics of it, the only parts coming forward from what you see will be the motor mount and the big timing pulley. The next step in the project is to take a better quality 5C spindle from a Phase II brand indexer (sitting in my garage)
http://www.use-enco.com/CGI/INPDFF?PMPAGE=453&PMITEM=240-3226
. . . . and mount it on sealed ABC3 grade deep groove ball bearings in a new head made from a solid block of metal (I haven't decided on the material). For the level of this project, those bearings are a reasonable compromise. I used them on the original X2 spindle and they have done very well. SKF calls their version 'Precision Plus' . they run in the $50 per bearing range.
bearings
-
Great little machine, well Dun :)
Can you please tell, how you controll the spindle I mean as an axis or as spindle in mash ?
If it is controlled as an axis do you have an issue with number of steps then milling?
If it is as a spindle how you controll an indexing motion?
Kind regards
Ilia
-
That is very cool! And just think how much money you can save by making all your own bolts from now on! :-)
Regards,
Ray L.
-
Himy,
That's exactly my devious plan. I'm SICK of spending over $1 for a friggin' bolt at the hardware store. I only need to make about 23,000 bolts and I'm breaking EVEN, baby!!!
CoolFox,
As shown in the video, it was acting stricly running as the A axis with all rotations pre calculated, and what you see is the max speed severely limited by the Gecko340.
Later, I learned how to do some 'semi - documented' tricks in Mach that improved the process:
With appreciation for all those who helped me figure it out, let me say it is a pretty slick process. Only thing left to work out is the screwy feed rates. XYZ and A are from different planets in that regard.
Primarily I had to write some VB macros:
One handles the switching the A axis and the spindle using Mache's Swapaxis function.
Another swaps back and the re-homes the A axis (the secret that makes everything work) . . . and zeros the machine and DRO.
Another takes arguments and caclulates threading moves and passes the variables back to mach for use in a gcode program.
Friggin thing is REALLY fun to play with :D I just ordered a Viper servo drive that will double the speed to the full 4,000 RPM
-
Simpson36,
Thanks for reply I was asking as I'm in process of upgrading my second lathe Smart&Brown L-model to CNC and wanted to add some active tools with extra Z axis.
I'm jeweller by trade and specialise in making of full eternity rings there turning and milling/drilling process all day (this is my web album http://www.pixum.co.uk/slide/3540283 in case you interested :-) on the moment I turn blanks on the lathe and then transfer them on 4axis cnc mill (made my self using THK KR slides and Mitsubishi melservo drives motors) for milling/drilling this take time and alighntment not allways perfect.
In regards of spindle homing after axis swap Great idea as I have on bough lathes (The other lathe I have Schaublin 102) servo drives 2.2 kW Servostar CD series with Gold line motors and incremental rotary encoders have K output impulses witch acquire once per revolution! I'm going to try this.
Is it necessary to zeros the machine and DRO after swap home axis? Can you use fixture offset for milling operations?
-
Coolfox,
Some encoders have a separate 'index pulse' in addition to the regular incremental quadrature. My encoder has this, but the motor is geared down to the spindle, so I can't use it for indexing and therefor I added a separate photointerruptor to the degree wheel.
It is not neccessary to zero the machine after homing unless you are recutting a part that has already had indexed operations done to it, or if you want to use absolute azimuth locations. Another reason would be to have the DRO numbers be meaningful for monotoring purposes. The Mach manual describes the A axis as 'infinate', and I have taken it to the tens of millions, so I don't see a practical limit there, and if your Gcode utilizes only relative (incremental) moves, and a DRO showing rediculously huge numbers is not a problem, then you would not need to zero the A axis.
Fixture offset is unaffected by the A axis shenanigans . . so far anyway.
My objective is to make a sort of 'mini machining center' and emulate as many of those funcitions as time and budget (and Mach3) allow. A combined spindle/indexer is, I think, the first and most important step, and I have that pretty much worked out. I've just been prototyping to test feasibility up to this point, but I'm satisfied now that the concept is proven out sufficiently to start designing a sturdy permanent head with reasonable quality ball bearings and alignment pins and table mounting. I'm going to have two gear ratios 3:1 and 2:1 to accomodate different materials and stock diameters.
EDIT: checked out your site. That is some seriously unique and beautiful jewelry! I pray my wife never discovers your site . . . :-X
-
Simpson,
regarding the gearing and encoder Index, it may have been possible for you to utelise the pulley ratios in Mach so that the correct speed was shown. Probably not needed now as you have the interupter but just thought I would mention for your next build as it may save time or hassle.
Hood
Edit
Just thinking again, dangerous I know ;)
I think you are just using the parallel port so maybe the above info will be useless as chances are the index pulse from and encoder will be too short for the PP to see accurately, you could of course electronically lengthen the pulse but probably easier just to use the interupter after all.
-
Simpson36,
From my experience. I have 1;1 ratio on the machines as if you use powerfull 1KW+ servo for the drive you normally wouldn't need more torque unless you want to cut 1.5mm deep cuts on tool steel 10mm diameter :-) my 2.2kW servo I have to limit torque 1/2 as it rip the 25mm wide AT10 cam belt in shreds then I accidentally start the locked spindle :-( .
You want to build a lathe headstock why not buy ready made one?
In regards of ball bearings have to say if you have time to make plane bearings it work better in spindle from experience of use Shaublin have double row roller on front and two angular on the back and cost of etch in £200+ compare to plane bronze taper bush bearings in Smart&Brown doesn't stand a chance in stiffness and precision you my say constant lubrication requite for plane bearings in Smart it organised with simple capillary effect felt pad inside bronze bush and not requite attention for weeks of constant use ( I have used it every day for 10 years and no wear on spindle or bearings noticeable!!!).
I'm planing to upload some photos, as would do maintenance ( want to respray it), of internal construction of headstock in the mean time you can have a look here usefull info http://www.lathes.co.uk/smartbrown/page3.html
Would you mind to share macros for your machine and sample of code (for that infamous screw :-)?
Kind regards
Ilia
-
Hood,
I don't need to read the speed since there is an encoder on the motor. The issue is homing te axis. There are some symantics do deal with . . the 'index pulse' is an ADDITIONAL once per rev pulse built into the encoder. I can't use that with the gearing , because Mach has no way to know the position of the spindle from the position of the motor. i.e. index pulse from the motor can mean the spindle is at 0, or 120, or 240 degrees. Mach would no know which, so I have to have a n index on the spindle itself (as far as I know, but you've enlightened me before).
You are correct about the form of the index pulse. It is not something that can be connected to the BOB in any simple fashion (I don't think) as it is a tiny ampitude square wave or something like that. If I could have used it, I would have already been pestering the electron guys for a how-to.
-
Coolfox,
I agree with you completely, but you are on a different level. Many moons ago I had my own machine shop to develop prototypes and am familiar with the construction of machine tool spindles.
I would love to use specific purpose machine spindle bearings, but the cost is way out of line with a hobby level toy sized machine. Even angular contact bearings would be a bit of overkill for the amount of stress I can actually put on them with the miniscule power that I have available. I also know that preloading deep groove ball bearings is not proper, but that's how these little mills are set up and installing some decent ABC3 sealed bearings ($100 worth of spindle bearings in a $500 machine) is as far as I think is reasonable for an upgrade. I have to stick with what I can fit in my very crowded garage with my wife's classic Vette (which is not going outside), and what I can spend on a hobby and not end up living in said garage (we have no dog house). I've already passed double what I said I was going to spend on the CNC conversion. This is not a good strategy for getting my favorite cookies. ;)
This is all the motor I have to work with for the 4th axis power:
http://www.homeshopcnc.com/servo2.html
This equipment will never see production levels nor will it be called upon for precision work (.0005 and closer). Those are the parameters for the project.
Yes, I would share the macros and other stuff, but it would not be useful unless you had the same setup as I have.
-
Ok, point taken. I won't be pestering you no more :-)
The code and macros I want to have a look as a sample with info to think about and may be cut few corners :-)
I just was looking through my stuff and came across spindle just like you described deep groove sealed bearings 6203RS screw shaft M19X1.3mm cans say for sure may be imperial for the chuck or else with 15degree looks like for ER 16 or similar collet other end is 15.85mm diameter 21mm long hoke through 8.5mm
center height 65mm clamps on 60mm wide dove tail with I also have I used this for grinding once 5 years ego :-)
Interested?
-
Yes, I would share the macros and other stuff, but it would not be useful unless you had the same setup as I have.
Hi simpson36,
I think your setup is great! I'm intermiittantly building a similar system also using a cheap spindexer...
I've never actually written macros on my own, but have modified a few.....
Seeing your macros and learning how you integrated your moves, axis swaps, and homing would be highly benficial ..... I think..
I've often only dreamed/ponderd of doing just what you have done....
So, if you wouldn't mind, I like to see your macros....
Thanks,
John (Bloy)
-
Coolfox, You're not pestering me. I enjoy conversations and debates on topics that interest me. There is always something new to learn. That's one of the enjoyable aspects, second only to actually designing and building stuff.
Thanks for the spindle offer, but I want something that is reproducible . . i.e. something that someone could make for themselves if they chose to follow the project and duplicate it.
Coolfox and Bloy, I'm definitely interested in starting a thread on the programming/control aspects of an combination indexer/spindle.
The macros I've written so far are just the basics to get things working. I'm going to want to talk to the Viper drive thru the serial port, and get Mach to adapt based on the data . . . . . that should be an interesting challenge. Hopfully enough to lure Hood and Vmax or some of the other Mach programming gurus into the discussion.
Question is, where to start the thread. General Mach Discussion, Mach Programming, or Show and tell.
-
Hood,
I don't need to read the speed since there is an encoder on the motor. The issue is homing te axis. There are some symantics do deal with . . the 'index pulse' is an ADDITIONAL once per rev pulse built into the encoder. I can't use that with the gearing , because Mach has no way to know the position of the spindle from the position of the motor. i.e. index pulse from the motor can mean the spindle is at 0, or 120, or 240 degrees. Mach would no know which, so I have to have a n index on the spindle itself (as far as I know, but you've enlightened me before).
You are correct about the form of the index pulse. It is not something that can be connected to the BOB in any simple fashion (I don't think) as it is a tiny ampitude square wave or something like that. If I could have used it, I would have already been pestering the electron guys for a how-to.
Ah sorry, was not reading your previous reply correctly, was thinking you had fitted the opto for an index pulse, should have remembered it was as a home switch ;)
-
Hi everyone,
After much study and work and help from this forum and other resources, I have my servo powered 4th axis up and running. It does everything I wanted it to do.
In the current arrangement, the mill spindle stays active and still functions as normal.
http://www.youtube.com/watch?v=2KNit__LJE4
For now, some interesting programming is needed along with a custom macro to do the thread passes, but in time, I think Mach3 will have features to make this much better.
I don't plan to make aluminum bolts from round stock. The part just demonstrates the various functions I wanted to get working.
I just can't get over your video. it's so neat!
I want to apply this to my router to accomplish custom threaded dowels, or spirals, along with the other variations in the piece form you demonstrate(hex head in your case) ..all done in single code file. That's why I am so interested in your "basic" (as you say) macro coding.
....gotta play catch-up....been doing other things while my machines were covered with sheets for way too long a period.
-
Viper servo drive is on order. Should allow me to use the full power and speed of the servo motor with the 1,800 line encoder.
Design is finalized after about 3 iterations. Here are the first final parts: the motor mount and the belt tensioner.
(http://www.thecubestudio.com/pictures/4thAxis/BeltTensionerWEB.jpg)
(http://www.thecubestudio.com/pictures/4thAxis/4thAxisMotorMountWEB.jpg)
With the Viper and permanent motor mounting and belt drive, I'll stop and do some more testing and post another video.
Then it will be on to making the permanent spindle head.
-
That is very cool! And just think how much money you can save by making all your own bolts from now on! :-)
Regards,
Ray L.
This is one of the parts I had in mind to make when I was messig with code to make the 'bolt' . The latest iteration of the 4th axis made these with a single program and one setup, including tapered NPT threads, hex flats, the thru hole and cutoff. It's coming along . . . .
(http://www.thecubestudio.com/pictures/4thAxis/CorvetteVacFittingcommented.jpg)
-
Hi, Simpson36
Very Nice job you've done, Show us more if you get time.
Chip
OK, I'll be making another video soon. Very cool as you can clearly see the swap-axis and re-homing functions as it makes the part in the previous post. For now, here is the final drive mechanism. This will be moved over to the new head once I have that built:
(http://www.thecubestudio.com/pictures/4thAxis/FinalDriveDesign01commented.jpg)
Poor thing after many hours of abuse:
(http://www.thecubestudio.com/pictures/4thAxis/FinalDriveDesign03WEB.jpg)
This is still one of the last temporary 'get-er-done' quickie that needs a final design:
(http://www.thecubestudio.com/pictures/4thAxis/FinalDriveDesign05commented.jpg)
-
For those who asked, I will also be posting the swap axis macro and mayby some others.
After seeing Ray L.'s beautiful macros, I'm just too embarrassed to publish my sloppy code until I get it spruced up a bit . . . :-[
-
For those who asked, I will also be posting the swap axis macro and mayby some others.
After seeing Ray L.'s beautiful macros, I'm just too embarrassed to publish my sloppy code until I get it spruced up a bit . . . :-[
Hi simpson36,
When you post your "improved" code, would you also include the original that worked for you before "cleanup"?
It would be interesting to see what works without being "elegant". Don't be embarrassed.... I'd just view the process as a "development". ;)
Great details in the photos. Nice setup!
Thanks,
Bloy
-
Simpson, RayL is a software writer to trade, I think, so you should feel no embarrassment :)
Hood
-
Here is it . . . will be avail in HD after YouTube is finished processing, but is viewable now.
http://www.youtube.com/watch?v=NzY-OqckqpI
Special thanks to Hood for the re-homing info. Also Vmax, RayL, Machinemaster, and others . . .
BTW, lots of light cuts this go round. I will have a lot more power once I get the Viper drive (delayed waiting on parts :() With the 1800 line encoder, the Gecko faults with a normal cut.
-
Simpson great job. (;-)
One last note. THere is room in the housing to install oil seals. Then you fill the shaft area with 90 weight gear lube through the port on top (lockdown screw) It will run a very long time without wearing out. I use a simalar setup on special grinder application where the indexer has to rotate at 60 rpm continuously while grinding.
Just a thought, (;-) TP
-
Simpson,
would be good to see you add the vids added to the Mach youtube site :)
Hood
-
Vmax, I recall you mentioning that before. I have installed needle roller thrust bearings on each end, those are neccessary for operating as a spindle and they actually peeked out past the surface in one spot . You can see the oil leaking out at that spot in one of the photos. If not for cutting the ends for the thrust washers, I could do as you suggest.
A new servo drive (Viper) and a little more voltage should get the 4th axis to a max of 2,500 RPM with a 2:1 ratio. Sustaining that speed for hours on end would cause trouble with the setup you are suggesting. At a minimum, it would consume a lot of power and I don't have so much I can give it away to a plain bearing or to a non bearing thrust. Used as an indexer or at low speeds, I agree with your assessment.
A full ball bearing head with sealed bearings is the way to go for a reliable zero play spindle on a budget, methinks. I'm going to 'ghetto' it a little with just some quality SKF 'Precision Plus' bearings and not angular thrust or high precision. No point with the current machine that flexes all over the place anyway . . . :-\
Hood, how do I post something specifically to a Mach3 group on Youtube? I have Mach3 as a keyword, but that's all I know to do. Not a big 'tuber. Still dinking around trying to get the best HD quality too.
-
If you click on the banner where it says "See Mach3 Machines Here" you should be able to join the group, you can then choose to add a video and it should take you to your videos page. But its been a long time since I joined so the steps are maybe not exactly as I said.
Hood
-
Here is it . . . will be avail in HD after YouTube is finished processing, but is viewable now.
http://www.youtube.com/watch?v=NzY-OqckqpI
Special thanks to Hood for the re-homing info. Also Vmax, RayL, Machinemaster, and others . . .
BTW, lots of light cuts this go round. I will have a lot more power once I get the Viper drive (delayed waiting on parts :() With the 1800 line encoder, the Gecko faults with a normal cut.
That is seriously cool! I hope to do something similar someday, if I ever finish my toolchanger.... I've got some simple parts I sell that could be done very easily like that.
Regards,
Ray L.
-
For those who asked, I will also be posting the swap axis macro and mayby some others.
After seeing Ray L.'s beautiful macros, I'm just too embarrassed to publish my sloppy code until I get it spruced up a bit . . . :-[
Hi simpson36,
When you post your "improved" code, would you also include the original that worked for you before "cleanup"?
It would be interesting to see what works without being "elegant". Don't be embarrassed.... I'd just view the process as a "development". ;)
Great details in the photos. Nice setup!
Thanks,
Bloy
Here are the two basic macros with minimal comments added:
'M940.m1s 06/01/09 ©2009 www.theCUBEstudio.com
'Macro to make 4th axis become spindle
'use M950 to swap back and re-home
resetaxisswap()
SwapAxis(6,3)
End
'M950.m1s 06/01/09 ©2009 www.theCUBEstudio.com
'Macro to swap 4th axis back to A and re-home
'use after M940 to restore 4th axis to A axis
'Waits until axis stops moving before reset
'Moves axis 20 degrees if home switch is active
'Re-homes A axis
'Resets machine A to zero
While ismoving()
sleep 100
Wend
resetaxisswap()
'CHECK IF A AXIS HOME SWITCH IS TRIGGERED
If getLED(39) Then
here = GetABSPosition(3)
here1 = here + 20
Message ("Axis home switch triggered - moving off to " & here1)
code "G0 A" & here1
While ismoving()
sleep 100
Wend
End If
'RE_HOME A AXIS
DoOemButton(1025)
While ismoving()
sleep 100
Wend
'RESET A AXIS TO MACH ZERO
SetMachZero(3)
End
http://thecubestudio.com/Mach3/macros/Mach3Mill/M940.m1s (http://thecubestudio.com/Mach3/macros/Mach3Mill/M940.m1s)
http://thecubestudio.com/Mach3/macros/Mach3Mill/M950.m1s (http://thecubestudio.com/Mach3/macros/Mach3Mill/M950.m1s)
-
Thanks loads!
...just a question about the motor mount parts...did you bead-blast those to get the color/texture? Or maybe you vibrated/tumbled them? You might have already mentioned someplace but I didn't see how you "finished" the parts.
Nice!
Simpson36, I'll probably never make the "fitting" that you carved out in your super video, and the actuall G-code in its original state would never be used "as is" by me. However, a step by step study of that g-code in conjunction with the macro usage might really make this process come clear.
.....you've undoubtedly got many things going on, so with reservations, am I asking too much to post the actual g-code running in your video.? Would this be possible? Somehow, in the future, in some way I hope to return your generous favors. (Continuing to compliment your work as you demonstrate it just won't suffice...:) )
Thanks again,
John
-
Bead blasting is for delicate stuff. I just use fine sand and then coat with a product from Eastwood called satin clear for bare metal. There is not a lot of paint that will stick directly to aluminum. This is an amazing product . . expensive though. 'Self-etching' or otherwise 'stick to bare metal' products are caustic (have acid in them) . . . . VERY nasty stuff . . respirator only, please!
As to the G-code, it calls a few other macros and internal subs and uses a lot of variables within the code itself so it will be very difficult to decipher without a lot of proper commenting.
I used to program a lot . . . 25 years ago . . and I was sloppy even then about commenting programs. A well deserved criticism I would get from colleagues was that my programs were difficult if not impossible to maintain because I would do very complex things and not explain what I was doing or how. I often could not even go back and figure out WTF I had done ???
If I can get time to comment the code sufficiently for it to be understood and maintained in some reasonable way, I'll make it available individually for those who may want to peruse. I do not have time to answer a lot of questions about it though.
-
Sure! Put me down as one of those individuals who will graciously accept your code for the sole purpose of solving a puzzle. I'm one of those that use ink and wouldn't think of asking another for the answers........yah, sure. ;).
Seriously, I promise not to ask ANY questions(within the first year).
What you are doing control-wise, is exactly what my machines are looking for.
Thanks,
John (bloy)
-
Update:
I have the Viper drive and it will run the motor fast enough. That's about the only positive thing I can say about that product at the moment. I have some more homework to do on the Rutex drive, but I think I'll try one of those next.
The final design for the new head is finished and the material, bearings, and sensors have arrived for the final prototype. Stay tuned . . . .
-
Sure! Put me down as one of those individuals who will graciously accept your code for the sole purpose of solving a puzzle. I'm one of those that use ink and wouldn't think of asking another for the answers........yah, sure. ;).
Seriously, I promise not to ask ANY questions(within the first year).
What you are doing control-wise, is exactly what my machines are looking for.
Thanks,
John (bloy)
I just found an early version of the g-code prior to using the new macros and variables. It has only the swap and swap back macros that I have already posted. This code is *somewhat* decipherable even though it is not numberd nor commented in any reasonable fashion, and the treading and peck drilling are verbose g-code for tuning purposes. But if you want to noodle over it, e-mail me your e-mail address and I'll send it to you.
-
Update: 4th axis completed. This will run at a max of 1,300 RPM or 2,000 RPM depending upon which motor pulley is used. Should be able to run all day long at top speed. There is a steel part that I make from 3/4" 12L14 steel at 1,800 RPM. it has 1/4-28 internal threads and flats on each side. This final iteration of the 4th axis will be expected to turn, drill, tap and cut the flats using one continuous program. We shall see . . . .
Basic drive parts moved over from previous prototype:
(http://www.thecubestudio.com/pictures/4thAxis/HeadDrivePartsWEB.jpg)
Final assembly:
(http://www.thecubestudio.com/pictures/4thAxis/FinalAssemblyBackWEB.jpg)
(http://www.thecubestudio.com/pictures/4thAxis/FinalAssemblyFrontWEB.jpg)
Homing sensor moved to back and oriented down to keep out as much swarf as possible.
(http://www.thecubestudio.com/pictures/4thAxis/SensorCloseWEB.jpg)
I have my new Rutex servo drive and so far it is doing everything it claims to do, so I should get full speed and power from the motor. When I get it all back on the mill, I will make one final video and post the link. Questions, comments , criticisms or suggestions welcome.
-
SImpson you may want to consider a sleeve in between the inside of the two bearings. AND compress the bearings together(preload). This will allow BOTH bearings to take the side loads instead of just the outer bearing.
Looking GOOD dude, (;-) tp
-
The purpose of the threaded collar is to preload the bearings. The spindle runs silky smooth with zero play. The bearings are so big, they will literally last indefinately in this application.
I don't understand the purpose of a sleeve, unless you are suggesting a 'crush' sleeve as used for example in older independent rear suspensions (Corvette, etc). Those are to maintain a specific clearance as opposed to preload.
The reason I chose an open frame instead of a solid block is the difference in expansion between the steel spindle and the aluminum head. I rely on the 'spring' of the open frame to accommodate the differential without binding the bearings. I considered a steel or cast iron solid block, but I decided against that because of the extra mass it would add that has to be moved around by the little steppers on the table. Ideal would be a hollow cast iron box similar to the x2 head, but I have no way to make such an item. I did consider using an X2 head casting. The open frame also provides more mounting options. With slotted mounting holes, I can mount the 4th axis anywhere from 0 to 45 degrees on the table. although there are keyways for quick alignment with the x axis.
I've had the motor up to 3,500 RPM so far at 62V. In service, it will run at 72V, but with a 1 ohm braking resistor, so I'm not sure yet what the top speed will be. Without the resistor, it calculates to almost 4,600RPM at 72V. I'm looking for 4,000.
-
YEP the sleeve would go between the two inner races of the bearing just like the Chevy rearend. This would spread the load of the inner races across both bearings AND allow you to tighten up the shaft to the bearings MORE. The standard bearings were NOT designed for much side load so spreading the load across the bearings would help to reduce stress and wear on them.
(;-) TP
-
Imagine the weight of the armature of an electric motor with a 45mm shaft . That's what the bearings are made for . . my little X2 is not even going to get their attention . . . ;D
I have the Rutex driving the assembled 4th axis now up to 350,000 steps per second with awesome silent holding power, and decent accel. I can set the accel up higher in Mach by being mindful to ramp up the speed from a dead stop to max RPM if the code calls for same, but that's a compromise I can live with for now.
Still running only 62V 5A on the bench. Will get 72V 20A when on the mill. Once i get it there, I'll be able to see my max speed and cut some metal with it . . should be fun.
QUESTION (for anyone):
I have a 1 ohm 100W power resistor in series with the motor per Rutex spec. It is in a finned aluminum housing and it gets friggin HOT! I have no accurate way to measure how hot this thing gets, but you definitely cannot touch it. How hot do these things run normally?
-
Simpson,
Those resistors are intended to be heat-sinked, to keep the temperature reasonable. I would expect probably up to about 70C to be a reasonable max operating temperature.
Regards,
Ray L.
-
This thing is a lot hotter than 70C so I better do something about that.
The resistor was mounted on the outside of the aluminum box that houses the 4th axis servo drive, but it was heating up the entire box!
I have some huge old finned heat sinks from Pentium 3 CPUs. Methinks perhaps I'll bolt the resistor to one of those and mount it remotely away from the control box . . where I won't burn the crap out of myself by accidentally touching it :'(
. . . -or- . . . I could mount it to the mill itself behind the Z axis post. That should be able to absorb a days worth of heat and would not result in having longer wire runs to the motor.
. . . -or- . . . I may just run without it and risk toasting the drive with deceleration voltage, although in my application, I just don't see that happening since Mach controls the decel rate.
Decisions, decisions . . . .
-
Below is my solution to the resistor heat issue, FWIW. The fan at the bottom blows into the box and cools the Rutex drive inside and the air exits the top and flows past the resistor, which is bolted to the box along its rear edge. It stays cool enough to touch most of the time.
Final score on the servo drive. With the Gecko I could get a .015 cut at about 700 RPM max spindle speed. With the Rutex drive, I am getting an .080 cut at 1,100 RPM max spindle speed. There seems to be power available to exceed this, but the maching rigidity is now the limiting factor. Needless to say, I am well pleased with that result ;D
I am sort of a servo junky now and I'm waiting on yet another servo drive, this time a Leadshine 810, to put the small NEMA 23 servo (from iteration 3 or 4 of the 4th axis project) on my X axis. Now I have discovered a small very reasonably priced ($180) AC servo drive made by the same outfit . . so probably I'll be playing with that sooner or later . . . my wife just got a new dining set, so I may be able to spend a little more on my toys . . :-X
Review and comparison of the servo drives I have played with so far is here if anyone is interested: http://www.thecubestudio.com/ServoDriveReview.htm (http://www.thecubestudio.com/ServoDriveReview.htm)
(http://www.thecubestudio.com/pictures/4thAxis/ControlBoxWEB.jpg)
-
SImpson you are not from the land down under are you??? That is the only reason I know to run the NOTICE sticker upside down (;-) LOLOLOL
Good Job, (;-) TP
-
Project completed! Here's a new video link showing the finshed 'mini machining center' doing its thing with Mach3. ;D
Feel free to ask questions or make comments. I tried to keep the 4th axis project uncomplicated so that anyone who wants to build something like this can duplicate these results. I don;t have drawings to share, but would be happy to provide sources and part numbers to anyone who wants them.
Once again, many thanks to all who helped this project along with information, ideas and generally sharing knowledge. I'm a newbee at CNC retrofits and could not have accomplished this without everyone's help :-*
Here is the link . . . just pay no attention to my wife's excited chattering in the background . . she thinks it's 8) and wants to know how to use it to do her wall art scuptures. . . which of course it will do beautifully . . . just as soon as I do this one last modification . . which won't cost much . . . really! :D.
http://www.youtube.com/watch?v=b2-Kdud7eiA (http://www.youtube.com/watch?v=b2-Kdud7eiA)
-
That is just SOooo Cool! Great!
-
Looking and working great and whats even better is you can now gift this machine to the good lady for her wall art and you can go get yourself a full sized mill while she is busy using that one and not paying attention ;D
Hood
-
That plan is Outrageous! Despicable! Sneaky and Underhanded!
. . . . I like it . . :D
-
You better just hope she doesnt read these pages ;D
Just wish I had the room for a real mill (big bedmill)l but they are all too tall for my workshop :(
Hood
-
new dining room = new machine, matching chairs= cnc conversion, vase with flowers=drive
you sitting in dog house eating...........priceless ;)
RICH
-
Project re-opened . . . . two new wrinkles for those following along;
I have avoided testing the 4th axis with a heavy lathe chuck holding a heavy workpiece because of concern over BEMF smoking the Rutex drive during a fast decel from 3,000RPM motor speed.
Two possible solutions have surfaced:
1) I have figured out how to change the axis accelerating parameters with a macro. This will be much better than trying to program a velocity step down in G-code (my previous solution). I'll be experimenting with that in coming weeks.
2) A new servo drive from CNCdrives that has an automatic brake resistor feature that monitors voltage and dumps excess automatically to a power resistor. This is nothing new, but it is new at a $170 price point. This new drive was shipped to me yesterday along with the same company's mid range drive. Anyone interested can watch the review as these drives are next up.
-
Thanks! ....I'll be watching...
Bloy
-
Duno what is cooler, the machine or having a wife that is so enthusiastic about the machine !!!!!
Mine just says, "so what" :(
Derek
-
OK folks, I got the Dugong drive and ran the 4th axis up to full speed with a 4" chuck and a hunk of cast iron in the chuck. I also changed from 3:1 to 2:1 ratio for the first time . . zoom! Max accell/decell it would handle. Shaking my whole workbench! I'd call it a success. The thing is scary . .LOL!
I'll make another video when I get a chance.
There is a lot of interest in 4th axis and I am considering making a limited production run of these machines, possibly for the European market, where most of the interest comes from. I am in the process of redesigning for production. The drive belt mechanism works fine, but is too complicated (read expensive) for the target market. Conventional trapezoid tooth profiles are also too noisy at full speed. I'll be going with the newer GT2 belt system and no idler. I'll be looking for a vendor to produce the spindles, preferably in Europe, so anyone interested in that can send me a PM.
The current iteration was intended as final so it is nicely finished and has the index photosensor on it and is ready for prime time. The only part used from an off-the-shelf indexer is the spindle and I already bought a Phase2 indexer quite a while back when they were on sale, so I have another spindle. I have to make everything else over again except the frame, so before I take the current 4th axis apart and trash all of the pieces, I thought I'd offer it up for sale if anyone is interested. I'd take a reasonable offer and just start over from scratch. It seems shame to waste a working prototype. Better someone using it than just having huge custom made timing gears and belts sitting on a shelf.
In any case, keep watching for more refinements and higher speeds. My new drive design is finished and I ordered the new belts, pulleys and aluminum stock today. If I have time, I'll design and build a prototype spindle; longer and with a threaded nose to accept a conventional chuck, but with my schedule that is unlikely this time around.
-
...been wondering what you've been up to!!
...patiently waiting for the video.... ;D
-
Current prototype is sold, so I will be fabricating a new 4th axis from scratch, this time with manufacturing considerations as a requisite.
Any suggestions for useful features are welcome at this point. For a production version, a belt guard will be incorporated into the design.
-
UPDATE: started new prototype. Again not doing anything exotic so anyone wanting to follow along should be able to duplicate these results and build yourself a nice addition to your mill. Material is 1" thick 6061 plate. Motor plate was planned as 3/8" thick. but is 1/2" due to 'scrap on hand' material sourcing. :-[
This is the new frame . . essentially the same as the first proto, just bigger to allow for larger chuck or workpiece. Also better fit to larger table on my planned next mill. What is very different is the motor mount. Expensive idler is gone and belt tension is now accomplished by rotating the entire mount. The big hole in the mount is not necessary. I put that in to ease belt and pulley changing if I want to experiment with different ratios. Currently 2.5:1 which is a compromise from previous 4th axis which had both 2:1 and 3:1 ratios.
(http://www.thecubestudio.com/pictures/4thAxis/Proto2%20Motor%20MountWEB.jpg)
-
These are the new drive parts. Older trapezoid tooth belt is gone, replaced by newer GT2 type. Quieter and can handle a lot more power. This time rather than making a pulley from scratch, I took an off-the-shelf pulley and modified it by cutting out the center and mounting it to a stock clamping collar which is comes already drilled for mounting a pulley. Between the stock pulley and clamp collar, I saved several hours over starting from scratch. Collar is bored for a slip fit on the spindle . . no more threads. Preload will be accomplished differently. More on that later:
Spindle is out of a PhaseII brand 5C spin indexer. Still an import, but about three times as expensive as the normal junk and it is much better quality. Roughness in center portion of the spindle is from me reducing the dia slightly in that area with a coarse zirconia flap wheel to ease bearing press and reduce the interference fit on the rear bearing for preload purposes. Front bearing (next to shoulder will have full press fit. I will probably be making my own spindle, but for now this is easier.
(http://www.thecubestudio.com/pictures/4thAxis/Proto2%20Drive%20PartsWEB.jpg)
-
WOW! I watched quite a few of your videos, excellent work! I either have a very long way to go with this hobby and your youtube videos were inspiring and frustrating at the same time. Or I need to sell this project and move on in another aspect of life, not! Btw your wife sounds like she is faking it! Now if you were making her jewelry? Excellent work and I cant wait til some of this is better understood instead of every step forward being so challenging.
-
Fabrication complete. Still needs a guard on the belt.
I am planning to build a tailstock for it by modifying another spin indexer.
I'll post a link to a video once I get it hooked up and cutting metal. Should be an improvement over the first proto, especially in accuracy due to the better spindle.
Detail of the motor mounting parts:
(http://www.thecubestudio.com/pictures/4thAxis/Proto2MotorMountStandoffWEB.jpg)
Final Assembled unit front and rear:
(http://www.thecubestudio.com/pictures/4thAxis/Proto2AssembledFrontWEB.jpg)
(http://www.thecubestudio.com/pictures/4thAxis/Proto2AssembledBackWEB.jpg)
-
OK, guys here it is! Video as promised. This thing kicks aluminum butt . . . :D
Video is real time and the part finishes in just a bit over 2 minutes . . i.e. twice as fast as the previous version. FINALLY, I am satisfied with it! A couple more very minor things and it's ready for prime time. Much easier to copy now with the idler and threaded pre-load collar eliminated. The entire 4th axis will also be available for purchase pre-made.
Next project I am currently working on is an all new scratch built table for the X2 and X3 mills which has X3 (or more) travel and ball slides on both axis. Here again, as I did with the previous 4th axis, instead of doing such extensive modifications to the existing working parts, I am selling the X2 CNC table (and saddle) as is for someone to drop onto an X2 and be ready to rock. Listed on Ebay for anyone interested.
Are we having fun yet? Yes, I think so . . . ;D ;D
http://www.youtube.com/watch?v=295Phu8GnjE (http://www.youtube.com/watch?v=295Phu8GnjE)
-
That's a wrap! Project finished. ;D
"Yeah, but can it cut steel?"
4th axis is now producing the little steel part for which it was created. Hard tapping is at the end of the video. Light cuts are due to lack or rigidity in the X2 table. Next project will solve that issue for X2 and X3.
This has been a fun project with a successful conclusion. Thanks to those who helped out with knowledge and ideas. Here is the final video:
http://www.youtube.com/watch?v=p0kIWpylrqk (http://www.youtube.com/watch?v=p0kIWpylrqk)
First run 30 front and 20 rear :
(http://www.thecubestudio.com/pictures/HingeCNC/Buttonsfrom4thAxis.jpg)
-
Very good Simpson. :)
Brett
-
Looking good Simpson, glad your happy with the outcome.
RICH
-
Thought you guys might enjoy seeing the little X2 doing precision grinding. I tried this with the original X2 table and no dice unless I locked the Y axis. Works great with the new table!
I figured out a way to run the 4th axis as the spindle and still have the X2 spindle running at the same time.
http://www.youtube.com/watch?v=tu20ndoPuS0
This is end prep for the permanent precision ground ballscrews for my new table. The new table had Thompson ballscrews temporarily since the post office lost the ground screws. I finally got them and they are on the table now. Zero lash and smooth as silk!
Surprisingly, I can actually hold less than .001" on the OD of the shaft I am grinding in the video. I have a Thompson on the Z axis also, but I put oversize balls in it to take out most of the lash and I have a gas spring (strut) on the Z pushing the head upward pretty hard, so as long as I sneak downward slowly, I can get fairly accurate results.
Notice the 4th axis is complete now with belt guards, the home sensor has been relocated under the belt guard and there are pin blocks to lock the spindle for collet changes.
-
Hi Simpson,
Nice work, creative thinking.
How are you getting 350,000 pulses per second from Mach 3?
Are you turning with the bottom edge of the carbide insert?
Ozzie
-
Sorry I don't understand either question. How are you arriving at the number of steps and what do you mean by the 'bottom edge' of the carbide?
Here is another video showing my new table runnibg 300 IPM rapids during PCB routing and drilling.
I think I can cut a lot faster, but I don't have much experience with PCB or FRP in general, so I'm sneaking up on it. If anyone has some feed suggestions, I'd appreciate hearing them.
http://www.youtube.com/watch?v=-6l7w3OwnC4
-
On your reply #42 you mention 350,000 steps per second. Mach 3 can't do that, I think.
In your last video of turning, I see a 60 degree carbide insert. The camera is looking at what is normally the top of the insert, (in a normal lathe setup). But if the spindle is turning counter clockwise, you must be cutting with the bottom edge of the insert. If it's not turning that direction, then you're cutting a left hand thread???
Ozzie
-
Ozzie,
Now I understand. I said the DRIVE is pushing 350K steps, not MACH. Many drives have step multipliers on the input side. If I set the multiplier to 10, for example, MACH puts out 1 step and the drive treats that as if it were 10 steps. You may see this referred to as 'electronic gearing'. Where the trouble come is when the drive itself has a bandwidth that will not allow it to pump OUT enough steps. In this case, you are limited in RPM.
I'm not cutting with the wrong side of the carbide. The threads are calculated by a macro that I wrote and can go either direction. You just enter + or -. Also the tool can be bolted to the mill facing 'in' or 'out', and the 4th axis can run either direction. The big inset obsures the cutting action, so the setup for the video is 'backward' so that you can see the cutting action. I didn't change the program, I just flipped the tool and reversed the rotation for making the video.
-
Got it! Thanks
Ozzie
-
Here are some details of my new XY table. I designed this for a client. There will be 4 of them to be put on new X3 machines for a specific purpose. This was probably the most fun thing I've done with the CNC to date. The Thompson ball screws are temporary because the post office lost the ground lead screws that I bought for the project. They finally did arrive and the grinding video shows the end prep. I sold the original CNC converted X2 table as a complete setup with all the CNC stuff included, includeing all three axis photointerruptors (home switches).
The new screws are pretty amazing. I have the table set now for 300IPM rapids. You can see this in the video showing PCB routing.
(http://www.thecubestudio.com/pictures/MillTableModular/BaseInstallHolesDoneWEB.jpg)
(http://www.thecubestudio.com/pictures/MillTableModular/BaseInstallCompletedWEB.jpg)
(http://www.thecubestudio.com/pictures/MillTableModular/AssemblyBase%20and%20Xslide.jpg)
(http://www.thecubestudio.com/pictures/MillTableModular/Assembly%20Yleadscrew.jpg)
-
A friend told me that CNC conversion is like a virus. Easy to get and once you get it and you can't get over it . . . :D
I have the slides for the Z axis . . . :o . . . . . somebody stop me . . . .
Continuing with the table photos . . .
(http://www.thecubestudio.com/pictures/MillTableModular/AssemblyFinalWithoutMotors.jpg)
(http://www.thecubestudio.com/pictures/MillTableModular/AssemblyMotorYaxis.jpg)
(http://www.thecubestudio.com/pictures/MillTableModular/AssemblyMotorXaxis.jpg)
-
Those hose fittings are being used as stand offs?
-
Those hose fittings are being used as stand offs?
YEP!! Waste not want not . . LOL! Those are test pieces from my 4th axis project . . . I have <ahem> quite a few :-[
Speaking of recycling, I just listed these Thompson ball screws in the 'Bargain Basment'. Somebody may as well get use out of them. They are brand new and I have no further need for them.
-
Nice job Simpson!
Looks like you lost a fair amount of travel on your Z with these slides ;)
Are the mounting plates and the bearing plate made of aluminum??? Or is just the photo?
Daniel
-
Nice job Simpson!
Looks like you lost a fair amount of travel on your Z with these slides ;)
Are the mounting plates and the bearing plate made of aluminum??? Or is just the photo?
Daniel
Well, lets find out. It is 4.4" from the dovetail flat (on the X2 base) to the top of the table. I sold the original table and saddle, so if someone with an X2 can measure, we'll have the answer. I had over 12" of Z travel, so I wasn't particularly concerned about it. The new base plate is 3/4" and doesn't need to be for an X2 (the table is designed for an X3), so an other .25" could be eeked out of there on an X2.
Everything is aluminum ground jig plate. The solid aluminum table weights only 0.4lbs less then the original cast iron table.
I will be building an entire new machine frame also from aluminum and making a complete new head from aluminum. The total weight of both head castings on the X2 is 16lbs. The new aluminum head will weight 24.6lbs . . both figures do not include the spindle, bearings, motor, drive parts, etc.
The stock X2 column weights 17.5lbs. The new column will be 20lbs. X2 base casting weighs 17.25lbs. New base will weigh 19.5lbs
-
I watched your video of the grinding on the leadscrew. Impressive adaptation as always. Is there any reason why you grind rather than turn? Is the screw hardened so much regular cutters dont work well? Also I see you using rotary tables and indexers, do you control the table with a distance on the DROS referring to inches traveled or degrees? Seems like inches traveled would be different depending on how far away from the center of the rotary table as the work is spinning. Does any of that make sense?
-
Simpson,
Can't see how the aluminum parts are heavier than cast iron ??? Are they that much bigger in size (almost 3 times) or what?
I will be building an entire new machine frame also from aluminum and making a complete new head from aluminum.
Got to find a new name for your machine... you can't call it X2 any more I guess ;)
Daniel
-
I watched your video of the grinding on the leadscrew. Impressive adaptation as always. Is there any reason why you grind rather than turn? Is the screw hardened so much regular cutters dont work well? Also I see you using rotary tables and indexers, do you control the table with a distance on the DROS referring to inches traveled or degrees? Seems like inches traveled would be different depending on how far away from the center of the rotary table as the work is spinning. Does any of that make sense?
Good questions. The shaft is very hard. In general: you can turn hardened shafts with ceramic or boron or diamond, but it takes an extremely rigid and powerful machine and the tools are very expensive. In the case of a leadscrew, you are facing an interrupted cut as well so to have success, you would probably need one of the reinforced versions of the inserts . . now you are talking 30 times the cost of carbide. :o
I'm not using a rotary table. I had a big one many years ago and recently I bought a small 6 inch rotary table (PhaseII brand, pretty nice actually). But I only used it once before I sold it after converting to CNC.
I have only turned small 3/4" and down alum and steel parts, so for turning, the 4th axis is always running flat out at approx. 1,200 to 1,400 RPM. Other than that I use it for indexing and position it via azimuth (degrees). I have not messed with trying to set a surface speed via Mach using the part diameter.
-
Simpson,
Can't see how the aluminum parts are heavier than cast iron ??? Are they that much bigger in size (almost 3 times) or what?
Yes, bigger. Bigger and also thicker walls. The table for example is much bigger. Aluminum is light, but not THAT light. You would not want to carry a 1" x 6" x 48" piece of it very far. The X2 is very light, very thin wall castings. There really isn't a lot of material there. I'm old enough to remember when cast iron was dirt cheap and aluminum was stupid expensive . . . and the very idea of an aluminum soft drink can was pretty outrageous . . . . . well, obviously . . . . things have changed. In designing with aluminum, it is cheaper to go heavy with a less expensive alloy than to get the same strength from smaller parts made from the more expensive high strength alloys. If you are designing aircraft parts, weight and strength are primary, so you usually stick to the 7000 series, but for a machine tool, you have room for 'fat' parts and the weight is an advantage, so it plays well. I am making the head heavier intentionally to absorb as much vibration as possible on the tool side of the slide bearings. That mass has to be moved around by motors and leadscrews, so you can't just go crazy with it. I'll be putting cooling fins on the spindle for sustained high RPM running.
I will be building an entire new machine frame also from aluminum and making a complete new head from aluminum.
Got to find a new name for your machine... you can't call it X2 any more I guess ;)
Daniel
Suggestions?
-
Well, with all those modifications it can't be an X series any more... should move on to Z1 ;) ("Y1" doesn't sound well)
Daniel
-
Well, with all those modifications it can't be an X series any more... should move on to Z1 ;) ("Y1" doesn't sound well)
Daniel
When I was a teenager, someone remarked that my car could be very easily fixed up nice. All I needed to do was jack up the gas cap, roll a new car underneath and then let the gas cap down. :D
If I use only the spindle from an X2 . . . . would that be considered 'contamination' of the 'Z1'?
-
:D
You got to leave something from the X2, otherwise we won't believe you it was an X2 ;)
Daniel
-
:DYou got to leave something from the X2, otherwise we won't believe you it was an X2 ;)
Daniel
I found a heat treating place about a hundred miles away here in Texas, so I *might* be making my own spindles after all. We shall see
Here is the latest video showing the final iteration of the PCB routing and drilling operations. http://www.youtube.com/watch?v=r9Zf_5yHB1I
Based on some feedback I got, I kicked up the feedrate and am using my mister to provide a constant air stream. I'll mount a vacuum nozzle on the fixture, but not make a new video just for that, so this will be it until I finish or make significant progress on the new frame including the all new Z axis.
After much whining on my part, I have convinced CNCdrives to provide these little circuits free to US customers of their drives until the function is built into the drives (planned for future version apparently) I even offered to make the parts (just for fun really, cause I think it's very cool that I can actually make an electronic anything that actually works . .LOL!!) Well, they said OK, but it would need a cap across the power pins and also a top quality IC socket. The one I used initially was apparently an embarrassingly cheapo version that they object to. So I made the changes and the part will be available to anyone who wants it . . free to US buyers of any CNCdrive Whale or Dugong servo drive.
-
I would assume that heat treating is the easiest step in making a spindle. If you can handle the machining involved, then why not.
What PCB are you talking about ??? and who is CNCdrives and what's their story?
Daniel
-
I would assume that heat treating is the easiest step in making a spindle.
What PCB are you talking about ??? and who is CNCdrives and what's their story?
Heat treating is easy enough if you know what your doing AND have the equipment. I sold the whole shop many years ago and the furnace went with.
Here on the Mach forum a search should turn up a thread where I did a review of several available servo drives. In that review I refer to a weakness in the error line of the DNCdrive products which prevented them from driving my optoisolated BOB reliably. I am not an electronics guy, but I came up with a sort of Rube Goldberg contraption that did fix the problem. CNCdrives showed me the better way to do it and provided the information I needed to design the physical PCB that would not only boost the error signal for an optoisolated BOB, but would also drive a remote LED indicator light that I insisted on having. Credit goes to them for the circuit, I just made the physical parts . . . which was actually big fun for me.
The circuit was born of those necessities, but it has many applicatons since it can take in a pretty wide vaiety of voltages and signal strengths and convert them all to a clean 5V signal, as well as split out a second identical clean 5V to power a remote LED from the same input. For example I had a lot od problems with false e-stops from the limit switches. One solution is to run the switches at 12V, but you then must convert that back to 5V for the BOB or you burn things up, so this little board is a perfect solution to that.
My BOB is a CNC4PC brand and has lots of LEDs on it. I love that whole concept of montoring what's going on, and in it's present iteration, Mach is pretty vaugue as to why it e-stopped and an indicator is not much good under the counter in a closed steel box. This little cicuit board can tap into literally any signal in your boix and drive a remote LED. I think it is a very important solution.
-
Ah.... Lucky CNCdrive to have customers like you :)
Routing PCBs on the mill is fun indeed. Did it once and really enjoyed the process.
Daniel
-
I am getting multiple inquiries on programming for the 4th axis and as much as I would like to I cannot always respond individually. I do try to respond to notices from this forum, so posting questions here is the best route. Here are the answers to the most FAQ:
Everything you see in the videos is using strictly the Mach mill program, never the turning program. There is only one setup and one profile. I wrote a simple macro to implement the MACH 'swapaxis' function to run the 4th axis as a spindle. (thanks goes to Hood for this tip). I recently learned to also turn on the mill spindle while the 4th axis was still in 'spindle' mode. You see that with the grinding video.
I also wrote macros to calculate and perform threading. I participated a bit in discussing theory on the threading topic but in practice I do not use the Mach threading, so unfortunately I can't help with that.
Primarily I design assemblies in AutoCAD and then isolate the parts, move them to 0,0 and dxfout the geometry of each part in this way. Then using LazyCAM to pick up the dxf and generate the basic Gcode. I then usually find it convenient to make resulting 'programs' into subroutines and set up variables and write a small 'control header' to call the subroutines in order, and that is pretty much the technique.
Turning profiles are sometimes generated for MachTurn and then run on the mill by swapping the axis around with a G-code command for that part of the program, again usually called as a sub. Sometimes it is easier to lift a turning profile directly off the design in AutoCAD and then simply generate code from that profile for the Z axis.
-
Thanks for the details! Cant wait to try some of that.
-
Hi Mike,
After reading your excellent trade study on servo drives, I've just ordered a Dugong Servo drive that I plan to use to drive my Lathe spindle which is a 400W 90Vdc motor.I've just had a thought. Do you know if the Dugong drive capable of being continuously run in 1 direction? Some servo drives fault after running too far in one direction.
Cheers,
Peter.
-
Hi Mike,
After reading your excellent trade study on servo drives, I've just ordered a Dugong Servo drive that I plan to use to drive my Lathe spindle which is a 400W 90Vdc motor.I've just had a thought. Do you know if the Dugong drive capable of being continuously run in 1 direction? Some servo drives fault after running too far in one direction.
Cheers,
Peter.
Peter, I suspect you are addressing that question to me, so I should tell you my first name is Steve.
There must be some upper limit of encoder counts that the drive can handle, but I have not found it yet. Good question . . I'll get an answer to this and post it.
EDIT: The answer is no, the register is 32 bits wide and rolls over when full so the drive will never fault from running on one direction too long.
One reason I liked the Dugong for a spindle drive is the built in braking resistor. A heavy chuck and workpiece can generate a lot of back current and the Dugong dumps it overboard instead of burning out the the drive. Using a servo motor on the spindle eliminates all of the threading problems and also opens the door for some interesting possibilities with cutting tools and grinding wheels attached to the cross slide, yes?
Incidentally, I have one of your excellent PWM speed controllers, which finally solved my speed control problems. Unfortunately, I can't use it with my 4th axis because of the swap axis function . . . :'(.
-
Hi Steve,
Not sure where the Mike came from. Sorry. I checked with CNCDrive, they confirmed that there should be no issue. There shouldn't be as a step/dir interface is a reletive one, as opposed to an absolute one.
With step/ Dir, it is move this increment from where I am.
I have one of the small servo drives from ImService. They fault after a short while of continuous running in one direction. Poor designas I see it.
Cheers,
Peter.
-
Hi Steve,
. . . . . a step/Dir interface is a reletive one, as opposed to an absolute one. With step/ Dir, it is move this increment from where I am.
Peter.
The drive maintains an absolute position internally. Some drives have a built in homing capability, which one might assume would complicate a 'roll over' of the position register, but I have no experience with those drives, so I don't know how they address that issue.
I find that there is no standard criteria in designing servo drives (speaking only of those that I tested). One manuf may consider an internal homing ability more important than continuous running, if in fact that is the trade off, while another may decide the opposite. Leadshine, as an example, thinks in terms of OEM only and concludes therefor that the ability to reset a faulted drive is not an important issue, because (in their view) once the specific 'single purpose' of a machine is set up properly, there should be no faulting. This unfortunate thinking leaves the Leadshine drives usesless for a 'general purpose' machine tool, in my opinion, even though they are otherwise very good drives.
These are the kinds of differences that make evaluating and selecting a drive much more complicated than is generally thought. Your 'continuous running' question is a very good and valid one that did not surface during my review unfortunately, so I did not investigate that ability in each drive. It is a caveat for sure.
-
Homing sensor moved to back and oriented down to keep out as much swarf as possible.
(http://www.thecubestudio.com/pictures/4thAxis/SensorCloseWEB.jpg)
I have my new Rutex servo drive and so far it is doing everything it claims to do, so I should get full speed and power from the motor. When I get it all back on the mill, I will make one final video and post the link. Questions, comments , criticisms or suggestions welcome.
may i ask what typ of homing/index sensor that is?
Cheers
Spunk
-
Hi Spunk,
Looks like an optical sensor. A photo LED is used on one side and a photo transistor on the other side.
Daniel
-
and perhaps also the answer to how do you wire it into mach3? you'll need to power the LED thats the easy part...
how to get the transistor into mach?
Spunk
-
The output of the opto just goes to a pin on your interface (Parallel Port, SmoothStepper or whatever) and then you set that pin up as the Index in Ports and [ins Inputs. Simpson also is using it as a homing switch so he will also have that pins set up as the A axis home switch, again in Ports and Pins.
Hood
-
Spunk,
As you say, you power the LED and the transistor gives you the output signal, which goes into Mach3 like Hood said. for any given sensor the manufacturer would supply a typical wiring diagram. You'll need to use resistors in line with the transistor and the LED to limit the current to the nominal. The resistor values (or the equivalent current) would usually be specified be the manufacturer.
The output signal from the transistor would usually be too low requiring some kind of an amplifier before it's fed into Mach. A simple solution would be using a schmitt trigger for that purpose.
Daniel
-
I use an Optek OPB917B it has everything you need except for a limiting resistor for the LED, 220 Ohm is what I use when powering from 5V.
Hood
-
Photointeruptors come in a lot of different flavors, some costing under US$ 1.00
Some have lenses over the LED and some do not. Some have wires attached and some have only bare pins. Some have polycarb housings and some use much cheaper plastic. Some are rated for vibration and some not . . etc. etc.
In order, I have had lots of false signals from stray light and dust with the un-lensed parts, pins break off very easily on the parts with bare pins sticking out (in fairness, these are meant to be mounted on a PCB, not have wires hanging off them), plastic housings that dissolve seemingly in anything except distilled water, and finally, some just litterally fall apart. The 4th axis was the 'acid test' that killed a number of these parts, some of which were actually working fine as homing switches.
After the above experience with various recommended products, I have settled on the following rather expensive unit US$ 7.25 approx. It has wire leads, lenses covering the optics (to keep out dust and swarf), an indestructible polycarbonate housing that can bend a .060 aluminum tab that gets out of line and hits it, and can tolerate vibration.
It is a Honeywell model HOA 1887-012
I now have these on all of my machine axis and also on the 4th axis and they will be used on the commercial verison of the 4th axis as well. You simply need a 100 to 150 ohm resistor on the 5V that feeds the LED (red wire) You can use 5V or up to 30V for the trigger side. For example if you use 5V from a BOB, you can daisy chain 3 of these and still have 3.5V return to the BOB. I have a little "Remote LED driver board" to boost signals, but it has not been necessary to use with these Honeywell interrupters and the Homann designs BOB.
-
You have done a great job, and I hope if you have time you show all your works. I think it's been a long time when you finish this study. I am very interested in what you have done.
-
Thanks!
The X2 featured here is now sold and gone. My new mill is scratch built completely from aluminum and is slightly larger than an X3. There are still some details to finish up and the high speed pulleys are not done yet, but I have just started cutting metal with it. I will be posting photos and videos soon. Unfortunately I do not have time to do a build thread on CNCzone or anything like that.
Here is the latest video which simply demonstrates the newly finished pneumatic spindle lock on the 4th axis. I now have a larger disk and an new belt guard (not shown in the vid) extends to cover the new disk.
http://www.youtube.com/watch?v=glYDe94hxb0
-
Hi Simpson
I am also building an indexer based on your design.
I think I may be using too small a servo (150watt) so was intending to use a brake or locking system.
I don't intend using heavy cuts but most of the parts will be 5" aluminium.
How does you brake work is it just by putting pressure on the outside of the belt?
Regards
Arthur
-
It is a disk brake. The disk is .062" 2024 aluminum and is intended only for holding and not for stopping. The servo motor provides deceleration. The mechanical disk caliper is from a 'pocket bike'.
The lock was 'added on' and not part of the original design, so it could be better integrated, but other than having the disk extend past the frame (which requires an elaborate belt guard), everything worked out satisfactorily.
The servo drive is a 'Dugong' model from CNCdrive and they are modifying the firmware for me so that the drive's max amps can be set back a fixed amount under software control. This will coordinated with the lock event. The purpose of this is to prevent the motor from overheating if it disagrees with the location of the lock event. Keeping the amps below the motor's continuous rating will prevent possible overheating and damage to the motor.
-
Here are a couple pics of the new mill showing the spindle lock. In this case, I may want to use the brake for stopping the spindle, so it is a thicker steel part.
(http://www.thecubestudio.com/pictures/MillSimpson/HeadDetailBackWEB.jpg)
(http://www.thecubestudio.com/pictures/MillSimpson/HeadDetailFrontWEB.jpg)
Next photo shows nearly fial assembly. Belt guards nor installed yet, etc.
(http://www.thecubestudio.com/pictures/MillSimpson/MillOnBenchWEB.jpg)
-
Thanks for the pictures.
I can see the brake clearly now.
You mention the problem of the servo dithering while at rest. If was only going to use the 4th axis as an indexer would I be better using a stepper motor instead?
Regards
Arthur
-
Quality job there Simpson :)
Thanks for the pictures.
I can see the brake clearly now.
You mention the problem of the servo dithering while at rest. If was only going to use the 4th axis as an indexer would I be better using a stepper motor instead?
Regards
Arthur
You could use a stepper but also with a servo you could have a macro so that when you apply the brake you disable the drive a split second later, that would solve any tripping problems.
Hood
-
IF you have no need for using your 4th axis as a lathe spindle, then it is a perfect application that takes advantage of a stepper motor's strengths, which have been discussed numerous times.
Flying Scott: You mentioned 'small' servo motor, but a direct comparison between stepper and servo power cannot be made based on published specs because they are not measured in the same way. Sort of like comparing Gallons to Feet.
Steppers are very good at holding (big steppers, that is). That is where the torque spec is measured and they do not dither or vibrate or have any 'slack zone'.
You need to also consider the drive mechanism involved. Many 4th axis are worm gear driven. I would speculate that this would swing the advantage back to the servo motor, where the vibrations would not be an issue, and the far higher speed of the servo motor would be a great benefit.
Hood: If we may refer to the drives that I reviewed as 'hobby level', then there are none that can have their amp output commanded in real time with Mach3 software. I'm not saying it is impossible, as there may be some method that I am not aware of, but I did investigate 'real time' reconfiguration via Mach3 at the time of the review and I know of none. CNCdrives is adding the capability and I don't know yet exactly how it will be implemented, but it will thru a macro for sure and will use the currently unused 'aux' connector on the drive. .
-
Hood: If we may refer to the drives that I reviewed as 'hobby level', then there are none that can have their amp output commanded in real time with Mach3 software. I'm not saying it is impossible, as there may be some method that I am not aware of, but I did investigate 'real time' reconfiguration via Mach3 at the time of the review and I know of none. CNCdrives is adding the capability and I don't know yet exactly how it will be implemented, but it will thru a macro for sure and will use the currently unused 'aux' connector on the drive. .
If this is with reference to my last post then I was meaning disable the drives, I would imagine most servo drives require an enable signal, but maybe not?
Hood
-
Very cool axis brake.
Let me just make sure I understand:
You've got a disc brake off a motorcycle or somewhere.
Instead of operating it via brake fluid, you've threaded a rod into the caliper that directly acts on the puck when your air cylinder rotates it?
I'm also curious about the dithering concern with servos. Yes, they dither, but are you seeing some effects in your work without the brake, or are you just concerned there might be? The reason I ask is that the dither is pretty darned small, should be no more than a step or two. I still would understand the desire for a brake to hold still for indexing.
RE stepper vs servo, stepper sure has a lot less resolution than an encoder. If you've only got a timing pulley reduction, it's kind of hard to get good resolution with a stepper. You can compare these resolutions versus various gear ratios in a spreadsheet I did for my own 4th axis design project:
http://www.cnccookbook.com/CCMillCNC4thAxis.htm
Cheers,
BW
-
Very cool axis brake.
Let me just make sure I understand:
You've got a disc brake off a motorcycle or somewhere.
Instead of operating it via brake fluid, you've threaded a rod into the caliper that directly acts on the puck when your air cylinder rotates it?
Cheers,
BW
Bob,
The disc brake is off a bicycle, or mini-bike. They are typically cable-operated. This is a GREAT application for those! I've considered putting a spindle brake on my mill, and one of those woudl be just the ticket.
Regards,
Ray L.
-
The disk brake caliper is mechanical from the git-go. These are small calipers for bicycles, ATV's, mini bikes, etc. I purchased several varieties and the one featured on the 4th axis happens to be for what is called a 'pocket bike', which is a ridiculously tiny motorcycle. They all have similar internal cam arrangements.
I found the quality of the calipers to be unacceptable, so I had to extensively modify it. So much so that in the future, I plan to use the brake pads, but fabricate the caliper from scratch.
The disk is cut from 2024 of 7075 aluminium sheet. I must emphasize that this is a consideration for mass and these disks are not intended, nor capable of acting as brakes to slow a load. The disk on the milling machine spindle is another 'pocket bike' item. It simply needed a small hole for the sensors and have its mounting hub removed. It is the smallest pre made disk I could find at 5.5" OD
The servo motor will *try* to maintain one count on an encoder, but that is not going to happen if the load is too great and the amps too little. I have 35amps available and I can force the motor off center by hand with less than a 12" lever. The normal dithering is tiny with my 1800 line encoder (7200 with quadrature), but 'tiny' is relative. Cutting the teeth on the GT3 timing pulley was done with a tiny 4 flute ball end mil ll at 28,000 RPM. If you have a chip load of .0002, even a .001 oscillation is significant.
My 'anticipated' issue is off center drilling of fairly large holes in stainless or tool steel. The force needed for that would be greater than it takes me to move the spindle off hold. This is a heavy unit intended for 24/7 production, so it needs to be able to handle tough jobs.
-
Hood: If we may refer to the drives that I reviewed as 'hobby level', then there are none that can have their amp output commanded in real time with Mach3 software. I'm not saying it is impossible, as there may be some method that I am not aware of, but I did investigate 'real time' reconfiguration via Mach3 at the time of the review and I know of none. CNCdrives is adding the capability and I don't know yet exactly how it will be implemented, but it will thru a macro for sure and will use the currently unused 'aux' connector on the drive. .
If this is with reference to my last post then I was meaning disable the drives, I would imagine most servo drives require an enable signal, but maybe not?
Hood
Some can be stopped with an external signal, but if I remember correctly, restarting any of them requires a reset and they loose reference. None of them track the encoder during a fault, so even if you could restart them, results would be unpredictable. methinks.
If I recall correctly, you have high end stuff and it may have a more sophisticated capability in this regard. I do know that some drives continue to track the encoder even when they are faulted . . . totally logical . . . and totally missing from the low end. Of the hobby level drives I tested, only the CNCdrive product was acceptable and it can be stopped, but must be reset to start up again. Unlike others, at least the CNCdrive products can be reset without powering down the whole system and waiting for all the CAPs to bleed off.
-
Yes my drives are Industrial type and do track the encoder in all states, I just presumed even the cheaper drives would need an enable signal which you would use to allow movement but could also take that signal away and the drive would stop holding the motor. If that was the case and you were using a brake then the motor couldnt move anyway so as long as you enabled just prior to releasing the brake it should not lose position.
From what you say though it looks like that is not the way the hobby drives work and if you need to power down before re-enabling its not very good and obviously what I was thinking will not work..
Hood
-
I have looked far and wide throughout this thread and more .
I cant find how its achieved running
both spindles togeather without the axis swap.
Iam after a solution like the very first post,ie
controlling A axis as a continuous spindle (lathe mode) and also using a router spindle
at the same time.
Any info would be great.
Regards
Whatda
-
You can run a router and use the 4th axis in 'lathe' mode now unless you want to use Mach to control the router speed.
This project is alive and moving forward. The current 4th axis version will become a standard part of a new commercial machine and I am currently building a new 'Super Duty' version with double reduction and a robust steel disk brake. I am pursuing a couple of different options for controlling the spindle without using the swapaxis function.
One possibility is a solid state device which will switch the servo motor's controller from a servo drive to a speed controller.
The second involves another modification of an existing servo drive to perform the speed control function internally and outside of Mach, but using an output of Mach to engage the 'lathe' mode.
A third is to use a pulse generator to create the pulse stream for 'lathe' mode and only switch the encoder output between Mach and the pulse generator for 'indexer' mode and 'lathe' mode respsectively.
There are advantages and challenges with each method. It is an interesting challenge Anyone who wishes to share some creative ideas on this challenge is certainly welcome to do so.
-
Thanks for the info.
".......You can run a router and use the 4th axis in 'lathe' mode now unless you want to use Mach to control the router speed."
How is this Achieved. ????
Regards
Whatda
-
Simply use Mach's 'spindle' to run the 4th axis and just operate the router off a manual switch (or via a relay if you want CNC control).
-
Thanks,
But my router is controlled through a Vfd and would run at 18000rpm all the time.
What I would like is the opposite ,ie the rotary axis ran at max speed and then be able to control
the router spindle speed.
Iam thinking about just conecting a simple single phase motor thru a series of pullys to the 4th axis and
some way of disconecting the belt so I have a indexer and a lathe spindle,not high tech,but it should work.
Regards
Whatda
-
Steve, could you post the gcode just for an example on how to do it. No need to add comments.
Thanks
Bob
-
whatda: My 4th axis is being made into a standard part of a larger new machine being devoloped now. For that project, I am developing a hardware 'swapaxis' that will work outside Mach so that my 4th axis is not tied to using Mach. There are also other people who want the 4th axis and do not use Mach3. That may do the trick for you.
Also I am deveoping a 'Super Duty' version of the 4th axis which has a 'back gear' double reduction belt setup for heavy simultaneousl 4 axis milling and for high toque applications like drilling large diameter holes in the center of tough material or turning large diameter stock. The high speed capability remains and switching between the two is a simple belt move. You may be able to do something similar, but swap motors instead of reduction sets. I'll be posting videos soon on Youtube when the new 4th axis is finished so you can see how it is set up.
stnc: What is it exactly that you want to know how to do? The swapaxis function is part of Mach3 and is not done in g-code. I wrote a pair of macros to call it. You call the macros from G-code using the 'M' command. It that what you are after?
-
Steve, that's right. I don't know how to call the macro.
Thanks
Bob
-
As an example, my macros at issue are the following. I have posted the macros previously, but if you can't find them, I'll be happy to post them again. Note that the numbers are not special, you can use any unreserved number you want:
M951 initiate swap axis (A axis becomes Spindle and Spindle becomes A axis)
M950 Cancel swap axis and re-home the A axis
M961 apply spinfle lock (just activates an output that I have controling a relay which in turn controls a solenoid valve which in turn operates the air cylinder)
M960 release spindle lock
Example of G-code would look like this (Note that 'M' codes call external macros):
N50 M951 (initiate swap axis)
N55 M3 (turn on spidnle)
<add your turning code here>
N65 M5 (turn off spindle)
N70 M950 (un-swapaxis and re-home A axis - note that the A axis has lost its reference due to being a spindle, so re-home puts it back to a known zero)
N75 G0 A45
N80 M961 (engage spindle lock)
<cut a tooth, drill a hole whatever>
N150 M960 (release spindle)
N155 G4 P.1 (delay *might* be needed depending on how fast the lock is vs your accell of the A axis)
N160 G0 A90
N165 M961 (engage spindle lock)
etc, etc
-
Thanks alot Steve, thai's what i needed. :D Good luck on your up coming project.
Bob
-
Update on the swapaxis question: the third of my little PCB projects is now complete and tested . . .
The new 'SWAPAXIS' hardware board is completed and working!
This should work with any CNC software and also with the Smoothstepper. The rehoming macro still works, but the swapaxis command is now superfluous, of course and will be removed.
I extend an offer to Hood, who has helped me a lot with my projects, to receive a free swapaxis board to keep in exchange for testing it with smoothstepper. If Hood declines then the offer is open to anyone with a smoothstepper who is willing to do some testing.
I am currently accomplishing the axis swap via the M9 command for flood cooling. I have the output mapped to the SwapAxis board. ANY output will work fine, I just used a 'standard' CNC output that every system should have available.
Photos of the little PC board follow. The .032" end mill I normally use was too wide for the very small traces needed for this little guy, so I did my first rounting with a 30 degree router bit to cut the traces. NOw I have 'ebgraver fever' and I want to engrave everything. Trying to figure out how to fixture the dog . . LOL!!
(http://www.thecubestudio.com/pictures/PCBprojects/SwapAxisCloseWEB.jpg)
(http://www.thecubestudio.com/pictures/PCBprojects/SwapAxisBackSideWEB.jpg)
-
Simpson, looks good :)
Thanks for the offer but I am not sure whether I can use this or not. I use differential Step/Dir signals and the board I made for that connects direct into my BOB (see pic)
I presume your board just accepts single ended step/dir inputs?
Hood
-
Hood, every time you speak, I realize how much I have to learn ???
I am aware of differential signaling being used for encoders (I use that myself), but are you saying that your servo drives also use differential step and direction signals? Does smoothstepper support differential step and dir signals?
My board does not really 'accept' anything, it is just a signal switcher. It would not have enough lines for differntial signals, but I suppose you could use two of them. I don't think I would be inclined to pursue that unless your setup is common. I have not heard of using differential signals for step and dir. What would be the purpose except to mount the drives far away from the BOB?
-
Most industrial rated drives have differential Step/Dir inputs, they can usually be used single ended but I prefer to make my machines as noise immune as possible so I chose to use the differential capability.
The SmoothStepper doesn't have differential Step/Dir outputs, future hardware may be different though ;)
Wouldnt think many people use differential Step/Dir as I dont think any of the non industrial servo drives support it.
Hood
-
Is correct, output diferential in step/dir signal is more resistant to the noise that step/dir a ground.
I did the test putting the signal step/dir single ended hit to the motor and it lost many pulses,
but the differential sign didn't lose none pulse.
regards
-
OK, since the smoothstepper does not support differential signals, you must be doing the conversion downstream. My swapaxis board simply switches the signals unchanged, taking inputs from two axis and swapping them to the servo drives. The switching is very fast and should be transparent to the system, particularly to the smoothstepper driver.
Mach3 is unaware of the swapaxis board and even 'hot switching' with the axis moving seems to work OK. I have tested the board with the spindle axis active and running the spindle. Switching the axis in this situation simply starts the 4th axis running as the spindle. Switching back does not seem to confuse the system at all. The 4th axis is simply, silently, the A axis again, with the caveat that the position is no longer valid, so indexing requires a re-home. Also a 'hot swap' does not utilize accelleration settings to the axis sort of 'slams into gear' without a clutch . . . so to speak, if you hot swap.
It seems to me that your setup *should* work like this: CNC->SMOOTHSTEPPER DRIVER->BOB->SMOOTHSTEPPER->SWAPAXIS BOARD->DIFFERENTIAL CONVERTER->SERVO DRIVE
Now that you have made me aware of differential step and direction signals being used on industrial drives, I have yet more homework to do. My new 'Super Duty' 4th axis is almost complete and it is not a hobby level machine, so I will need to support differential signals in the accompanying swapaxis board, so I will look into what is needed to accomplish that. Hopefully it will be something simple like dual multiplexers triggered simultaneously.
-
Simpson, if you look at the pic I posted you will see that the board I made connects directly into my PMDX 122 BOB with solid copper legs so no way I can take single ended signals off without removing that board :(
Hood
-
I saw the picture. Very professionally done and neat as always, but not welded together from what I could see . . . ;)
Hood, it was my impression that you have several or many machines all with smoothstepper and that you were interested in swapaxis capability except that the smoothstepper did not support it. I think it was you who said that anyway. Apparently one or more of those impressions in incorrect. Incidentally, I believe the swapaxis board could also go between the smoothstepper and the bob.
Eventually I will test is myself, but I am too busy right to inject the reportedly troublessome smoothstepper into the mix. The board works great between the BOB and drives using Machs port driver, And I have sent some out for testing on specific systems that will be getting the new 4th axis. The swapaxis function is critical to the usefullness of my 4th axis so all three of the little boards (TTl splitter/booster, relay and swapaxis) are being supplied with the machines. Several guys are building their own 4th axis based on this design and I may make these board generally available for people who tackle this themselves. None of these guys has a smoothstepper, but I just would like to know if it is compatible for future reference. It *should* be, but it would be nice to hear it from someone who tested it.
So, anyone with a smoothstepper or equivalent (if there is one) and preferably also a 4th axis, send me a PM if you want to play and I'll send you a free swapaxis board to test and keep.
The machines themselves are coming along nicely. The lock is now also a brake and is completely enclosed in the frame. This is the Standard model. I'll be posting a new video soon of the Super Duty guy with the double reduction belts. That guy is a brute. I can't keep the main drive pulley locked onto the spindle . . so that model is 'still in the oven' so to speak until I get that issue resolved.
(http://www.thecubestudio.com/pictures/4thAxisSTD/FrameEngrave01WEB.jpg)
-
Wow, very professional looking unit. That's the bearing right there in all its glory, eh?
I'd be tempted to put some sort of ring shaped cover over it, even though it is sealed.
Cheers,
BW
PS I think US Digital has relatively inexpensive and compact converters to go from single ended to differential and back. Like everything else, it's just a chip.
-
Wow, very professional looking unit. That's the bearing right there in all its glory, eh?
I'd be tempted to put some sort of ring shaped cover over it, even though it is sealed.
That's a good suggestion, thanks!
PS I think US Digital has relatively inexpensive and compact converters to go from single ended to differential and back. Like everything else, it's just a chip.
Yes they do, and Rutex also sells them and CNCdrives gives them away free with their drives. The converter boards are very simple (right up my alley :-[), but I would prefer to support the differential signal directly on the swapaxis board, if it is not too complicated . . . I can forsee just switching the extra signals or perhaps adding the differential converter chip to have a 'combo' depending upon what people want, but I just don't have time to address that at the moment. You are right about 'just a chip' :D My little TTL splitter and also the Swapaxis are just chips also. The boards are basically just a 'connector' for the chips. The solid state relay is a little more complicated, but still pretty basic stuff. Anything electronic is a challenge for me and I have help in figuring these little guys out, but still it is fun and satisfying when they actually work! The PCB routing is still fun and has now lead me into engraving . . which is also big fun! CNC is a very addictive hobby, I must say.
My current 'rouge's gallery':
(http://www.thecubestudio.com/pictures/PCBprojects/ThreeBoardsWEB.jpg)
-
Good looking boards, Steve and the 4th axis is just professional!
How did you do the printings on the boards?
Daniel
-
Dan, I tried a number of methods including special ink in an injet printer and using a modified DVD printing tray, but in the end I settled (for now) on common rubber stamps using a 'permanent' ink. I would rate the results as 'satisfactory'. Silk screen would provide a better result, but I don't really want to mess with silk screening. I am however contemplating purchasing a pad printer for another project. If I go forward with that, then it would also do an outstanding job on the PCBs!
Hood, It has occurred to me that your smoothstepper is ahead of your BOB, which makes sense considering the smoothstepper uses a driver to 'talk' to Mach3. The swapaxis should still work between the drive and the BOB -or- between the BOB and the smoothstepper, but in opposite order of the flow that I originally assumed.
-
Ah... Interesting! The results sure look good. Do you engrave the words in the rubber first, or how is it done...?
Daniel
-
Looks great :)
Yes I would like the Swap axis feature but without rewiring things I couldnt really use your board. The computer connects to the SS via USB then the SS connects to the BOB via a short ribbon then I have my Diff board direct into the BOB. I could hack into the ribbon between SS and BOB but would rather not do that as it would be a chink in the armour where noise could be introduced.
My Lathe has a similar setup except the BOB itself has differential outputs so again it would be the ribbon that needs hacked into.
I could use it on the Bridgeport but dont have a servo spindle on it, in fact Mach doesnt even control the on/off of the spindle, its totally manual. I am about to re-do the Bridgeport to use servos and have a servo for the spindle also but when I will get a chance to do it I dont know.
The coil winder has steppers and is single ended but its not mine, I made it for a customer and its in his premises so cant really alter it either.
Hood
-
Actually thinking about it I probably could test it out on the Bridgeport the way it is, I would just have to use it to swap an axis with another axis rather than spindle and A.
Does it swap both axis or does it just divert one axis to another. What I mean is does it for example take X and Y in and when you swap it puts X out to Y and Y out to X or does it just take X in and either put it out to X or Y?
Hood
-
Based on posts and PMs I will clarify a couple of things about the swapaxis question.
First, the swapaxis board does exactly in hardware what the Mach3 software swapaxis does, namely it 'swaps' the axis. Two axis inputs, two drive outputs. The board is invisible to Mach3. This caveat is that a user needs to be aware of the acceleration and step settings as they will not follow the axis swap, so a litte planning would need to be done, or adjustmnent made in the G-code.
This is not a complicated install. We are talking about taking 4 wires out of a $2 ribbon cable, cutting them and connecting them to a 1" wide PCB using a tiny screwdriver. Add a 5V supply, a ground and an on/off output signal from your BOB and it's done. In the unlikely event that there are not enough pins left on the BOB, a second serial port is about $15 and basic BOBs are under $10
-
A tester has volunteered who has a servo driven lathe spindle and has been trying to get swap axis for it for quite some time. Thanks to all who who responded.
I am not promising to make the boards generally available, but I will let everyone know the results of the smoothstepper testing. This will probably take a couple of weeks or more.
-
Hi, In the macro the swapaxis(6,3) What are the 6,3 refer to?
Thanks
Bob
-
Hi Bob,
The numbers in the SwapAxis command are numerical designations of the axes. AS far as I remember it goes like this:
0 -> X axis
1 -> Y axis
2 -> Z axis
3 -> A axis
4 -> B axis
5 -> C axis
6 -> Spindle
Daniel
-
Hi. I guess I'm not doing something right.
This is the gcode i'm using for testing the 4th Axis servo motor
M940
SwapAxis(6,3)
M3
Y.5
Y0
M5
M30
The servo does not run? Y Axis moves.
Does the servo require a Relat to run?
Thanks
Bob
-
Are you using the SwapAxis command straight in your gcode??!! You have to make up a macro that would contain the SwapAxis command and you would call it each time you need a swap the axes.
Daniel
-
Yes, I'm using Steves macro, M940.
I made a typoO in the last reply.
Do i need a Relay for the 4th axis Servo motor?
Bob
-
What is it exactly you want to do, Bob?
Daniel
-
I just want to run the 4th axis as a Spindle.
Bob
-
OK, you have to define your A axis - its pins and tune it under motor tuning AND enable it. Also do the same for the spindle - define the pins and tune it AND enable it, also enable the M3/M4 relays under ports&pins -> spindle . Then, if the macro you're using has the SwapAxis(6,3) command you should be able to program M3 S#### and the 4th axis would rotate at the commanded speed as if it were a spindle.
Daniel
-
Thanks alot!
I added the S500 to the M3 and it worked.
Bob
-
First video is posted on Youtube. Shows new Super Duty 4th axis. Rotary engraving demonstrated.
http://www.youtube.com/watch?v=sC2urdg_bqg&feature=related
Second video posted:
http://www.youtube.com/watch?v=zrFR_lwZBRA
Demonstrates spindle lock power during metal shaving operation.
-
Looks to be working great and now you have a CNC slotter/shaper so internal/external splines and keyways will be easy :)
Hood
-
Really nice work there- :)
Dave
-
Great demonstration and perfect videos as always!
Is it an oscillating head you're engraving with? Wonder why the engraved lines look black...?
Dan
-
It actually looked like a magic marker to me? Thought it was just a test and not a true engraving.
Dave
-
Thanks for all the comments. Couple responses:
The black 'engraved' line is indeed drawn by a pen for the purpose of 'demonstrating' an engraving process. I use the pen often to proof new programs. I have not had time to build my tail stock yet, so cutting on the unsupported end of a foot long tube would probably not be successful. The engraving video actually was done with belts set for high speed, but it still made pretty smooth curves. The 'backgear' setting is specifically for high accuracy (very smooth arcs) and/or high torque (drilling, trunnion table, etc). I am discovering that Mach has 'issues' with simultanwous 4 axis motion.
The tail stock will be built from a standard spin indexer and have very light duty ball bearings and a 5C collet. In this way I can pas stock not only the 4th axis spindle, but also the tailstock and work on very long pieces. I have an adapter from straight shank to MT3 that can be held in a 5C collet and then use 'normal' tall stock tooling. The tailstock spindle will have a conventional handwheel advance for drilling, etc.
As mentioned in another post, the shaving operation is really most usefull for internal stuff, but that doesn't make a very good video, so I just did something external and simple again for demonstration. Many of the videos on YouTube make me want to slash my wrists, so I really try to keep mine short and reasonably interesting.
-
Comment on the new spindle lock/brake for those of you who are folowing along and making your own 4th axis;
The new setup is completely enclosed in the frame and under cover so that flood cooling can be used. The disc is now hardened steel and the caliper is larger and more robust so that the setup can serve as a brake as well as a lock. A slightly larger and longer travel air cylinder is used. Repeated emergency stops with or without motor power are accomplished quickly and without damage to the brake.
With the new brake covered up, I found it impossible to tell if the lock was engaged or not, so I now have an indicator light next to the manual override button. I also suggest beginning each program with an unlock command. The motor is unable to turn the spindle against the new lock, however, in 'backgear' the torque was such that the setscrews on the main pulley were immediately dragged around the hard spindle, even after I ground small flats for them to bear on. This has been solved by going back to the steel clamp collar with the aluminum pulley bolted to it.
The current 'lock demo' video will be taken down as soon as I have a chance to make and post a video of the new lock, so be aware that the lock as shown in the current 'lock demo' video is NOT capable of holding the spindle for a shaving operation as demonstrated in the latest video. Sorry for any confusion there.
-
News:
SwapAxis board has been tested with the smoothstepper and it works fine!
-
Good to hear :)
Hood
-
Here is a fun accessory for anyone using a 4th axis:
http://www.cncwrapper.com/
Works great so far!
-
Hey Hood! Congratulations for hitting the 10,000 posts mark ;)
We appreciate your contribution!
Daniel
-
Thanks Daniel,
I got a lot of help when I first started, especially from John Prentice, and I am glad to be able to help others if I can.
Hood
-
Yes, when it comes to the development of Mach John Prentice is a legend, up there with Art.
Cheers,
Peter.
-
My hat's off to Hood also as well as Himmy and many others for the generous time they spend helping others. My 4th axis project would have died without the 'swapaxis' tip which came from Hood. Instead, the project is slowly maturing into an actual useable product with several, possibly many people following along and duplicating it for themselves . . not to mention a spill over into servo driven lathe spindles which are borrowing the techniques.
I think this is quite a success and the kudos go to a lot of people, not just myself.
Latest neat addition is an indicator light to show when the 4th axis (or lathe spindle) is in 'indexer' mode. An indicator monitoring the lock status has been added and is shown in the latest video.
Speaking of which, I posted another video to answer some requests to demonstrate more shaving operations, particularly splines (gear teeth, pulley teeth, keyways, etc are also made the same way):
http://www.youtube.com/watch?v=B8l6lH4ydd4 (http://www.youtube.com/watch?v=B8l6lH4ydd4)
I think that's going to be about it for a while. I am seeking suggestions on motion controllers/pulse generators equivalent to smoothstepper.
-
This is just a heads up for anyone seeking the currently backordered Dugong servo drives.
CNCdrive is making specially modified Dugong drives for my 4th axis project.
I have three brand new standard Dugong drives left that I will not be using and I have them for sale in the Bargain Basement here.
These are new and unused and CNCdrive is giving full new drive warranty to the buyer of these drives.
-
4th axis prototype (previous model) for sale.
The new Super Duty prototype is finished and running, so I am finished with the previous prototype. It is complete and working and ready to do 24/7 production work. I would rather see someone using it than to disassemble it for parts, so I have it for sale complete in the Bargain basement. I am only listing it in this forum initially.
The 3 Dugong drives are sold.
I am finishing modification to the new mil and some interesting new stuff will be coming soon on the new Super Duty 4th axis and 'semi-mini machining center'. Stay tuned.
-
4th axis prototype is sold.
Currently I am working on making the 4th axis spindle lock function automatic and transparet to the CNC control so that a user does not need to manually addi macros to the G-code. There are a lot of ways to skin that cat and I would be interested to hear any ideas on how to accomplish that goal.
-
I see 2 problems to overcome.
1. Mach is a buffered system all moves are planned adn then sent to the buffer for output. Once in the buffer there is little interaction back to mach as to when a Aaxis move is being sent out.
2. Latency of the mechnanism. You will need to be able to "start" to deactivate the brake long BEFORE mach makes the actual move.
There is a signal mach has "current hi/low" that can trigger a relay on axis movement, we use it to drive the timer for the lube pumps, BUT by the time the Brake reacts it it will be too late. Servo error or lost steps would be the results
We worked on this for a while a long time ago. Even Art suggested it would not work without driver level intervention.
NOW what we did get to work fairly well was rewriting the CAM post to enclude the Brake function before and after the index move that worked well for CAMed situations. BUT it could not do interpolated 4th axis moves well. Abig mess actually.
BUT you never know, maybe ????
-
Presumably you just want to look at the step/dir train incoming, unlock when it starts, and lock again when it stops.
The logic is pretty easy, but I'll bet the issue will be the time it takes to lock and unlock (that latency) and whether that creates a problem for your g-code programs or for the electronics. The electronics are just a matter of buffering the pulses until locked/unlocked and then resuming playback.
Issues with the g-code program would arise from trying to do coordinated moves between the axes. I don't see how to fix that without going to macros.
I guess what I would do is use an M-Code to close a relay that turns "indexing mode" on and off. When its on, the locking is enabled and the lock synchronizer will function whenever it sees step pulses. When off, the 4th axis stays unlocked for continuous machining.
That's my off the cuff. I'm sure I missed a lot. Will be interesting to hear more.
Best,
BW
-
2. Latency of the mechnanism. You will need to be able to "start" to deactivate the brake long BEFORE mach makes the actual move.
There is a signal mach has "current hi/low" that can trigger a relay on axis movement, we use it to drive the timer for the lube pumps, BUT by the time the Brake reacts it it will be too late. Servo error or lost steps would be the results
Yet lateency is an issue, but only with unlocking. Right now I use macros to engage and disengage and I have delays built into the macros. A G4 command can also be used in the G-code immediately preceding the move. My goals is to eliminate the need for manually inserting stuff into the G-code. To do that I need to find some available trigger that occurs at the correct point and use that to operate the lock.
What Mach signal are you using as a trigger?
-
Presumably you just want to look at the step/dir train incoming, unlock when it starts, and lock again when it stops.
Yes exactly, but only while in 'indexer' mode. In 'lathe' mode (including all simultaneous axis moves that include the 'A" axis), all locking would be disabled . . except perhaps for an e-stop.
I guess what I would do is use an M-Code to close a relay that turns "indexing mode" on and off. When its on, the locking is enabled and the lock synchronizer will function whenever it sees step pulses. When off, the 4th axis stays unlocked for continuous machining.
Correct again. My thought is to rig two 'modes' for the locking function. When the 4th axis is in 'indexer' mode, the automatic locking feature would be active and in 'lathe' mode all locking would simply be disabled. This would be easy to trigger simultaneously with the swapaxis hardware. i.e. I could split the signal to the Swapaxis SELect pin and use a mirrored signal to enable the lock solenoid, or something like that.
It is an interesting challenge.
-
I thought you had three modes:
1. Full A axis mode
2. Indexing mode
3. Spindle mode
Now you're saying you will need the locking only in the "indexing" mode. Can't see how the program "knows" whether you're in "indexer" mode or in "Full A axis mode"? Looks like a macro is needed...?
Dan
-
Current Hi/low is an ouput signal that can be mapped to pin/port, look in config,pin/port,output. BUT it signals on ANY axis movement and is WAY too far behind the data stream time wise.
I have seen this done from a motion card and a realtime system. I cannot fathom it working from mach(;-( WITHOUT the use of a macro. You would have to be able to TAP the function into the DRIVER level. AND as Art said once "That aint gonna happen".
BUT I will keep the thinking cap running(;-)
-
Is this something that could be configured into the post processor ?
Just rambling.
Russ
-
I thought you had three modes:
1. Full A axis mode
2. Indexing mode
3. Spindle mode
I don't know where this came from, but it was not from me. I don't recall ever saying 'Full A axis mode' and I really don't even know what that would be.
There is no point in getting into a semantics argument. There are no defined 'modes' so you can call them whatever you like, but there are only two. There is an A axis and a Spindle axis. One has postion and one does not.
-
Current Hi/low is an ouput signal that can be mapped to pin/port, look in config,pin/port,output. BUT it signals on ANY axis movement and is WAY too far behind the data stream time wise.
I have seen this done from a motion card and a realtime system. I cannot fathom it working from mach(;-( WITHOUT the use of a macro. You would have to be able to TAP the function into the DRIVER level. AND as Art said once "That aint gonna happen".
I'm not familiar with the 'current hi/low', but my ignorance there is irrelevant since by your description, it would be unusable in any case. Sounds like you have a visual on the challenges.
In order to function, it seems to me that the trigger must come from the servo drive. So I would need a drive that has or will have some function that identifies a condition where the lock should be applied and released. Custom firmware is certainly one path. A 'position reached' function looks like a promising possibility.
Another poster mentioned the need to disable the drive while the spindle is locked. The 4th axis has a belt drive and the lock is on the spindle and *so far* there is enough 'give' in the belt that the drive is not too unhappy with a count of two off an 1800 line encoder . . . BUT . . I am also pursuing the solution mentioned where the drive is disabled during the lock event and re enabled. I am waiting to get one question answered by a vendor to determine the feasibility of that solution.
Overloaded: you are exactly correct. Right now I have to manually insert the macros int the G-code. Automating that process would require messaging a post processor, as you correctly point out. I want to avoid this. That is what drives the goal to make it automatic and transparent.
-
Doing it at the servo drive level won't work either, Tried that as well using the servo brake
Anything you see at the drive level is already BEHIND the time curve of the buffered data AND mach has no idea what is happening at the servo level movement wise so it cannot STOP the data flow out of the port.
-
Doing it at the servo drive level won't work either, Tried that as well using the servo brake
Anything you see at the drive level is already BEHIND the time curve of the buffered data AND mach has no idea what is happening at the servo level movement wise so it cannot STOP the data flow out of the port.
My goal it so make the function transparent to Mach so what mach knows or doe not know would be irrelevant. What is or is not in a buffer would also not have any practical effect.
By way of example, a hypothetical scenario would go like this:
Premise:
4th axis is in 'indexer' mode connected to the A axis drive. G-code commanded position: G1 A425 The servo drive only sees A axis steps and no other.
1)Upon reaching azimuth 425, the drive reports 'position reached'
2)That signal (+5V) is used to trigger the lock ON (** and perhaps simultaneously disable the drive)
3)new A axis steps come into the drive.
4) within 1ms (I am told) the 'position reached' line goes low and the HV to the motor come on (**IF the drive is enabled)
5) that signal (0V) released the lock (** and perhaps enables the drive)
6) Axis moves to new azimuth and the process repeats.
Other than the axis attempting to move before the lock can release, I don't see a fatal flaw in the process, but I would definitely want to know if anyone else does.
Can you elaborate more on what it is you tried at the servo level?
-
I thought you had three modes:
1. Full A axis mode
2. Indexing mode
3. Spindle mode
I don't know where this came from, but it was not from me. I don't recall ever saying 'Full A axis mode' and I really don't even know what that would be.
There is no point in getting into a semantics argument. There are no defined 'modes' so you can call them whatever you like, but there are only two. There is an A axis and a Spindle axis. One has postion and one does not.
Sorry, a bad interpretation of mine of your previous post ;)
Dan
-
Other than the axis attempting to move before the lock can release, I don't see a fatal flaw in the process, but I would definitely want to know if anyone else does.
Exactly! This is the main problem and especially because of the requirement to make it "transparent" to Mach. You got to have something in Mach that would trigger just before it starts sending pulses to the 4th axis. Once it's started sending pulses to the drive it's too late to try and do something on the hardware level to unlock the brake. That is unless may be if you have a servo with an incorporated brake and some clever drive that can take care of it all.
Dan
-
Here's a thought, probably not fruitful...
First, you need to know precisely how long it takes the brake to lock/unlock.
Second, what's the pulse rate and how far does it make your 4th travel during that lock/unlock time.
Remember, your axis must also accelerate and decelerate, so it will be moving slower than the rapids rate when it starts or stops.
Now the probably useless thought:
Given that information, can you just let the servo drive catch up? It's a function of the pulses during the lock/unlock not exceeding the follow error and causing a servo fault. Obviously a drive that allows you to adjust that would be helpful.
Depending on your feedrates, it might be workable. It surely would be simple if it is workable.
Cheers,
BW
PS As you have surmised, lock is probably ok, its unlock that is touchy.
-
Other than the axis attempting to move before the lock can release, I don't see a fatal flaw in the process, but I would definitely want to know if anyone else does.
Exactly! This is the main problem and especially because of the requirement to make it "transparent" to Mach. You got to have something in Mach that would trigger just before it starts sending pulses to the 4th axis. Once it's started sending pulses to the drive it's too late to try and do something on the hardware level to unlock the brake. That is unless may be if you have a servo with an incorporated brake and some clever drive that can take care of it all.
You do not have to do anything in Mach. I already know this can be done with custom firmware from at least two different vendors, so there is no need to discuss whether it is doable at all, only what approache might be best and why. This is just a dialogue and I am interested in what others have tried and what their issue were.
To addresss your comment specifically, if in theory the lock could be released instantaneously with zero latency when the very first new step crossed the drive's input terminal, then you would have to agree that everything would work just peachy, yes?
So then it is actually only a matter of timing. A servo motor driving a 4th axis with a workpiece does not instantaneously accellerate even to a low G1 speed. If for example, the actual accelleration of the axis is '*********' and we assume for the sake of argument a zero latency for the electrons flying around within the drive, then we only need consider the time to move the physical plunger in the pneumatic solenoid valve and the time it takes for few CCs of air to move thru about 5" or 1/8" tube to atmosphere. I have not measured either of these, nor do I have equipment to do so, . . BUT . . it IS working now using macro commands embedded in the G-code. I started with delays in the macros and then also have used G4 delays. I have whittled those now down to nearly zero and the lock seems to be working fine. There remains a sleep of only 10 in the macro, but that is there to wait on MACH to do something, not the lock mechanism
Interestingly, it has occurred tome as I am typing that I have a video of the entire cycle which I could simply count the number of frames needed and get at least a rough idea of the latency. Look here and observe the red indicator light which signals that the lock is engaged.
http://www.youtube.com/watch?v=B8l6lH4ydd4
-
Here's a thought, probably not fruitful...
First, you need to know precisely how long it takes the brake to lock/unlock.
Second, what's the pulse rate and how far does it make your 4th travel during that lock/unlock time.
Remember, your axis must also accelerate and decelerate, so it will be moving slower than the rapids rate when it starts or stops.
The only part of this that is not 100% accurate is where you say it may not be fruitfull. You are exactly on the mark.
Now the probably useless thought:
Given that information, can you just let the servo drive catch up? It's a function of the pulses during the lock/unlock not exceeding the follow error and causing a servo fault. Obviously a drive that allows you to adjust that would be helpful.
Depending on your feedrates, it might be workable. It surely would be simple if it is workable.
This is my conclusion as well. By both theory and observation. You give me another idea of how to quantify. The Dugong drive has a real time Encoder DRO in its tuning software which can be active while the drive is ruining . . a very neat feature that I have used many times. I could simply monitor the error during program execution. I don't know why I did not think of that :-[. . . so . . I'll just blame Dan . . . Dan, this is all your fault . . ;)
PS As you have surmised, lock is probably ok, its unlock that is touchy.
Batting 1000
-
"THE "problem we ran into was that the servo BRAKE as fast as it is was way too slow in releasing the brake to prevent the servo from erroring on position error.The next moves were already at the drive BEFORE the brake could react. You have to be able to regulate that data stream of pulses to alow time for the brake to react BEFORE it gets the next move command. That fact that we had high resolution servos that require high step feedrates may have contributed to the problem 45000 steps per inch.
At full speeds that is not a lot of time in between pulses to activate the brake(;-)
Now sometimes it would work fine depending on the code and other times it would error depending on the code. TO me if it cannot operate 100% of the time it is a bust.
OF course IF you used a special buffered input servo drive THEN it may work on simple A axis "non interpolated" moves as the drive MIGHT be able to buffer the data stream enought to allow the brake time to release before it errors on position.
BUT THEN you have created a unique realtime mini drive system that the average user does not have. (;-)
Like I said earlier it worked fine in a reatime enviroment as the controller (PCcontroller or motion card) can regulate the closed loop data stream.
BUT hey maybe you have figured out a better mouse trap, GO FOR IT and let us know how it goes.
By the way good job on the 4th axis setup. Have you looked at the system that TORMACH uses with their 4th axis indexer/lathe set?
-
OF course IF you used a special buffered input servo drive THEN it may work on simple A axis "non interpolated" moves as the drive MIGHT be able to buffer the data stream enough to allow the brake time to release before it errors on position.
It had not occurred to me that you were trying to do this on a linear axis! That would be a LOT more crucial since you would be cutting metal during the period. It is impressive that you could get that to work at all. My application is strictly 4th axis and within that, only non simultaneous A axis rotations, with no cutting going on during the period. Following error can be very high allowing, as Bob said, the servo to fall behind and catch up after the Lock released. This would have no practical effect on the machining phases because the period in question (possible fall behind) would strictly be at the beginning of positioning moves between cuts. I have no intention to have spindle locking active during simultaneous multi axis cutting. If fact, I specifically designed a double reduction belt drive for that purpose.
It would be necessary for any servo drive to buffer the input stream to some degree in order to track following error, I would imagine. However, I can see where a fixed small following error that some drives have together with a high res encoder would leave precious little time for releasing a mechanical lock.
BUT THEN you have created a unique realtime mini drive system that the average user does not have. (;-)
I am going to try and avoid this if possible, but I am only looking at drives that have field flashable firmware. If there is no other alternative, I could provide a custom firmware for certain drives, if not the entire drive. I am already be providing the swapaxis hardware, TTL splitter, and solid state relay.
-
By the way good job on the 4th axis setup. Have you looked at the system that TORMACH uses with their 4th axis indexer/lathe set?
Is this the one where they have an entire small lathe that is mounted on the mill table?
-
I don't know why I did not think of that :-[. . . so . . I'll just blame Dan . . . Dan, this is all your fault . . ;)
Huh... ??!! ???
-
actually the 4th axis was a rotary the steps per inch was a slip. The Tormach approach uses a lath BUT it also has a rotary axis hookup that has a quick attachment to convert it to an indexer/interpolated 4th.
Their setup requires manual setup to convert but it only takes about 10secs or so to convert it.
IF you are providing the drive then this is a moot point as you can get the drive setup with a configuralble buffer system that wil wait on the brake. BUT you will have to be able to turn the function off IF you want to do interpolated 3d work.
I think most users are going to want it to do both functions once they start using the 4th(;-) it is like Crack once you get the 4th YA CAN"T LIVE WITHOUT IT. (;-)
-
BUT you will have to be able to turn the function off IF you want to do interpolated 3d work.
I think most users are going to want it to do both functions once they start using the 4th(;-) it is like Crack once you get the 4th YA CAN"T LIVE WITHOUT IT. (;-)
Yes on all counts.
I did some testing and have some quantified results. Settings:
Step per degree: 25.6
Step multiplier: 5x
Velocity: 82k
Acceleration: 14k
The test program did nothing but G0 moves with lock on before and off after - G0 move is next line in code after lock release macro. Macro has sleep(10) after DeActivate. No programmed G-code delays. Heaviest chuck I have (a customer's special wood holding chuck). The machine was literally slamming around in a way that it would never experience in normal use.
Encoder 1800line (7200CPR)
Absolute MAX following error on G0 moves ranging from 30 to 180 degrees each direction:
Without lock: 151 to 189
With lock @50psi 171 to 203
Notes: Above are MAX and occur only once during the run. Most times error is under 10 between moves with or without the lock.
With acceleration set to something reasonable, max error is in the 50's with most moves generating about 5. Negligible difference lock on or off.
Looks like custom firmware is not going to be needed.
-
Is that running with the macro to lock/unlock ???? Using the macro it is a sure thing no problem
TO be seamless you have to eliminate the macro right?
-
Is that running with the macro to lock/unlock ???? Using the macro it is a sure thing no problem
TO be seamless you have to eliminate the macro right?
There will always be error on a servo, so the goal was to see how much error was added by the lock. The test program contains lock and unlock macros, with no g-code commands between the unlock macro and the 'A' axis move. The release Macro and a G0 'A' move back to back is the worst case scenario. The same program was run with and without air to the lock. Without air, the lock is inoperative and cannot add following error due to release latency.
If by 'seamless' you mean transparent to the CNC control software (MACH in this case), then yes, the Macro has to go away.
There are three basic challenges to making the 4th axis work as I want.
1) a way to run the A axis continuously.
2) operation of the lock function
3) ability to run a high res encoder at high RPM.
For use with Mach3, all of these have been solved. (1: swapaxis board 2: embedded macros 3: step mulitplier, smoothstepper, kflop)
For 'universal' use with other CNC setups, only the first is solved and I am working on the other two. Solving the second challenge for 'universal' use (my current focus) will also benefit Mach by eliminating the need for macros embedded in the G-code by manual editing or modified post proc.
-
Looks like custom firmware is not going to be needed.
Cool!
Every now and again we get lucky. Always brings a smile to my face when it happens!
Cheers,
BW
-
NOW the big Question(;-)
If you have that much error using the macro with the delay that the macro creates, what will the error be without the delay from the macro???
"THAT" was where I ran into trouble(;-) I guesss it will depend on how much following error "your" drive allows before it errors out(;-)
-
I don't think using sleep in a VBscript contributes to following error. I don't know that for a fact, but sleeps are inserted to put Mach on hold while functions execute, otherwise it can get ahead of itself. Another example is that you have to put G4 delays in the G-code or Mach won't update the variable monitor.
Inserting sleeps and 'while is moving' and so on are execution holds that have nothing to do specifically with the 4th axis or the lock. Those are in there because Mach needs them. Maybe a scripting Guru will chime here and quantify the time period of sleep(10), but it ain't much . . few milliseconds I would guess.
I hear you on the fault issue, I had some Gecko 320s . . briefly. The drive I am using is a Dugong which has a settable following error up to the 10s of thousands. These drives are tractors. Not too fancy, but just big brutes. I'm using a NEMA34 DC brush motor with a 40A max draw and the Dugong just slaps it around. A momentary error of 200 on a 7200CPR encoder is not much considering the test parameters. Swinging a heavy chuck with stupidly fast acceleration and instant direction changes and long moves . . actually I had the thing pushed to a limit just below where the servo motor would go ballistic and break into violent oscillations. i.e. the test was intentionally not real world. The test settings were quite in excess of anything you would use in normal machining operation.
-
Simpson36,
I don't know if I understood correctly, but maybe if you use some kind of sensor at the brake lever system cappable to stop the program execution using "feedhold"...
Maybe using two microswithes connected in a way that the cnc is in feedhold when the lever is in transit - the command is free only at the end of each move - braking or not.
-
Here is the latest video of the 4th axis. This one is by request also. I've had quite a number of queries about form tools ever since I mentioned it a while back. This vid demonstrates a form tool that I finally got time to make for cutting curvilinear form teeth. Also shown is the hideously painfull way I was making them before the form tool. The contrast is . . . well, just take a peek.
http://www.youtube.com/watch?v=ViN4t4YI_T4
Thanks for the suggestions and comments on the lock operation. It is pretty much sorted out now. For Mach3, a couple of options are available already. It is just a matter of seeing how 'automatic' I can make the operation for the purpose of making the dual indexer/lathe capability easily implemented in a variety of setups, not just Mach. While I am waiting for the modifications to the Dugong, I am going to try other drives that already have the features I need to automate the lock function.
Also I am thinking now about making a trunnion table. That was the reason for the 'Super Duty' version of the 4th axis, so I may as well have one for myself. I do have applications for it. I would love to hear from people who are using or would like to be using a trunnion table. What's good What's bad. What would you like to see?
I was looking at Tormach's 'Duality Lathe' approach. Anybody have comments on that setup? Tormach has the lathe tools in a fixture that appears to be clamped to the spindle. My spindle is servo powered and has a spindle lock like the 4th axis. It occurs to me that a relatively simple fixture clamped to the spindle could hold a bunch of different tools in a carousel arrangement.
-
These pictures show how I set my X2 4th axis up for Tools. I mounted the quick change tool post on the side.
(http://)(http://)
-
That's kind of entertaining.
Although, if you're willling to stick the part in the spindle instead of a 4th axis, the table makes one heck of a gang slide! :o
Cheers,
BW
-
These pictures show how I set my X2 4th axis up for Tools. I mounted the quick change tool post on the side.
Two of your photos are not showing up (at least for me)
Awesome idea! A lathe quick change (or a somewhat less versitile 5C collet holder) is what I had in mind to do until it occurred to me that I might be able to clamp something to the spindle and rotate it to bring multiple tools to bear. I would need to replace the current spindle lock with the new Super Duty lock and perhaps replace the lower ball bearing with a tapered roller.
Another related plan I have is to add a high speed spindle to the head in a horizontal orientation. I have aquired an ER32 spindle for that project. Currently I bolt my die grinder handpiece to the side of the head. It is visible in the last video cutting the pulley teeth with a tiny ball end mill. The X2 is long gone though.
Bob, there is an advantage to keeping the 4th axis free for indexing and to drive a trunnion table (probably my next project after a tail stock) I am currently in the process of reinforcing my mill for heavier work. I have already braced the column and now am moving on to add two more trucks (bearings) to the Y axis ball slides. This will singnificantly increase ridgity as well as add another 4+ inches of Y travel. Nothing like a few hundred pounds of 1/2" and 3/4" 1018 steel plate to stiffen things up.
-
Here is an update: I am going forward with the trunnion table project. The tailstock will be integrated with that project, so I will be doing both simultaneously. I am still planning to make the tailstock from a 5C spin indexer so that long material can pass thru both the 4th axis and tail stock spindles.
Today I purchased the materials and new heavier ball slides beef up the mill and build the trunnion table.
As part of the shuffle, I am selling off some of the other new ball slides which I aquired and will not be using. You can find these in the bargain basement if interested.
I will document as much of the project as I have time for. Photos and text in this forum and videos, of course on Youtube. Nothing is cast in stone at this point, so any 'wish list' or other interesting ideas are welcome. I am of course especially interested in any problems with current trunnion tables or indexers that might be overcome in a new design.
-
Another Update: I've been collecting parts and materials for mill modifications, and as usual, I've gone way overboard to the point where it is no longer really a 'modification' but building a whole new mill.
And of course also a whole new control box to go along with it. My current box is a mess becuase it has been changed and added to so many times.
The result of this is that I am selling two more DUGONG drives (in the bargain basement) and I am also considering selling the existing milll rather than 'modifying' it to such an extreme. Just two photos here. If interested, send me a PM or e-mail and we can discuss.
I have two more projects to finish and then I will be building up the new mill and 4th axis trunnion table. Probably two weeks out.
(http://www.thecubestudio.com/pictures/MillSimpson/FramePerspectiveViewWEB.jpg)
(http://www.thecubestudio.com/pictures/MillSimpson/MillOnBenchWEB.jpg)
-
Steve, team, I'm trying to figure out the options for running my stepper as a spindle and getting confused (it happens a lot). I've read this thread and I know I can do what I want with Mach and macros, but I'm not sure exactly how to do it.
I want to run it continually, like a spindle, pretty much just off and on. I don't need to zero it or ever know where it is. I know I need to configure it and the spindle in Mach3, but beyound that I'm not sure.
I see M951, then I see a bunch of references to Swapaxis(6,3). I can’t find the source for M951 (did it get renamed ?), and I’m wondering if I need the little swapaxis board to do what I want.
It's set up, tuned, enabled and running now as 'A'. I know I need to configue the spindle and enable it.
Past that. I'm not sure.
Thanks !
-
Quick update: Servo drives are sold.
Curmudgeon; In a nutshell, MACH does not run any axis continuously, but it does have an interesting ability to swap the axis around and this can be used to make MACH treat your 4th axis is if it were a spindle . . by 'swapping' the A axis and the Spindle (which for this purpose is treated as an 'axis')
I wrote macros to do this and posted them. I will post them here again. Note that 950 and 951 are not special numbers. You can use any non-reserved number you like.
In order to not be tied exclusively to a MACH specific feature, I developed the 'swapaxis' capability in hardware (the little board).
(http://www.thecubestudio.com/pictures/PCBprojects/SwapAxisCloseWEB.jpg)
A side benefit of this method for MACH is that it works with the smoothstepper board. The swapaxis hardware board simply takes Mach's step and directions outputs from two axis and routs them to two drives. It can then, on command,(any Mach Output) swap the signals to each drive. It is extremely fast and is transparent to Mach (and the smoothstepper).
As I made more macros, things got confusing, so I standardized on Mxx1 to turn something ON and Mxx0 to turn it OFF.
-
So the SwapAxis() and ReSetAxisSwap() are Mach functions then, and functions must be called from a macro, Ahhhh it's all comming together now...
I think I have everything I need.
I can't do any testing till I get home tonight, but I beleive you have me pointed in the right direction, THANKS !
Richard
-
Hi, Great looking machine! The Spindle has an encoder on it. So if you want to Tap. What is the largest Tap you can drive? What is the top and low RPM?
Bob A
www.cad2gcode.com/cncprojects
-
The mill is not completed yet in the photo. I later added way covers and finished the drive belts, added the spidle brake, etc. The spindle has two ranges 0 - 3,600 and 0 - 8,500
The bearings are not fully broken in yet, but I have had the spndle a bit over 7,000 RPM so far running the motor from a 10A 120V variable speed controller. The spindle is smooth enough at most speeds, but there are some 'illegal' speeds where harmonics come in. I am building a dynamc balancer now that hopefully will allow me to smooth it out more.
When I was running the motor with a servo drive, I tapped 3/8 - 16 in aluminum. I don't know the limit yet.
Here are some pics of the drive development:
(http://www.thecubestudio.com/pictures/MillSimpson/PulleyBlanksWEB.jpg)
(http://www.thecubestudio.com/pictures/MillSimpson/PulleysGT2WEB.jpg)
(http://www.thecubestudio.com/pictures/MillSimpson/HeadPulleys01WEB.jpg)
(http://www.thecubestudio.com/pictures/MillSimpson/HeadPulleys02WEB.jpg)
-
So the SwapAxis() and ReSetAxisSwap() are Mach functions then, and functions must be called from a macro, Ahhhh it's all comming together now...
I think I have everything I need.
I can't do any testing till I get home tonight, but I beleive you have me pointed in the right direction, THANKS !
Richard
OK. Some success, some stuff still to be worked out.
Spindle swap worked flawlessly, reset, not so well.
Not sure why, but resetaxisswap() never put things back completely. I didn’t try just doing another swap from 3 to 6.
Also I have both the ‘A’ and the Spindle motor tuning set up exactly the same, but the spindle turns extremely slow unless I use numbers like S40000, which is equating to about 200 RPM’s ??
The DRO never reads any RPM’s for the spindle.
I probably have something configured as angular instead of rotational or otherwise set incorrectly.
Not a lot of time to fuss with it last night, but I’m confident I’m very close.
Thanks for all the help !
PS
Your mill is amazing.......
-
I think you use servo motor for Spindle.
Constantly moving, does not hurt to Servo motor?
-
I think you use servo motor for Spindle.
Constantly moving, does not hurt to Servo motor?
I'm sure your right.
I had this stepper/gearbox from a previous project and just decided to use it for a proto-type. Originaly I was going to have a real indexed axis, thus the stepper. In the end I just decided to use it for a spindle.
I actualy have a real spindle with a 3 jaw chuck and a matching tailstock that's off an old Sherline that would work much better. Speed control and on/off all right there on the front panel of it. But what fun is that ?
-
I think you use servo motor for Spindle.
Constantly moving, does not hurt to Servo motor?
Servo motors do have a duty cycle and are expected to have 'rest' time and therefor have no provision for cooling. Running constantly might cause heat build up in the motor. The motor is rated 6,000 RPM and I have had no problems thus far running continuously at about 4,000 (a guess) with the mill on 'low range' if you will. On high range, at 7,000 RPM, both the spindle and the motor get hot after a while. On my previous X2 mill I ran 7,500 RPM for long periods and had an aluminum finned block on the side of the head to draw off the head. The spindle motor of the X2 has active cooling and it got hot, but not overly so.
In my application I do not plan to run the spindle at high speeds for more than 10 minutes or so at a time, so it is not a priority to investigate that. If I do need to run at the MAX 8,500 design speed for extended periods, I would need to add cooling to both the motor and spindle lower bearing. The lower bearing is currently pressed into a 1" thick block of aluminum which dissipates the bearing heat quite well enough for what I am currently doing.
Active cooling would be relatively easy to add.
-
thx for information
-
AC servos have a continuous rating and an intermittent rating.
AC Servos (or DC Brushless, same thing ;) ) have their windings in the casing of the motor which means they are very efficient at dissipating the heat.
Hood
-
i have two yaskawa AC servo motor. and i want use them Instead Spindle
With This description should not be a problem :)
-
AC servos have a continuous rating and an intermittent rating.
AC Servos (or DC Brushless, same thing ;) ) have their windings in the casing of the motor which means they are very efficient at dissipating the heat.
Good point, Hood. I should have mentioned that the motor that was asked about is a DC brush motor. It has continuous 7.6A and max 38A. This motor is Keling and they provide only minimal information on their motors. Higher end DC brush motors provide more information about how long the motor can be at max amps or how hot is is allowed to get without damage. I have a DC brushless motor to play with but it is far too small for the spindle. Perhaps I should consider and AC servo motor for the spindle.
-
I do know that the stepper gets hotter just holding perfectly still than it does spinning. Sort of makes sense, but not intuitively.
-
Update on 4th AXIS (rotating, but not indexed).
Everything is working as hoped for! Not sure why the resetswapedaxis() was not working originaly, or why the spindle speeds were not displaying correctly, but the next time I tried it everything worked fine. I suspect I had changed things so much it just needed to be restared to get back in sync.
I'll post up a video when I get it working on the machine.
Thanks to everyone for the help.
-
Development has continued during the last several months and I finally have a few picture of the latest iteration of the 4th axis. It is quite a different animal at this point. I do not have time to do a 'build thread' and I will have very limited time to answer questions, but I will try.
The 4th axis is bigger, has a 6" 6 jaw chuck and an entirely new pneumatic lock system. The previous pocket bike mechanical caliper actuated by a linear air cylinder works OK, but had reached it's max potential and I needed more holding power. The solution finally was to design and build a new integrated pneumatic caliper from scratch. Subequently I went with a new rotor, this time from a highway legal vehicle and as large as I could possibly cram into the enlarged frame. I then went on to design an articulated version of the caliper which not locks up the spindle so tight you cannot turn it by hand even with a 12" long lever.
The single and two stage reduction now use the same belt. You just need to move it.
So, the 4th axis development is now completed and I am moving on to work on the 5C tailstock and then the trunnion table. These new items are what most people are interested in. I have included a photo of the first tailstock mock up for proof of concept. It works great. Instead of an aluminum 'nonsense' test piece, what is shown here is an actual finished 4thaxis part. It is an integral one piece shaft extension with the pulley teeth cut right on the same piece of steel. This pulley is for a 750 watt Mitsubishi industrial AC servo motor that is going on one of the 4th axis. I am currently using a 400 watt Mitsubishi AC servo motor on my new prototype. Pictures in next post.
-
5C tailstock mock up:
(http://www.thecubestudio.com/pictures/4thAxisNextGen/TailStockMockUpWEB.jpg)
Little better shot of the part:
(http://www.thecubestudio.com/pictures/4thAxisNextGen/ShaftExtensionIntegralPulleyWEB.jpg)
Brand new integrated pneumatic caliper. Latest articulated verison on the right:
(http://www.thecubestudio.com/pictures/4thAxisNextGen/SuperCasliperPadsCompareWEB.jpg)
Lock and disc:
(http://www.thecubestudio.com/pictures/4thAxisNextGen/SuperCasliperwithDiscWEB.jpg)
New belt drive setup:
(http://www.thecubestudio.com/pictures/4thAxisNextGen/BeltDrivesProtoWEB.jpg)
FInally just some frames and various components:
(http://www.thecubestudio.com/pictures/4thAxisNextGen/4thSDframeComponentsWEB.jpg)
I'll update this thread with the new tailstock and trunnion table progress and development continues on those projects.
-
Wow!
This was a fantastic thread! I have been looking for someone doing what you are doing for a long time now. And finally I found your clips on YouTube. From there I managed to find this thread by googling your nick.
It’s a long time since you last posed, so I’m assuming you have other things to attend to like most others. But like many others I hope you haven’t abandoned this project, and will return with an update and an even better setup!
You are able to do what most people, me included, are not so good at. Get things done and put to life. I must say your design really looks good, and you do know what it takes to get things to do what you want them to.
Anyway, I have an idea that your design could benefit from. I intend to do this when I make an attempt at building my own turning 4th axis as well.
It’s a design used on commercial lathes, so it’s a proven design. It’s also easy enough to get the parts needed.
The idea is to have an automatic gear changer. In your present design you have to loosen the motor and middle shaft assembly to change over the belt to change gear. What you could do is implement the design shown on Kirk Wallace’s page http://www.wallacecompany.com/cnc_lathe/HNC/ (http://www.wallacecompany.com/cnc_lathe/HNC/)about his Hardinge lathe conversion. He describes and show pictures of this design on his lathe. He then goes on to explain that the clutches should be easily obtained from car A/C units. As he mention the size on A/C clutches are sufficient for his sized lathe which is bigger than yours. But at the same time they aren’t so bit that it should be a problem. At least to me as my mill is somewhat bigger and I intend to put a 3-5 KW motor on my turning axis.
I hope you read this and take it in to consideration soon. I think it could be a great improvement on the design.
Then there is the encoder. I would like to get an absolute encoder on there with enough resolution to make it usable as a full 4th axis. And that require more resolution than the encoders you have on there now. The use of absolute encoders was founded in the idea that you don’t need to read them more often than the servo thread is run. And the resolution needed demands so high that turning at full speed would require MHz bandwidth if an incremental encoder was used.
I started a thread about wanting to build your design over at practicalmachinist. But they seemed most focused on telling me why not to do it, to get a millturn unit or farming the work out. That wasn’t my idea, and I tried to get them to understand that I’m doing it for the fun of it. But nobody seemed to think there is a point in doing this on a hobby basis.
That might also be why I haven’t found anyone at all except your attempt at doing this. I don’t se the big problem in doing it, and you have even proven there is none. Except for having to do all kind of hacking on Mach to get it to drive the whole thing that is. That is why I’m so glad I found your design and proof that it is in deed possible to do. And it works GREAT!
BTW, I REALLY like the clips where you show how to use a mill for shaping work. I hadn’t thought about doing things that way. But you have shown me that it’s both doable and works great! Thanks! Now I know it might be possible to do that hex bottom hole I need.
-
Thanks for the compliments. I do not have time for forums these days, but I can provide a few design tips that may prove useful if you intend to build your own machine. First, there will always be 'doom sayers' who will condemn any idea as unworkable or unnecessary . . usually because they have not been able to accomplish the same task. Some people demeaned the idea of this 4th axis, yet I cannot keep up with the demand for them. I am finishing the last three of another batch of 7 and when shipped in a week or so, that will be 14 units shipped so far this year. And there is a waitring list for more. So, do not be discouraged by the untalented who preach impossibility. Even the infamous R8 power drawbar and other 'impossible' projects are very doable.
That being said, every machine designer works to a spec and design criteria. Pretty much all things are doable, particularly in the realm of prototypes, but nobody will purchase a 4 million dollar mouse trap so cost is usually one of the criteria. Because of the pneumatic disk caliper and custom spindle drive pulley, this 4th axis is not an easy DIY project.
Note in the photo that there is no real estate available internally for additional components for an auto shifter mechanism. The spindle lock is a critical need and cannot be eliminated. In fact, the current model is even more crammed as it has a larger air cylinder which required the frame to be expanded slightly. Automatic shifting IS doable, but all features must be evaluated based on cost/benefit and there is also no 'room' in the pricing to add such a feature and remain affordable to the target market. Some challenges with your proposed scheme will be retaining a hollow 5C spindle, very high rotating mass and perhaps there would be consequences to having large electro magnets in close proximity to CNC equipment. I think syncronizers might be a reasonable alternative if automatic shifting is a requirement in your application. Note that on my 4th axis, you do not loosen the motor or the idler. Both are mounted on a moveable plate which is how the belts are changed/adjusted and the same belt is used for both high and low ranges.
You are quite correct in that integrating the unique capabilities of a combined lathe/indexer into Mach3 (or others) is the largest challenge. Lastly a comment on encoders; the motor you see in the above posting is a 400 watt Mitsubishi industrial AC servo with an absolute encoder of over 100,000 counts. That motor has been replaced now by a 75-0watt version and I have acquired three of the current Mitsubishi J3 series for my new mill. That series has 262,000 pulses per rev absolute encoders. However, in practical terms, Mach3 cannot utilize an absolute encoder in a useful way and a typical 2000 line (8000 in quadrature) encoder is completely adequate for most indexing.
Good luck in your project and please do post a link to a build thread if you start one somewhere.
(http://www.thecubestudio.com/pictures/4thAxisNextGen/UltraDutyInternalsWEB.jpg)
-
Videos showing the new tailstock and trunnion table:
Tail stock is finally completed. It has the original planned features as well as some new features. The ball bearing supported hollow 5C spindle is probably the largest difference between this tails stock and any other. The version shown in the video is the first prototype whick has now been sold and the latest version is similar, but the leadscrew redundancy has been removed and a new adjustable drag and spindle lock have been added. Some new videos of that version will be added soon.
http://www.youtube.com/watch?v=DJXg1u4D3Y0 (http://www.youtube.com/watch?v=DJXg1u4D3Y0)
Some testing of the new tail stock:
http://www.youtube.com/watch?v=yM06Ro7y_oU (http://www.youtube.com/watch?v=yM06Ro7y_oU)http://www.youtube.com/watch?v=K4rziDNTu5A (http://www.youtube.com/watch?v=K4rziDNTu5A)
And finally the first tests of the trunnion table:
http://www.youtube.com/watch?v=itJqDwl7X2o (http://www.youtube.com/watch?v=itJqDwl7X2o)
-
More videos:
Use 4th axis for sharpening
http://www.youtube.com/watch?v=Y3PIdnNwsAg (http://www.youtube.com/watch?v=Y3PIdnNwsAg)
Internal shaving of keyways
http://www.youtube.com/watch?v=Nnfcc2fqeTA (http://www.youtube.com/watch?v=Nnfcc2fqeTA)
Major bench mill upgrade phase 1 table
http://www.youtube.com/watch?v=HTQ7u3pcdYs (http://www.youtube.com/watch?v=HTQ7u3pcdYs)
Major bench mill upgrade phase 2 column and head
http://www.youtube.com/watch?v=QsuJgCCPSl0 (http://www.youtube.com/watch?v=QsuJgCCPSl0)
-
Amazing work there on your new mill, Steve!
I have been thinking of using a 4th axis as a tool grinder for some time now. Nice to see you actually doing it and getting good results. Gives good motivation.
You're videos are very professionally made as always and a pleasure to watch. Thanks for showing!
Dan
-
Some pics of the latest tail stock. Redesigned to eliminate positioning redundancy and to add spindle grag/lock mechanism.
(http://www.thecubestudio.com/pictures/TailStock/TailStockFrontViewCommentedRev3WEB.jpg)
(http://www.thecubestudio.com/pictures/TailStock/TailStockRearViewCommentedRev2WEB.jpg)
(http://www.thecubestudio.com/pictures/TailStock/TailStockTrunnionTableCommentedRev2WEB.jpg)
(http://www.thecubestudio.com/pictures/TailStock/TailStockLeadScrewDriveCommentedRev2WEB.jpg)
-
It looks like the tail stock is about 5" wide? This could work on an X2.
To use this on a mini mill you really do not need the lock or break. But you would need to figure out a belt and motor. I have also though of using a X2 spindle box and remove the gears then mount a pulley on the back with a belt and motor. Problem is that I do not have the equipment to mod the parts.
http://littlemachineshop.com/products/product_view.php?ProductID=1906&category=-269978449
Issue with any of the existing Phase II style tables that are retro fitted with a stepper motor is the backlash is difficult to control if you put any real hours on them.
-
The tailstock frame is only 4" wide. In the case of using the tail stock as a 'mini' 4th axis, there would be no need for the rails and therefor, the base could be 4" wide and bolted down from the front and back like the 4th axis base.
I would not consider using a mill spindle because you are going to have R8 or MT which are not good choices for a 4th axis. These are designed to hold cutting tools, not workpieces. 5C is better because you can pass stock thru it and also there are an amazing array of collets available for 5C, including square and hex, as well as 5C mounted chucks.
There is a fundamental difference between a motorized rotary table and my 4th axis in that a rotary table is strictly for indexing. It cannot rotate any where near fats enough to do turning. With a 5C spindle sitting on ballbearings and a tooth belt drive, you have indexing and also turning in the same device.
A paradox inherent in this chameleon is that to get speed useable for turning, you need a servo motor and servo motors, while vastly more powerful and faster than steppers, do not hold as well. that is the reason for the spindle lock.
-
I agree the 5C system is superior. The main purpose I would see for using a 4th axis on a mini mill is for indexing. Most people using a mini mill are cutting light materials such as wax, plastics, brass, etc.
I would forgo the servo for a stepper. Steppers are much more compatible with small home systems. The addition of the belt I assume fixes the back lash issue by eliminating any and increases the positioning speed.
-
Finally, the last piece of the puzzle; a comprehensive motor controller specifically designed to take advantage of the 4ths axis Indexing and Turning capabilities.
Check out the YouTube video - link here: http://www.youtube.com/watch?v=vBLuF_F2_qs (http://www.youtube.com/watch?v=vBLuF_F2_qs)
This new motor control took 200 hours to develop and is a complete system consisting of the control box seen in the photo below, 15 or so 'control' macros, another 20 or so embedded macros ('M' macros to embed in G-Code) a Mach brain program, and a set of custom Mach screens.
The controller has its own PLC and generates its own step and direction signals independent of Mach, but it is also tied into Mach to retrieve and deliver data for two separate speed modes and also retains all normal indexing 'A' axis functionality.
My favorite feature (other than being able to control the 4th axis with screen clicks instead of manually entering macros into the MDI) is what I call 'AutoSpeed'. This mode monitors the distance from center of the cutting tool on either the Z or Y axis and automatically calculates the 4th axis spindle RPM to maintain a specified SFM. A significant portion of the 200 hours of development time went into creating this mode.
Over 2,000 lines of C code run on the internal PLC and the controller is connected to Mach thru ModBus. The PLC can also communicate using Modbus Ethernet, but that version of ModBus lacks some features that I need for certain operations, so I had to stay with the slower interface.
(http://www.thecubestudio.com/4thAxisMotorContol/ControlBoxCloseBlueWEB.jpg)
-
It is a Honeywell model HOA 1887-012
I now have these on all of my machine axis and also on the 4th axis and they will be used on the commercial verison of the 4th axis as well. You simply need a 100 to 150 ohm resistor on the 5V that feeds the LED (red wire) You can use 5V or up to 30V for the trigger side.
@ simpson36 (or others)
Can you be a little more descriptive of your wiring scheme (http://"http://www.machsupport.com/forum/index.php/topic,11422.msg88150.html#msg88150")? There are more wires attached than what you describe. And with me being eletronics challenged and I frequently blow things up when I don't know what I'm doing.
-
I *think* you are talking about the sensor?
If so, then I think a description of how it works is better than a simple 'hook wire A to terminal B' type of answer.
These type of sensors are quite simple. They work by shining a light at a photosensitive 'receiver', which is just a switch that turns on when you shine a light on it. Note that the light is in a range not visible to humans.
The light source is on one side and the 'switch' on the other side. Between those goes your shutter wheel, a tab on your mill table. etc. This breaks the bean and turns off the switch. When you remove the obstruction, be it a tab attached to your table or a hole or slot in a shutter wheel, then the light reaches the switch and it turns on. That's all there is to it.
So you have a power and ground on the light source side (sometimes, but not always red and black). You have to obey the rules on this as you would with any other LED. You MUST restrict the current to a level that is tollerated by the LED or you will burn it out. In the case of the referenced sensor, if you feed it with 5V, then you need between 100 and 150 ohms resistor in the wire to 'slow down' the juice to a level that keeps the LED healthy. This resistor can be anywhere in either the positive or negative wire and the value is not super critical.
The receiver side has sometimes (but not always) a green and a white wire. Since the device burried in the sensor and attached to these wires is just a switch, you don't need to be so careful. If you are only feedijg a pin on a BOB, just do not exceed the sensor's volatge spec and you are pretty much golden. The switch does have polarity, meaning that the power will only flow in one direction, so if you get it hooked up backwards, it won't work. Good news is it probably won't be damaged so just reverse the wires.
TIP: with the light side powered up and no obstruction, the switch should be ON. You can check this with an ohmmeter, continuty tester, etc. Put the probes on each way and you will get continuity in one direction and not the other so you will be able to see which way to hook it up. A catalog cut (spec sheet, data sheet, etc) also will define which wires are for what purpose.
Hope this helps.
-
HARD TAPPING!! Finally, I got some time together and added the last few parts to convert the mill spindle to full servo power. this opens up a lot of possibilities, but for now, I am basking in the hard tapping capability. I despise tapping. I have TapMatic heads, and Custom made floating tap holders and a reversing manual speed control on the mill spindle . . which makes thing go faster and with a lot less effort, but it still requires an operator full time to . . operate.
No more . . . This video is a 'by request' to overview the concept of parametrics, but the new servo driven spindle is the perfect vehicle to demonstrate some basic functions:
http://www.youtube.com/watch?v=Cu61oBY5-rw (http://www.youtube.com/watch?v=Cu61oBY5-rw)
-
The 4th axis is getting a new custom designed and built spindle. The motivation for launching into this prokect was three fold;
1) need for strong, safe, standard chuck mounting for large chucks.
2) larger size to allow the frame to be lengthened and eliminate the cantilevered main drive pulley mount.
3) consistency and precision.
When I have more time, I will try to do a blog on the build but for now here are just a few photos of the finished spindle. I am currently building a new prototype 4th axis to house this new spindle. The spindle nose is 'universal in that it can mount standard D1-4 and also A or B style chucks.
Photo 1 compares the new spindle to the previous. The nose and front bearing surfaces are ground, hoever, both the Chuck mount and 5C tapers as well as the face of the flange will all be ground in-place after the machine is assembled and under its own power. This will provide for the runout that equale the bearing spec. The new spindle also can accomodate ABEC7 angular contact pairs for a high precision option.
(http://www.thecubestudio.com/4thAxisSpindleProject/SpindleCompareHorizontalWEB.jpg)
Photo 2 shows the 'universal' mount. Note the cam loc parts and also the threaded inserts. More on this later.
(http://www.thecubestudio.com/4thAxisSpindleProject/SpindleFlangeWithNutsFrontW.jpg)
Photo 3 shows the new spindle on the balancer . . which needed some modifications to hold the new spindle.
(http://www.thecubestudio.com/4thAxisSpindleProject/Balancer01WEB.jpg)
-
Drive parts are completed. Some more pics:
First photo shows the monster main drive pulley. Big pile of aluminum after making this guy.
(http://www.thecubestudio.com/4thAxisSpindleProject/SpindleComponentsAll01WEB.jpg)
Motor extension is needed to hold dual pulleys. In this case, the small pulley and the shaft are made from a single piece of steel. I am using a clamp collar in this design and there are still set screws just for redundancy. While very robust, I think it will be difficult to balance this arrangement, so it may not make it past prototype. The ratty looking larger aluminum pulley is carried over from the current prototype.
(http://www.thecubestudio.com/4thAxisSpindleProject/MotorPulleyCompWEB.jpg)
The new counter shaft is heavier and is also made from a single pice of steel for the shaft and small pulley. The bearing case is a stock item from the previous design, just put in here to show the components. A new casing with built in belt tracking adjustment will be used on the new 4th axis.
(http://www.thecubestudio.com/4thAxisSpindleProject/CounterShaftComponentsAllWE.jpg)
-
Thank you very much for taking the time to share. You have built quite a 4th axis! How does your balancer work? I will have a few projects coming up soon that I will need to balance electric motor shafts.
-
The balancer duplicates the functions of a commercial balancer. The end support are free to swing on ball bearing fulcrums and an accelerometer measures the movements imparted by imbalance. On one end is an absolute encoder which provides azimuth data so that the exact position of the imbalance is known.
A little PLC collects the data in real time and then sends the data to the computer via USB where it is analyzed by a separate Windows program. The data is checked for validity and then reduced to a specific relative imbalance 'severity' number and the azimuth of the imabalance. Dynamic balancers are potentially dangerous machines and you may notice the nylon wire ties which are just for safety in case the part being balanced decides to try to become 'free to move about the cabin'. For liability reasons, I do not sell balancers nor share any technical information about them. I was quoted between US$150 and 200 per spindle to do the balance, and the cheapest used (and working) industrial balancer I could find was about US$35,000 so it was worth my while to build one for my own use.
-
Couple more photos:
The meet spec for precision bearings, a new method for machining the 4th axis frames is shown here.
The frame uprights are bolted to the fixture (here a 6" chuck and later will be a 12" face plate), indicated and the top and bottom faces trimmed parallel to the centerline. The frame is then assembled and torqued in-place on the fixture. A heavy steel (.875" thick) top plate is installed to stabilize the frame uprights and provide a good balance for the spinning frame. The top plate is later removed and replaced with the permanent .160" 7075 aluminum top cover.
With the assembled frame in the fixture, the base is faced to be perfectly parallel to the centerline.
(http://www.thecubestudio.com/4thAxisSpindleProject/LineBoreFaceBottomWEB.jpg)
The pre-roughed bearing pockets are then line bored first the rear bearing:
(http://www.thecubestudio.com/4thAxisSpindleProject/LineBoreBackWEB.jpg)
And lastly the front bearing pocket is brought to size. The image is captured from a video so the quality is not great. When I have time, I will do a blog and/or video on the entire new machine including the new spindle and the new frame.
(http://www.thecubestudio.com/4thAxisSpindleProject/LineBoreFrontWEB.jpg)
-
Project is just about completed. This new machine will be operational in a day or two and the current 4th axis prototype is now for sale if anyone is interested. It is the machine making the parts for this new prototype and in the latest videos. PM me for details.
A couple of interesting componenets are the belt tracking adjuster built into the counter shaft housing. The housing is mounted via a ball socket and can be angled a couple of degrees by alternate tightening of the flange bolts to effect belt tracking.
(http://www.thecubestudio.com/4thAxisSpindleProject/BeltTrackerCloseUpWEB.jpg)
Looks like this assembled:
(http://www.thecubestudio.com/4thAxisSpindleProject/CounterHousingInMotorPlateW.jpg)
Balanced main drive pulley and hub. Note the drilled holes in the hub for balancing. In this close up, you can see the super straight teeth afforded by the spindle lock.
(http://www.thecubestudio.com/4thAxisSpindleProject/MainPulleyAndHubBalancedWEB.jpg)
Some 'undressed' shots of the finished new 4th axis are next. The new 'Mega' 4th axis has pretty covers with the name imprinted on them, but I do not have photos of that yet.
(http://www.thecubestudio.com/4thAxisSpindleProject/4thMegaBackNoCoversWEB.jpg)
(http://www.thecubestudio.com/4thAxisSpindleProject/4thMegaRearNoCoversWEB.jpg)
(http://www.thecubestudio.com/4thAxisSpindleProject/4thMegaFrontNoCoversWEB.jpg)
(http://www.thecubestudio.com/4thAxisSpindleProject/4thMegaMotorSideNoCoversWEB.jpg)
-
Nice! Are you going to reuse the servo motor?
-
Yes, the new 'Mega 4th Axis' has the Servo motor from the previous prototype. It is now installed and has begun its working life. The last steps were to to drill the base, drill a couple of additional holes in the outrigger plate (the 'Mega' is longer), align the machine, do the final cuts on the tapers with the machine installed and under its own power, mount the D1-4 Chuck adapter and then cut and final grind to mount the 6" Bison chuck. I also now have a balanced 12" cast iron D1-4 mount face plate. You can see that piece and also build details and install details in an upcoming video.
Here are a couple of selected final shots to finish the update to the thread. First is the in-place cut of the tapers:
(http://www.thecubestudio.com/4thAxisSpindleProject/SpindleTapersCutInPlaceWEB.jpg)
Next is a shot of the ground cam-loc face of the completed D1-4 chuck adapter:
(http://www.thecubestudio.com/4thAxisSpindleProject/D1-4AdaptGroundCamLocSideWE.jpg)
THe completed adapter is then installed on the re-cut spindle and then rough cut and ground to receive the chuck:
(http://www.thecubestudio.com/4thAxisSpindleProject/D1-4AdaptGroundOn4thAxisWEB.jpg)
And lastly a glamor shot of the finished product . . . which I will have to do over because it had not occurred to me to wait until the spindle was final cut:
(http://www.thecubestudio.com/4thAxisSpindleProject/4thMegaFrontWithCoversWEB.jpg)
-
The video:
http://www.youtube.com/watch?v=TlQgMk6te_Y (http://www.youtube.com/watch?v=TlQgMk6te_Y)
-
Every time I go through this thread, I'm awed. I'm about to start making chips for my own 4th axis. So I visit this thread for inspiration!
Mike
-
Hi Steve,
This was an awesome thread from start to finish! Your work and perseverance is amazing!
Anyways just wanted to know if the latest version you need to swap belts? I noticed two difference size pulleys on the motor.
I am also wondering what would be needed to complete the setup, motor, driver, relay etc with the production unit?
-
Just a tid bit here and a few answers.
The machine has two very different speed ranges. The belt swap is to move from 'back gear' which is now 12:1 ratio for 'direct drive' which can be a variety of ratios from about 2:1 to 4:1 for high speed turning. The spindle and all rotating parts are balanced to 4K so for example with a 6k motor and 2:1 'direct drive' (one belt from motor to spindle) you can run 3K spindle with collets or a small chuck. In 'back gear', the same motor would run the spindle at only 500RPM for use with a large (6" or 8") chuck or large (12") faceplate of the trunnion table to do low speed high torque operations.
I am now working on a BT30 high speed spindle and Automatic tool changer. The first prototype is good for maybe 7k to 10k. To go faster, in my opinion, a stronger spindle is called for and to that end I have made a prototype spindle completely from A6 tool steel, hardened, properly tempered and ground. I made it for the InTurn™ 4th axis not because I intend to run the 4th axis at 15k, but because the 4th axis is capable of providing a brutal torture test in strength of the spindle.
The new spindle is ready to install in my own InTurn™. the final grind on the 5C internal and D1 external tapers are done with the spindle installed. The spindle from my machine will get a new improved D1-4Flange and be put in one of the two extra frames that have from making the last batch of machines, so I will have one additional InTurn™ available if anyone wants it.
Here is the new spindle:
(http://www.thecubestudio.com/BT30SpindleProject/Spindle_InTurn_GroundWEB.jpg)
-
all I can say is wow I have read the thread backward and forward. everything is unbelievably well documented except the balancer it looks home built as well. I am very interested in the harmonic balancer A spindle and some deep groove bearings have followed me hone. I have a large DC servo and made some sketches. I had not thought of the challenges of the software side of the equation. I have 2 horizontal manual mills so this will be ran stand alone as well as hooked to the bridgeport mill as a indexer and sub spindle. I am planning to use the Audrino to control it as a stand alone indexer. http://www.homemodelenginemachinist.com/f39/electronic-dividing-head-using-arduino-17896/
the spindle is out of an emco myer lathe, it is a D1-4 and is the same spindle that is in My south bend 14 lathe so I already have chucks and faceplate ect. that fit. ( it may give me an excuse to buy that 16C collett chuck I have been looking at) as it will fit the lathe as well.
your efforts to impart this to the rest of us is above and beyond
thank You
archie =) =) =)
-
all I can say is wow I have read the thread backward and forward. everything is unbelievably well documented except the balancer it looks home built as well. I am very interested in the harmonic balancer A spindle and some deep groove bearings have followed me hone. I have a large DC servo and made some sketches. I had not thought of the challenges of the software side of the equation.
That spindle looks perfect for a 4th axis project!
I am unclear if you have asked a specific question and also unclear if the software you are referring to is for the 4th axis or the balancer, so I will address both. First the controller:
The InTurn™ 4th axis motor controller is a fairly complex device in its current form. There are approx 3,500 lines of C code running on an Atmel processor. The controller has been thru one very major upgrade wherein the signal generation was moved off the processor to a separate digital signal synthesizer with MHz capability. I am holding the signal speed to 500KHz for practical reasons. The controller is about to go thru another fairly major update. This next update will not require any new hardware.
1) 'Manual' mode is gone. The existing speed dial and switches have been re-purposed and now provide real time variable speed for both set speed and autospeed modes.
2) The modbus comm to MACH3 has been changed from Serial Modbus to the newer serial plug-in modbus.
3) The use of brains if much more extensive and replaces the use of the 'legacy' serial modbus register commands
4) Most of the code has been streamlined and optimised now and overall response time is improved for most functions, especially for E-stop.
There is no info posted on the new upgrade, but you can read the specs and see a video of the development and operation at www.theInTurn.com
The servo drive that you chose should have enable/disable capability. None of the hobby level drives have this so far as I know. The drive I recommend for DC is the Copley Accelnet. If you are accustomed to hobby level drive prices (Gecko, Dugong and the newer versions, Viper, Leadshine, etc), the cost of the Copley might seem quite high, but it is a commercial/industrial level drive and it priced competitively and actually far below the AC drives from Mitsubishi, Yaskawa, AB and so on.
-
all I can say is wow I have read the thread backward and forward. everything is unbelievably well documented except the balancer it looks home built as well. I am very interested in the harmonic balancer
I think you are referring to the dynamic balancer. Yes, it is a 'home brew' machine. I wanted to purchase a commercial balancer, and I still may at some point, but the machines are just way, way too expensive to justify at this point. I am able to do an acceptable balance with my balancer, it just takes longer . . a lot longer actually. However the end result is similar. Having a spindle balanced at a commersial shop runs between $200 to $250 for the level needed by a 4th axis.
For liability reasons, I do not release any information about the balancer, and I do not sell them. It is going thru a rebuild and upgrade right now to increase its balance speed from 4k RPM (InTurn™ 4th axis ) to 8K RPM (new BT30 spindle).
What I can do is describe the basic operation using a broad brush and you would need to take it from there.
1) A 'floating' carriage holds the item to be balanced
2) when the item is spun, the carriage oscillates as a result of the imbalance
3) a sensor, either pressure or accelerometer, records the movements while an absolute encode simultaneously records the azimuth.
4) Software calculates the amount and position of the imbalance.
5) the item rotation is stopped and the item is turned by hand to the exact imbalance point as shown by the controller.
6) weight is added to the light side or taken off the heavy side and the process is repeated.
An Atmel processor (on an Arduino MEGA development board) is used for data acquisition only and the data (sensor reads) that is collected on each run is sent over a serial connection to a program on the PC for the calculations.
My upgrade of the balancer involved beefing up the floating carriage and swapping out the current 16bit 16MHz processor (Arduino MEGA) for the new 32 bit 84MHz Atmel processor (Arduino DUE) which is 4 to 5 times faster and moving from analog to digital accelerometer. Hopefully these improvements will allow fast enough data collection for an 8K RPM balance.
I hope this is of some help to you
-
An Atmel processor (on an Arduino MEGA development board) is used for data acquisition only and the data (sensor reads) that is collected on each run is sent over a serial connection to a program on the PC for the calculations.
Is it a commercial program or some code of your own?
Dan
-
Yes that was the two questions and you have answered them. I will be making a balancer at some point.
I would to do this from inside mach3 I want a true 4th axis ( as a
indexer and a spindle with brake) to run on my bridgeport R2E4. I have
no requirement to have separate electronic box to control the 4th axis. while mounted on the cnc mill.
understanding I want this for my own use.
thank you for the help
archie =) =) =)
-
An Atmel processor (on an Arduino MEGA development board) is used for data acquisition only and the data (sensor reads) that is collected on each run is sent over a serial connection to a program on the PC for the calculations.
Is it a commercial program or some code of your own?
Dan
My own.
-
Yes that was the two questions and you have answered them. I will be making a balancer at some point.
I would to do this from inside mach3 I want a true 4th axis ( as a
indexer and a spindle with brake) to run on my bridgeport R2E4. I have
no requirement to have separate electronic box to control the 4th axis. while mounted on the cnc mill.
understanding I want this for my own use.
thank you for the help
archie =) =) =)
The control box turns over control to MACH for indexing, so when in INDEX mode, it operates as you describe, i.e. within MACH3. MACH has no ability to continuously rotate an axis so you will need to do some 'smoke and mirrors' tricks to get that to happen. There are a couple of ways to go about it and I think those are detailed early on in this thread. I was doing the INdexing TURNing using MACH only for quite some time before I made the control box. It is doable, just a bit inconvenient and limited. It is a good place to start. If the limitations become a problem, you can then create a separate controller for the TURNING operations.
Lack of enable/disable is a serious handicap. Cost condiserations may force the use of a hobby drive to get started, but the more you use the machine, the higher the priority will be to upgrade the drive.
-
Is it just that the changeover is slow. or is it a programming issue IE having to massage the code by hand.
Thank You
archie =) =) =)
-
I do have a untested yaskawa servo drive under the bench that I may break out if nessessary it is supposed to drive a 5HP spindle I was sort of holding that back for a lathe project but who knows.
archie =) =) =)
-
Is it just that the changeover is slow. or is it a programming issue IE having to massage the code by hand.
Thank You
archie =) =) =)
I'm not sure what you are asking here. There is a feature in MACH called 'swapaxis' that can be used alternately connect the spindle axis and the A axis to your 4th axis. I used that initially and then later created a small circuit board to accomplish the same feature in hardware. Either method is essentially instantaneous. There are some caveats to using the internal MACH command which I think are detailed elsewhere in this thread.
You have to message the code in any case, so that is not a factor in deciding how to control your 4th axis motor. On that topic, a couple of my buyers are working on post processors that embed the codes for the InTurn™ and there are three editors that make it quick and easy to embed the codes.
*The most sophisticated is an editor by CNCcookbook. Bob Warfield added a feature for the InTurn™ (or equiv) code 'massaging' that that actually looks at the context of the G-code and can differentiate where a command should be embedded and where it is should not. A 'smart' editor, if you will.
*There is free editor called NotePad ++ that has the ability to replace a line with several lines. This is the 'secret' to fast and easy editing to add InTurn™ commands.
*A third CNC code editor has promised to add a similar feature, but so far I have not heard back from them and I have not gone searching to see if it is done yet.
-
I do have a untested yaskawa servo drive under the bench that I may break out if nessessary it is supposed to drive a 5HP spindle I was sort of holding that back for a lathe project but who knows.
archie =) =) =)
??? Your 4th axis project IS a lathe project. Once you have your 4th axis doing turning on your mill, as Scarface would say "I don need no steenkeen LATHE!"
Yaskawa AC drive would be an excellent choice. If you have 5HP (continuous), you might get away without dual range belts, depending on the type of work you are doing.
-
Simpson, I just wanted to say something here........... THANK YOU! This is a very, very, very good topic. Inspirational to say the least. Great work, creativity, originality, implementation, can all be found right here. Thanks for taking the time to share. :)
Brett
-
I couldn't resist.
my Stinkin lathe
-
Do you have the part number for your sealed AC 45mm ID Bearings? I read through the thread but didnt see it. All I can find is non-sealed versions.
-
Do you have the part number for your sealed AC 45mm ID Bearings? I read through the thread but didnt see it. All I can find is non-sealed versions.
If your question is directed at me (the OP), I don't understand what you are asking exactly, and you did not specify the application you have for the bearings.
You are probably not going to find sealed AC bearings, nor sealed tapered rollers. There are DUAL row AC bearings that are sealed. These are typically 5000 series although the numbering system has changed a couple of times.
-
I was directing it at you. I am going to build a somewhat similar setup for myself using a spin index spindle. In several of your pictures you show what appears to be a sealed angular contact bearings. Are these not angular contact then? Or maybe you are using two dual rows? If not, what are they, deep groove?
-
After doing a little bit more reading on another thread you participated in I see you are using 6000 series SKF Deep Grooves (on the first few prototypes anyway, the new ABEC 7 one sounds sweet btw!). I ordered up some 6209's just now. Was gonna go with the 6309's, but I wanted the higher RPM rating for some of the turning i do. One other question I had for you, in some of your pics it shows you line boring the bearing bores in the aluminum housing. It was my understanding that in order to preload these bearings, the outer races would have to be seated against something (horizontally). In other words, I would have assumed that you would have had to counter bore the bearing holes and leave a bit of a backing for the outer race to sit on rather than boring the hole straight through. Is the friction of the bearing fit enough to hold the outer races for the preload?
If my question doesn't make sense let me know and ill whip up a CAD model to show what im trying to explain.
-
Al of the 4th axis that I have made use sealed beep groove ball bearings. Ball bearings can be side loaded and that is part of the spec. Early on I spoke to the SKF application engineers and they have no problem with pre=loading the deep grove balls provided the side loading spec is not exceeded. Because the spindle is hollow, it is very large in diameter and therefor takes a very large bearing. In fact, far larger than would be needed even in seriously abusing these machines. These big bearings have such high side loading numbers, that angular contact bearings are not needed in this application.
You can pre load a dual row AC bearing also, but doing so would result in only utilizing one of the rows. The side loading spec would be unaffected, but I would half the radial load spec. in that arrangement.
Good catch on the line boring!. The rear bearing is retained by three large socket caps with hardened washers. I later replaced this with an internal snap ring in the bearing bore. I have acquired a surface grinder and that changes the process considerably. It would take too long to explain here, but I am currently doing a video series on the new InTurn™ ULTRA model and, by request, focusing more on the design aspects than the machining of parts. There is a lot of information on bearings and the seals associated with AC or tapered rollers. The top-of-the-line ULTRA-T model features Tapered rollers which require 4 separate shaft seals.
The video series begins with a slide show and progresses to live narration video (again by popular demand) here: https://www.youtube.com/watch?v=TdU2VD3UdVw (https://www.youtube.com/watch?v=TdU2VD3UdVw)
-
Wow! From what you started with in the first page of this thread to what you are building now is SERIOUSLY impressive. My hat is off to you. I really appreciate your help with my questions.
I'm going to be building something similar to your first aluminum framed prototype. Ill be using a spin index spindle, SKF 6209 bearings, and probably this Servo (http://tinyurl.com/nh6j2bo). I'm planning on gearing it to 3:1 only, for simplicity. I'm thinking that should give me enough torque and hopefully get me around the 1400 RPM max speed range. I will be using it to machine mainly 6061/7075 and 1018/1020, nothing too crazy. Most parts will be a couple inches in diameter or less. Anything that is bigger will only be indexing.
- I'm unsure on the calculations needed to get the rotational accuracy I should expect.
- For small diameter (3 inches or less) work pieces, and 2550 oz in of torque (850 oz in peak * 3), will I absolutely need a brake?
- I really love the Gecko products, and all of my builds have Gecko drives. I know you said the G340 didn't meet your requirements, but based on what I'm building, do you think the new G320x will accomplish what I'm after (bearing in mind that my 4th axis will not be anywhere near the Mega and Ultra stuff you are building now, but more so inline with your first aluminum house prototype)
-
. . . I really appreciate your help with my questions.
Your questions have become component specific and I am not inclined to spec a machine for you over a forum, however, I will answer conceptual questions that are useful to the majority or people reading the thread. For example, unless you need to weld, 1018 is a poor choice of material for a low powered machine. You will fare much better with a 'free machining' steel like 12L14. That is my advice. You can answer 'why is it better' yourself by doing some research on materials.
Gecko makes fine products, but for a 4th axis that spins, you will want a drive that has enable/disable capabilities and as far as I know, Gecko does not. For DC servos, I recommend Copley AccelNet drives. These can be purchased used for about what a new Gecko would cost.
http://www.ebay.com/itm/Copley-Controls-Accelnet-ACP-090-36-Servo-Drive-CNC-High-Current-with-Warranty-/261684304231?pt=LH_DefaultDomain_0&hash=item3ced999967 (http://www.ebay.com/itm/Copley-Controls-Accelnet-ACP-090-36-Servo-Drive-CNC-High-Current-with-Warranty-/261684304231?pt=LH_DefaultDomain_0&hash=item3ced999967)
Here again, this is my advice, but do not ask me to spec the exact drive you need. Visit the Copley site. Learn about the drives. Do your homework.
Instead of endorsing a specific motor, I will tell you that Torque is a static measurement that does not define available power. 'Power' (Horse Power for example) is work done over time. Without the time component, torque is a useless number unless you are tightening bolts. The formula for calculating horsepower is easy to find.
CNC cookbook has a speed and feed calculator that will tell you how much power you need to make pretty much any cut in any material.
Yes, you will need a brake. There is no way your servo motor will be able to hold the spindle still for machining operations with a 3:1 reduction. However, you can use a simple band brake like I use in the Tail Stock II. A design goal for the Tail Stock is to make the machine as compact as possible so a big rotor is not an option. The band brake is intended as a 'damper' an not a powerful precision spindle lock, but it will do a good job for light duty general-precision work.
-
Simpson,
Thank You for your time and energy wisdom you have shared with us. all of this info is special. and much appreciated by myself for one and I am sure many others.
Happy Hunting
archie
P.S. I have a pile of parts including a D-4 camlock spindle waiting in the wings.
-
^ Agreed, thanks again for all the help!
-
I am in the middle of refurbishing an old lathe headstock to use on my bf20l/grizzly to be my 4th axis/ So far the mechanical side of things is pretty easy going with a few replacement bearings for the headstock and probably a 6:1 or 8:1 timing pully setup.
I've been wondering how you might think a nema34 stepper would go instead of a servo drive? Simply the price is the attraction not to mention I can get one with about 11NM of torque. I cant seem to find any servos anywhere near the price range of around 200 USD that the stepper would be(including the driver and power supply).
I actually thought about using a MTB disk brake with a small ball screw to adjust the drag/brake.
Anyway these are things I am mulling over now perhaps let me know if you think there are obvious problems I may be over looking.
Ive given the link to the specific parts im looking at on ebay at the moment anda pic of the headstock to make this more interesting
(http://i873.photobucket.com/albums/ab294/navanod/old%20emco%20lathe%20headstock%20and%20spindle/IMG_0668_zpsfa6e8b77.jpg) (http://s873.photobucket.com/user/navanod/media/old%20emco%20lathe%20headstock%20and%20spindle/IMG_0668_zpsfa6e8b77.jpg.html)
http://www.ebay.co.uk/itm/201217825668?_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT
http://www.ebay.co.uk/itm/261449060712?_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT
-
6:1 or 8:1 with a stepper will have good holding power, but it may not have the resolution you are after.
With a 1:1 ratio, each step of the motor is 1.8 degrees. With a 6:1 ratio, each step will be 0.3 degreess. With an 8:1 ratio, each step will be 0.225 degrees. With half steps, those degree numbers half. With microstepping further the same occurs.
If you assume full steps, and a 3" diameter work piece, each step would be:
- 0.047 inches of circumference travel per step if 1:1
- 0.007 inches of circumference travel per step if 7:1
- 0.005 inches of circumference travel per step if 8:1
Again, those number shrink if you half/microstep, but you will also lose torque. I decided to go with a harmonic drive coupled with a pulley ratio to get 66:1 which should give me awesome resolution regardless of workpiece size. For reference, i believe HAAS' 4th axis is 88:1. Also keep in mind that with a stepper you probably wont be able to spin it anywhere near fast enough to do any spindle work like Simpson is doing with his Servo setup. Lastly, 8:1 may be difficult to accomplish within a specific given space. You will probably need to have a multiple pulley setup which can potentially introduce more backlash.
A servo will allow you to get much higher resolution with much lower of a pulley ratio, and will also allow you to spin it up fairly fast. The downside is servo's dont have much holding power, which is why Simpson introduced a brake when doing 4th axis work vs lathe work.
-
Thanks mike,
I have also got a new harmonic drive for this project but was thinking it probably wont be any good for any fast spindle work. I will probably use the HD mounted to the back of a vertex 6 inch rotary that I have to do some pretty precise work.
I would have thought that a stepper can actually spin pretty fast if programmed correctly though perhaps im wrong there.
Any chance you know what servos Simpson is using anyway so I can spec things out a little more?
-
I do have a few older electrocraft e240 motors with some heds encoders but may be a bit less torque with these
http://www.ftp.cnchungary.com/Varsanyi_Peter/CNC%20vezerles%20-%20szervoval/ElectroCraft_E240_Motor/electrocraft%20E240.pdf
-
You can use any servo motor, really. The difference between Stepper and servo is what makes the difference. Which servo is important, no doubt, but not nearly so much as the fundamental technology of the motor type.
Steppers have great holding power and the torque rating is at zero speed and it drops off quickly as the speed increases. Whether a steppers holding torque is adequate for the task depends on the task. No stepper will hold as well as a spindle lock. If you are not going to use flood cooling, you can rig up an inexpensive spindle lock with a 4" band brake and hub for a mini bike or go-cart.
In considering servo motors, one should look at how they are rated. Hobby motors (Keling, etc) are rated at 'just prior to self destruct' levels whereas industrial servos are rated extremely conservatively. Usually at the power level they can sustain 24/7 at for 20 years without a break. Typically an industrial servo can pump out between 2.5 and 3x the torque rating in an 'intermittent' basis.
So, the numbers tell us that a 400watt rated industrial AC servo is equivalent to a 1,200 watt rated (calculated from the power draw) DC brush hobby servo.
A decent used 400 watt industrial AC servo motor and drive set can be had as cheaply as the equivalent new hobby DC servo setup (if you include a decent encoder for the hobby setup)
http://www.ebay.com/itm/MITSUBISHI-MR-J2S-40B-HC-KFS43B-AC-Servo-Drive-Motor-400W-Brake-Set-a18-/160880697451?pt=LH_DefaultDomain_0&hash=item25753ca06b (http://www.ebay.com/itm/MITSUBISHI-MR-J2S-40B-HC-KFS43B-AC-Servo-Drive-Motor-400W-Brake-Set-a18-/160880697451?pt=LH_DefaultDomain_0&hash=item25753ca06b)
http://www.ebay.com/itm/CNC-Machine-YASKAWA-AC-220v-Servo-Motor-Driver-400w-Replace-Step-Motor-Tested/321620537287?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D1%26asc%3D27675%26meid%3D871f20ebd6ed4b699b2af883250a6825%26pid%3D100005%26prg%3D11378%26rk%3D4%26rkt%3D6%26sd%3D160880697451&rt=nc (http://www.ebay.com/itm/CNC-Machine-YASKAWA-AC-220v-Servo-Motor-Driver-400w-Replace-Step-Motor-Tested/321620537287?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D1%26asc%3D27675%26meid%3D871f20ebd6ed4b699b2af883250a6825%26pid%3D100005%26prg%3D11378%26rk%3D4%26rkt%3D6%26sd%3D160880697451&rt=nc)
I would buy these in a heartbeat over a new hobby setup and never look back.
-
thanks sound advice and solves my hum har about the steppers.
Ive been thinking a lot about your awesome brake and how and if you can actually calculate "drag" for continuous milling?
-
I would have thought that a stepper can actually spin pretty fast if programmed correctly though perhaps im wrong there.
*Please anyone feel free to correct me if im wrong here*
I believe, that if you are running Mach at 35,000 mhz kernel speed, that implies that it can make 35,000 pulses per second, meaning 35,000 steps per second, which at a full step would be 175 rotations per second which is 10,500 RPM. Now, the caveat is, that is a 1:1 ratio which obviously wont work, and 10,000 is probably beyond the rated RPM for a stepper. If you throw an 8:1 ration on there, you're looking at 1312 RPM (at the 4th axis) with the stepper running 10,500 RPM. Also, as Simpson mentioned, the faster a stepper turns, the more torque/holding power it loses. So basically, you'd be at the point where it would be pointless not to do a Servo if you are planning to do high speed operations.
That all being said, I still went with a stepper because of cost, ease of setup, and I didnt need to do lathe work on my mill. Its all about what your intended use is for the machine.
-
My 4th axis I guess at the moment is planned for a wide variety of work. The initial stuff im doing is small diameter 1/2 inch with lathe operations and some 4th axis milling with slitting also. Gears and timing pulleys I want to do as well as some special collet making for carbon fiber robot arms say up to 20mm. Ill add another independent linear rail latter for small hobby servo internal splines as a type of shaper operation after general lathe milling to hold bearings so that its all as an inline operation as I can get.
I built medium sized robots and parts for artworks
-
For example here is a small 30cm robot I made a little while back on my bf20l converted to cnc
http://adamdonovan.net/files/Adam_Donovan_Work_2_Multiplexing_Tautophone.html
-
keeping my eye out for a bargain
http://www.ebay.de/itm/Rexroth-Servomotor-MSK060C-300-NN-M1-UG1-NNNN-/391015965804?fromMakeTrack=true
-
Ok not meant to be ebay spammy here but poking around all day for different deals and came across and AC servo with 4NM and a driver from china that seems quite decent for the near 500usd price. Cool thing about this driver is that its power supply is 220 240 which means for me no additional power supply needed and should be pretty easy to hookup.
I just asked them for the curve sheets and data now..will post it when I have them if anyone is interested.
http://www.ebay.com/itm/271735479951?_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT
-
I just asked them for the curve sheets and data now..will post it when I have them if anyone is interested.
The thread has been read 108,500 times because of the technical content and shared expertise and experience.
Please do not hijack this thread by posting ebay bargains on components you have no experience with. >:(
-
Truly inspirational work. I am itching to build one now as well!
A couple of questions if I may.
How is the 'universal' mount (shown on page 24) fixed to the spindle?
How is preload of the bearings done now you have done away with the threaded section on the spindle?
Thanks,
Dave.
-
There are two models of the InTurn™. The new ULTRA has a one piece Forged Chrome Moly Steel (4140) spindle with a D1-6/A6 mount.
The MEGA shown in this thread has a two piece spindle with a D1-4/A4 spindle.
D1 and A are standards with D1 being CamLoc and A being bolt on.
On the MEGA, spindle is two piece with the flange being a heavy press fit, heated to 550F and pressed on with a 20 ton press at that temp. It is exceedingly unlikely that you could dislodge such a press fit with the several times the 2HP available for the MEGA.
In addition to the press fit, the hardened steel cams for the CamLoc system extend thru the flange and into the spindle itself so the absolute maximum the flange could ever move is the .001" - .002" clearance in the Cam holes in the flange. The very first prototype MEGA was abused by my own hand for a few years before being sold when replaced by the new ULTRA. A few crashes bent the chuck's backing plate, but the spindle nor D1 flange was ever effected. A bent backing plate is easily skimmed to re-true it. Eventually, I made a thick cast iron backing plate which never bent even in severe crashes.
Early on, I spoke with the engineers at SKF about preloading the ball bearings. Deep Groove balls (as opposed to AC) are not specifically designed to be preloaed, but they have a lateral load spec and as long as that is not exceeded, there is no problem with preloading. Because the spindle is hollow, it has a very large OD which results in a huge bearing that will never see the loads it is capable of in this application, so there is plenty of excess load capability, even for drilling 1" dia hole into steel.
The preload is light and intended only to remove lateral play inherent in the bearings. The amount of preload is far less than what the bearings experience in operations like drilling into the center of a workpiece.
The preload is retained by the steel locking collar on the back of the spindle. The force exerted by backing out a stuck drill bit, heavy cuts on the back side of a workpiece, or using the wrong parameters for TPI on a large tap will exert significantly more force on the retaining collar that just holding the light preload.
The maximum predictable force on the locking collar is easily contained by the friction of the collar against the spindle, which is provided by a fine thread clamping screw. For the forces presented by the MEGA, a threaded connection is not needed, nor was there ever a treaded connection on this piece.
Perhaps you are confusing the MEGA with the new ULTRA. The ULTRA build videos clearly show the threaded retaining collar, but that is required by the optional tapered roller bearings, which are adjusted by clearance and not force applied. The 'micrometer' type of accuracy provided by the 20TPI locking collar is needed to provide this fine adjustment. The collar is still clamped in place using a large fine thread bolt.
In both machines, the locking collar doubles at the main pulley mount so there is also torque applied to the joint. Both applications work out fine on paper with 3x the horsepower ratings of the respective machines. There are lots of MEGAs out there and to my knowledge, only one has dislocated the locking collar. Most likely I did not set the collar tight enough from the get-go, but the user was able to re-establish the pre load and lock the collar down.
There are only a few of the new ULTRAs and only one had the tapered roller option, so it is a bit early to declare victory on that scheme.
-
I would have thought that a stepper can actually spin pretty fast if programmed correctly though perhaps im wrong there.
*Please anyone feel free to correct me if im wrong here*
I believe, that if you are running Mach at 35,000 mhz kernel speed, that implies that it can make 35,000 pulses per second, meaning 35,000 steps per second, which at a full step would be 175 rotations per second which is 10,500 RPM. Now, the caveat is, that is a 1:1 ratio which obviously wont work, and 10,000 is probably beyond the rated RPM for a stepper. If you throw an 8:1 ration on there, you're looking at 1312 RPM (at the 4th axis) with the stepper running 10,500 RPM. Also, as Simpson mentioned, the faster a stepper turns, the more torque/holding power it loses. So basically, you'd be at the point where it would be pointless not to do a Servo if you are planning to do high speed operations.
That all being said, I still went with a stepper because of cost, ease of setup, and I didnt need to do lathe work on my mill. Its all about what your intended use is for the machine.
Sorry, I missed this port on my last visit.
You invited: *Please anyone feel free to correct me if im wrong here*
Any while you are not wrong at all, there is a feature of most servo drives that you did not mention which (in many cases) negates the step frequency issue that you correctly identified.
There are a variety of names for the feature; 'step multiplier', 'pulse ratio', 'electronic gearing' and so on, but the function is the same: to alter the frequency or number of steps needed to rotate the motor. The limits and fineness of this adjustment varies greatly from perhaps a few pre-set choices in hobby drives, to being able to separately specify the numerator and denominator of the ration in industrial drives. Some drives it is as easy a specifying how many pulses per revolution of the motor as a single input number.
In the case of MACH running on the PP, typically the drive would have much higher frequency capability than MACH and the ratios for industrial drives can be very high indeed. "High-resolution encoder of 4194304 pulses/rev" taken from the Mitsubishi J4 series manual.
Without a very serious multiplication of the steps available from MACH, this motor would only be useful for moving something the speed of a farm tractor. :D
-
HIYA Steve, (OT) do yo have a picture of your 4th doing a radial type drill opertaion? I am working to finish a new WIZARD for 4th axis drilling and am in need of a good picture to go on the page.
A static picture (setup) would be fine if it just showed a round object in the chuck and a drill in the spindle just above the object. I will trade you a copy of the wizard for the picture (;-) I have a new TURN version for drilling as well for lathes to do offset face drilling. Do you have a picture of that as well?
(;-) TP
-
HIYA Steve, (OT) do yo have a picture of your 4th doing a radial type drill opertaion? I am working to finish a new WIZARD for 4th axis drilling and am in need of a good picture to go on the page.
A static picture (setup) would be fine if it just showed a round object in the chuck and a drill in the spindle just above the object. I will trade you a copy of the wizard for the picture (;-) I have a new TURN version for drilling as well for lathes to do offset face drilling. Do you have a picture of that as well?
(;-) TP
There are videos posted on both of these operations, but I should also have some quality pix around here, so let me dig a bit and I'll get you fixed up.
-
Terry,
I found a few very hi res photos of radial setups. As soon a sI get them edited and copyright notice on them, I will upload them to my ftp and put a link here for download (big files).
No pix of the horizontal, just video, but I have the original HD footage and can lift a decent photo. Here is the vid. Give me the number off the timeline of the images that you want and I'll pull a couple of pix off of it for you
https://www.youtube.com/watch?v=Dd_M6Li_OrA (https://www.youtube.com/watch?v=Dd_M6Li_OrA)
-
http://www.thecubestudio.com/pictures/RadialDrillSetup01.jpg (http://www.thecubestudio.com/pictures/RadialDrillSetup01.jpg)
http://www.thecubestudio.com/pictures/RadialDrillSetup02.jpg (http://www.thecubestudio.com/pictures/RadialDrillSetup02.jpg)
http://www.thecubestudio.com/pictures/RadialDrillSetup03.jpg (http://www.thecubestudio.com/pictures/RadialDrillSetup03.jpg)
-
Thanks Steve , here is what I am working on. IT can be a Wizard OR a page on the screenset
(;-) TP
-
Wow, Thet thar's one fancy Wizard!
Couple 'features' to consider adding:
*Hard tapping (a.k.a. Rigid tapping)
*Universal threading for the 4th axis (internal and external)
Also, it would be extremely convenient if you can add the lock/unlock macros via the wizard. If not, then provide instructions and links to editors that make this step very easy.
* CNCcookbook made special additions to their G-code editor to 'intelligently' add macros only where needed . . one would imagine by analyzing the context, but that's just speculation on my part.
*Notepad++ (this is free) has the ability to inject control characters into the 'replace' string so you can replace on line of Gcode with more than one line.
ex:
Find "G0A"
Replace with "M4010\nG4P.25\nG0A" in the entire file.
Anywhere there is N*********G0A*********
there will be
N*********M4010 (unlock macro)
G4P.25
G0A*********
Finding the spot fore the lock macro is trickier . . . needing that 'intelligence'.
One approach of many. Just a thought . . . ;)
-
The G84 button is on the list to do. You can already do it with a Bolt circle just input the 84.
I can do a button to insert the LOCK or UNLOCK into the saved gcode OR LIVE. AND make an auto insert function as well.
With this Wizard you have the option to run it LIVE or Save Gcode file (;-)
AND there will be a button to delete the last line of Gcode so you can erase the last line as many times as you wish to delete a function. It saves having to open the Editor and delete code.
I will LOOK at the threading but I am seriously out of room now (;-). I could expand it to 2 pages.
(;-) TP
-
Steve When Tapping do you prefer a { Feed per (unit)} feed OR {Feed per REV} feed ?
I prefer FPR as it is easiest to program AND you can SSO and the feedrate follows the RPM.
(;-) TP
-
Tapping preferences:
Based on comments thus far, it seems this wizard is MACH4 only? I was not aware MACH3 supported G84
That being said, my mill spindle is servo powered, so I use a custom macro for rigid tapping. The macro is a bit complex. It first gathers and saves certain settings, performs a software 'swapaxis' to 'C' axis, performs the tap action at a set speed and dwell and retract. Settings are then restored on exit to what they were on entry.
To 'jump' over obstacles, there is a retract parameter. The macro first compares the current Z with the retract and if they are exactly the same, it assumes the tap is at retract Z and does a rapid to zero before executing the tap action. With this macro all tapping begins at Z zero being the end of the tap at the top of the hole. This can be accomplished with a G52 if it is embedded in a larger G-code program. I tend to use it 'manually' most of the time.
With the exception of the retract mentioned above, all motion is incremental. This allows any arbitrary and precise tapping depth while maintaining the same starting azimuth. Which in turn allows more than one pass for deep threads in stringy or hard materials.
Having G84 available changes the landscape and I would agree that FPR would be preferable. I would only use G84 if M19 was also working and could be executed prior to the G84 for reasons cited above.
-
4th axis tapping:
ANY time I can TAP or DIE CUT threads, I take that rout . . . did I mention ANY.
Tapping to a center hole and/or driving a chucked workpiece into a die holder is stupidly simple so I tend to just write the code 'on the fly' into the MDI, but I think it would make a nice addition to a wizard.
The calc is quite simple and the whole operation is one line of Gcode . . well, two if you count backing out.
That brings us to single point threading, . . . the holy grail . . . . which I will comment on separately.
-
4th axis single point 'universal' threading:
This is the 800lb Gorilla, especially if you are trying to meet a spec. for thread fit. I have experimented and tried so many methods it's just ridiculous and while I have several successful 'universal' threading G-codes and Macros, including some that will cut any arbitrary tapered thread, the 'perfect' solution continues to evade.
I have tried the 'old school' 29 degree step down learned from crusty old machinists many, many moons ago. This works great for aluminum, but 'not so much' for stainless. Then 'quadrant' type cuts where the thread form is broken down into interconnected 'blocks' in the manner of an FEA mesh. 'Farming' or 'plowing' where successive vertical or horizontal rows are removed followed by a configurable number of 'spring passes'.
The two main positioning methods I have used are actual ID/OD or a 'surface start' at Z zero (or Y zero depending on where the tool is). I will provide, privately, some of the carcasses of these animals for you to dissect if you think it may be useful and of course, I am happy to discuss different methods, theories, or ideas on the best approach, but I do not have 'the' answer to this one. :-[
It would be great to get some fresh ideas on how to skin this cat. A theory to work on. I can cut threads, no doubt, but I don't have that grin . . . you know what I mean.
-
Mach does support G84 motions BUT it does not sync the Spindlle to the Zfeed. IT may be possible but i will not let that cat out of the bag.
A servo spindle can be refhome back to the index mark to set the Spindle to 0.000.
An M84 can be created to create a rigid tap function. It can have 3 outside param calls P,Q,R to set inside variables. The sequence can be done as a SUB so it SHOULD be able to simulate the G84 function quite easily.
Crusty OLD Machinist (;-) ???? I resemble that remark a great deal (;-).
I can shuffle some more room on the page so IF you get time send me what you have code wise for what you do. I will look it over and see what we can do with it. Top secret of course(;-)
Single point threading THE HOLY GRAIL. IT all depends (;-), we will just throw it all in the frying pan and see if we get fried fish.
(:-)TP
-
OK i made some room and have 10 extra function buttons available.
IF you can explain what axiss you use to do the internal external threads I will give it a go. Should not be that hard :D (famous last words).
A picture that shows a distant view that shows ALL of the machine would help.
(;-) TP
-
OK i made some room and have 10 extra function buttons available.
IF you can explain what axiss you use to do the internal external threads I will give it a go. Should not be that hard :D (famous last words).
The tool can be held in the Z or Y axis. Threads can be cut from the top or bottom (Z axis) or front or back (Y axis). Just what you wanted to hear . . LOL!
It sounds like unnecessary redundancy, but consider that cutting internal threads is best from a lubrication standpoint to have the cutter at the bottom where the oil collects, however, with a large diameter part, it can be difficult to reach the bottom with a cutting tool before the mill head contacts the top of the part with the tool bar on the 'far' side of the head. This arrangement is needed for long parts and puts the mill head above the part and the shorter tool overhang to cut the top of the hole becomes the determining factor for the setup.
Fortunately X remains X, so it is primarily a matter of exchanging Z for Y and positive for negative moves.
A picture that shows a distant view that shows ALL of the machine would help.
(;-) TP
As it happens, at this very moment I have an Ultra Spindle mounted for final turning and drilling the CamLoc flange, so I can get a lot of fresh photos showing exactly the setups you want.
8)
-
Should not be that hard :D (famous last words).
That's what I thought . . . . may I suggest that you not test with aluminum, but with DOM tube. Picture perfect threads on aluminum are all too easy. Clean internal threads on DOM tubing will earn you a case of beer.
Also remember that real threads are not pointy. They loose (((P/2)/TAN(30))/8) in height by spec. Given that bit of threading trivia, an uber convenient addition to any internal threading wizard would be a calculator that provides the correct actual ID for a given thread.
Then, when you are finished with the threading code, you can add the boring code ahead of it for 'on-stop-shopping'.
And as any good 'Crusty Old Machinist' knows, taper the back of the bore and lift the last thread to prevent creating stress risers in the part. ;)
-
Crusty OLD Machinist (;-) ???? I resemble that remark a great deal (;-).
(:-)TP
In addition to your machining experience, which is extensive, you also are a wizard at G-code, so I can think of no person better equipped to create a truly 'universal' threading program.
I prefer scripts to G-code because of access to better math, but I sent you a couple of my internal threading G-code programs as well as a G-code fragment specifically for calculating multiple passes in a specific way for specific materials. Hopefully there my be a morsel in there that will be helpful.
Now I am chomping at the bit to see what you come up with and test the whole system ( and I think system is the right word). Perhaps I can entice you somehow into expanding the system to take advantage of the InTurn™ Turning capability in prepping the part prior to the drilling/tapping operations.
Now, one more fish for your pan; provide the coordinated motion to bore and thread holes that are not in the center of the part. Example: simple flange with 4 tapped holes on a single BC. With NO horizontal spindle, this can be accomplished using a stationary cutting tool by coordinating AYZ motions to create an 'orbiting' motion in the part that will allow drilling/boring/tapping off center.
'Should not be that Hard' . . is a phrase that I heard somewhere . . . :)
-
A servo spindle can be refhome back to the index mark to set the Spindle to 0.000.
Yes, but is this part of G84? I was thinking that M19 should precede G84 in order to provide the functionality of my tapping macro . . unless I have the number wrong. M19 homes the spindle, is that correct?
An M84 can be created to create a rigid tap function. It can have 3 outside param calls P,Q,R to set inside variables. The sequence can be done as a SUB so it SHOULD be able to simulate the G84 function quite easily.
My macro is M9000 but it works via param passing with P being depth, Q being TPI and R being retract. I think I described the behavior in an earlier post. I'll take a peek and post it here if I am mistaken. Edit: yes response #291
If I stick the macro in G-code, it is typically arranged as a sub, so we are still on the same page with that . . . amazing . . :D
-
Yes the creast of the thread is FLAT not pointy . That is the OPERATORS responsibility to get the ID/OD correct for the thread.
The G84 routine would NOT refhome the machine in the real world it would invoke Spindle Sync where the spindle encoder drives the Z axis to make sure the motions of Z and the spindle are locked together in motion. It is NOT a problem with a servo spindle but is required with a VFD driven spindle motor. Might want to look into that part for your customers that do not have a servo spindle (;-).
M19 would be to position the spindle for a tool holder AND to REF the spindle to a known point for peck tapping. It can be 0.000 or any degree from there.It should be adjustable to allow for different tool changer setups. Simple Macro there.
G28.1 C0 ( Moves to C home index point)
G0 C180.000 ( Moves to orientation point)
The NICE thing about Mach3 is you are not limited to just 3 param inputs for a macro call the number is LARGE.
M9000 #1=1 #2=22 #3=333 ,etc,etc
You can input as many #vars as you like and MACH3 calls the #vars first then runs the Mcode (;-)
I would like to see the picture of the machine it will help me understand how it works axis wise. That way the code will be easier to write.
NOW it is a gamble to mix CB and Gcode in the same macro. Most times it works and SOME TIME it does not due to timing problems. (;-) IT can be PC dependant.
(;-) TP
-
Steve have you ever thought about thread milling instead of single point threading ??
Just a thought, (;-) TP
-
Yes the creast of the thread is FLAT not pointy . That is the OPERATORS responsibility to get the ID/OD correct for the thread.
That could be said for everything that your wizard . . or any wizard . . does. I don't understand the logic of taking one aspect of threading and excluding it so that the operator sits there with your wizard in front of him . . along with a calculator.
I was going to suggest another 'convenience feature' for you to add, that being an avoidance or clearance move added to the generated G-code to jump over or around chuck jaws. This occurred to me yesterday as I was adding these moves to a radial drilling program to keep the drill bit from visiting the chuck jaws. In this case it was not possible to simply increase retract because I was out of Z travel with this large part in the large ULTRA 4th axis.
The code to move 1" positive X before rotating the A and then returning had to be added to the sub. Granted, its not a big deal to add such things, but none of this stuff is particularly complicated , including any bolt circle, and it just boils down to convenience.
Whether clearance moves are 'operator responsibility' or not, they are frequently a necessary part of the process when you are work holding in a chuck.
Having to go back and edit the generated code is the difference between a good wizard and a great wizard . . . . just a thought. ;)
The G84 routine would NOT refhome the machine in the real world it would invoke Spindle Sync where the spindle encoder drives the Z axis to make sure the motions of Z and the spindle are locked together in motion. It is NOT a problem with a servo spindle but is required with a VFD driven spindle motor. Might want to look into that part for your customers that do not have a servo spindle (;-).
I added the servo spindle controls to the screen set for the people who purchased my spindles . . which are all servo driven. I decided to leave it there for customers to use if they ever move to a servo powered spindle. It is not actually a product in itself, so there are no 'customers' for that particular feature.
M19 would be to position the spindle for a tool holder AND to REF the spindle to a known point for peck tapping. It can be 0.000 or any degree from there.It should be adjustable to allow for different tool changer setups. Simple Macro there.
G28.1 C0 ( Moves to C home index point)
G0 C180.000 ( Moves to orientation point)
'peck tapping' . . there's a new one on me. I have not heard that term, but it is exactly what I do and why the starting point has to be maintained. The only difference is that unlike tool changes, the actual azimuth for tapping is unimportant, so long as it does not change. Well . . at least that's true for servo 8) powered spindles.
The NICE thing about Mach3 is you are not limited to just 3 param inputs for a macro call the number is LARGE.
M9000 #1=1 #2=22 #3=333 ,etc,etc
You can input as many #vars as you like and MACH3 calls the #vars first then runs the Mcode (;-)
Now you just made my day :-* I did not know this and have never come across anything that indicates this capability. Awesome Awesome. But wait . . . MACH4 also?
I would like to see the picture of the machine it will help me understand how it works axis wise. That way the code will be easier to write.
I was drilling on that spindle until past 7PM last night, but I'll get some shots today. Since this is probably general interest stuff, I'll post a low res version here and get a high res to you for use in your project.
NOW it is a gamble to mix CB and Gcode in the same macro. Most times it works and SOME TIME it does not due to timing problems. (;-) IT can be PC dependant.
I know it can be something of a Dance with the Devil, but I have been using named macros with MACH ever since I learned about them (credit to HimmyKabibble, Thanks Ray!). With the exception of the master screen macro, pretty much ALL of my InTurn™ control stuff resides in named macros and the numbered stuff does nothing but call the named stuff. CNC whisper down the lane, so to speak. Almost all of the timing issues go away because for whatever reason, MACH suspends until a named macro returns and it does not wait for numbered macros . . . unless something has changed.
There is a 'speed reached' LED on the InTurn™ controller and it indicated the status of a 'hold' on MACH processing until the 4th axis was up to speed. Otherwise, MACH would be moving a cutting tool into a part that was barely moving yet. Not good. Since converting everything to named macros, the LED is still there, but it no longer does anything except light up. The hard wired 'hold' to MACH is long gone.
[/quote]
-
Steve have you ever thought about thread milling instead of single point threading ??
Just a thought, (;-) TP
I own a couple of thread mills that see very infrequent use. I do development and prototyping under contract and for my own projects. Thread milling, in my opinion, is most applicable to production work. The problems I have with thread milling are manifold:
*Thread mills are stupidly expensive.
*One wrong move and they go to thread mill heaven . . . and I specialize in wrong moves
*Stupidly expensive
*Very limited reach
*Require uber accurate highly rigid machine. I have the accuracy, but 'not so much' on the stiffy :P
*Did I mention stupidly expensive?
-
Thread milling Remember this for a 4th/5th axis turn center not a simple CNC lathe
For large diam threads Thread milling beats single point cutting hands down. Manual TURN machines and simple CNC lathes did single point because they had no other choice. Your machine has a choice as it is NOT a conventional LATHE.
A single operation.
IF your machine can MILL an accurate circle you can do accurate thread milling. Does not have to be as rigid as you think
Cutter not any more expensive than a carbide boring bar with thread insert.
Can reach just as deep on large holes. Can even use the single point boring bar to do single point thread milling.
As to operators responsibility. All standard threads and profiles are listed in the Machinist Handbook. All crusty machinists will have a copy(;-). No calculator needed but you WILL have to put your glasses on , at least I do.
Creating the profile for the part to be threaded is normally a seperate operation as it requires different tools.
Before you begin the thread operation reguardless of what method the stock HAS to be the correct size. That way when the threads are cut to spec the profile is correct.
As to clearance moves a good wizard will let you EDIT the GCODE on the FLY to add such moves. This one does as well {EDIT GCODE} OR I can add an edit Button that is conversational in nature BUT you still have to be able to tell it WHERE and HOW to dodge the problems. The wizard CANNOT read the part in the Chuck. (well it COULD but (;-))
Rememeber Wizards are conversational AIDS to programming NOT a robotic programmer of parts(;-).
Just a thought, (;-)
-
Thread milling Remember this for a 4th/5th axis turn center not a simple CNC lathe
For large diam threads Thread milling beats single point cutting hands down. Manual TURN machines and simple CNC lathes did single point because they had no other choice. Your machine has a choice as it is NOT a conventional LATHE.
A single operation.
IF your machine can MILL an accurate circle you can do accurate thread milling. Does not have to be as rigid as you think
I have a 20TPI threadmill that I use for large holes in aluminum, where no tap is available.
Cutter not any more expensive than a carbide boring bar with thread insert.
Can reach just as deep on large holes. Can even use the single point boring bar to do single point thread milling.
Absolutely, however, I rarely snap off a boring bar. In terms of spinning a single point tool, I do this to cut ACME internal threads above a certain diameter. I have a solid carbide single point ACME thread cutter that I mount in a boring head. For shallow threads, I might use the boring head to increase the 'diameter' of the cutter, and for larger stuff I would interpolate.
Overall however, I think we are talking about different things. I agree with you as far as using a thread mill on the mill spindle. The topic was cutting internal thread on a lathe (or the 'lathe' setup of a mill/turn specifically). For starters, you would need some way to spin a threadmill. I do have a horizontal spindle setup for the mill head, but as you might imagine from the photo, typically it would not be able to be positioned to reach into the end of a workpiece.
You are speaking in absolutes, and in that context I cannot disagree with anything you have said. However, another conversation is in the context of what is practical on a given actual machine, and those two conversations are different. Like discussing steppers and servos in as the same topic because they are both 'motors'.
As to operators responsibility. All standard threads and profiles are listed in the Machinist Handbook. All crusty machinists will have a copy(;-). No calculator needed but you WILL have to put your glasses on , at least I do.
Wow, you really ARE old school! Machinists handbook next to a machining center. That is an amusing juxtaposition. Sort of like having rails in front of a saloon to tie your car to so it doesn't wander off. ;)
Creating the profile for the part to be threaded is normally a seperate operation as it requires different tools.
Before you begin the thread operation reguardless of what method the stock HAS to be the correct size. That way when the threads are cut to spec the profile is correct.
Well, a practical example is Collet drawtubes. I have to shave a tiny bit of the ID before threading. I use the single pint tool to do that <gasp. yeah, I said it! I don;t see any problem using the single point of a threading tool to shave small amounts to size the stock. Single point threading is the same operation. For those who get out of the shower to pee, this may seem like an abomination, but I just do it and not admit to it. :-X
So, seriously, the pint is that with a 4th axis capable of turning and indexing (InTurn™ or higher end HMC) the tooling can be the same, -or- you can rig a number of tools. Realizing that your scope is to write a Wizard and not solve world hunger, I still can throw in as many fish as I can find and see what comes out of the pan, to use your analogy.
Here is a link to a private video (i.e. you will not be able to find it on YouTube without this link) https://www.youtube.com/watch?v=wuzQhZnlBbA (https://www.youtube.com/watch?v=wuzQhZnlBbA)
Notice the multiple tools on the tool bar. In this case the hole is drilled for a tap, but it could just as easily be bored and single pointed. With Mill/Turn and gang tooling, you need to think outside the box.
As to clearance moves a good wizard will let you EDIT the GCODE on the FLY to add such moves. This one does as well {EDIT GCODE} OR I can add an edit Button that is conversational in nature BUT you still have to be able to tell it WHERE and HOW to dodge the problems. The wizard CANNOT read the part in the Chuck. (well it COULD but (;-))
Rememeber Wizards are conversational AIDS to programming NOT a robotic programmer of parts(;-).
Maybe you could just have a 'dodge' (I like that term) in X checkbox with a distance. That is really the only alternative when you run out of Z. The code would just as the X move out before the A rotation and then back in. These would occur in the same place as the unlock/lock commands. So basically
Unlock
--<is 'dodge' checked?>--
<yes> add G0X(distance var)
<no> no line added
Rotate A
--<is 'dodge' checked?>--
<yes> Add G0X(negative distance var)
<no> no line added
LOCK
-or-
universally add G0X(distance) where distance would default to 0 unless set by user
OK, enough chatter, here is a photo. You can download a higher res version here:www.thecubestudio.com/pictures/UltraRadialDrillingSetup.jpg (http://www.thecubestudio.com/UltraRadialDrillingSetup.jpg)
(http://www.thecubestudio.com/UltraRadialDrillingSetupWEB.jpg)
-
You best be "dodging" that chuck jaw in that photo (;-)
I'll add the button easy enough. It will ask you HOW FAR TO DODGE. And then write the dodge code OR do the move live depending on the record button's state.
WAIT:::: I already have a function to cover that. Look at teh MOVE TO POSITION section. You input where you want to move to and it will write the code for you to dodge things.
(;-)TP
-
Perfect!.
Let me know when I can have a copy of the wizard to play with. I am running a bunch of parts now that to not require the InTurn™ so it will be a while before I can do testing .
-
These are the new drive parts. Older trapezoid tooth belt is gone, replaced by newer GT2 type. Quieter and can handle a lot more power. This time rather than making a pulley from scratch, I took an off-the-shelf pulley and modified it by cutting out the center and mounting it to a stock clamping collar which is comes already drilled for mounting a pulley. Between the stock pulley and clamp collar, I saved several hours over starting from scratch. Collar is bored for a slip fit on the spindle . . no more threads. Preload will be accomplished differently. More on that later:
Spindle is out of a PhaseII brand 5C spin indexer. Still an import, but about three times as expensive as the normal junk and it is much better quality. Roughness in center portion of the spindle is from me reducing the dia slightly in that area with a coarse zirconia flap wheel to ease bearing press and reduce the interference fit on the rear bearing for preload purposes. Front bearing (next to shoulder will have full press fit. I will probably be making my own spindle, but for now this is easier.
(http://www.thecubestudio.com/pictures/4thAxis/Proto2%20Drive%20PartsWEB.jpg)
Hi all, apologies for waking up an old thread but it's an excellent thread.
How did you manage the preload - I can't find the answer anywhere :)