Machsupport Forum
General CNC Chat => Show"N"Tell ( Your Machines) => Topic started by: Hood on August 19, 2011, 08:15:55 PM
-
Just got a wee lathe yesterday, have been looking for one for a while just to play around with and got talking to Graham Waterworth and he mentioned he had one so I bought it :)
Its based on a Myford ML10 and had a 1/2HP DC motor on the spindle and a couple of steppers for X and Z axis. I have stripped things down a bit and have replaced the spindle motor with a 1HP AC servo I had lying around. I have a 400w AC servo for the Z axis and I am hoping to pick up a 100w or 200w one for the X.
At the moment the motor on X sticks out the front but I am hoping I can redo things so it sticks out the back, time will tell if thats going to be easy or not.
The toolpost is a Dickson quick change but I am thinking of doing a gang tool setup at the moment and maybe in the future I will make a turret.
Anyway heres a few pics of the lathe as was , the lasttwo are the original spindle motor and the servo I have fitted.
-
What a fun looking project, Hood, look forward to seeing it move along- :)
Dave
-
I am looking forward to seeing what it can do with the servos :) Think I will have to modify the ballscrew bearings a bit as at the moment they are only thrust bearings and on the Z the other end is just an oilite bush. Would be fine for steppers I suppose but think they may heat up a bit with servos, then again maybe I need some heating as its not exactly a tropical climate here ;D
Hood
-
Hood,
Welcome to the world of small and you will certainly have fun with it.
I put bearings on both sides of the Z and X axis on mine with one side able to adjust preload of both bearings.
The problem on having the X motor mounted to the back side is that there is not much to grab onto for mounting. I made a bracket
and things worked out. Another pain is ball screw size on the X axis since it's small in size ( small and accurate means expensive you will find ).
My x axis screw has a rather fine lead and didn't need any reduction / stepper is directly attached. Actualy it's a positioning ball screw and i was
after accuracy. Both of the screws have backlash adjustment for the nuts....so post tweaking the bearing preload i could then refine the nut preload.
For the x ball screw a tapped hole in the top of the carriage provides for getting to the nut adjustment. With both ends of the ball screw anchored / rigidly supported allignment becomes trial and error. Not much problem for the Z, but the x axis is very short and no give .......have gibbed x axis slide ...so the bearing mounting plates can be shifted some to provide for gib adjustment. In the end all this paid off.
I use a 1/2 HP motor on mine and when using the back gearing get plenty of torque and no spindle slow down when threading. Draw back
here is spindle speed though. But with CNC you can always use a fine feedrate to get surface finish.
A small accurate CNC lathe is just a joy to use,
RICH
-
Rich, its already got ballscrews on it, X is tiny as you say, its 8mm dia 2.5mm pitch. Z is a bit bigger think 12mm and 5mm pitch but not really looked closely as I have not had much time.
Mounting shouldnt be an issue at the back if I can do what I am thinking. The topslide had the motor at the front as in the second pic above, I reckon I should be able to just spin the top slide around so that the mount will be at the back. I have seen a few pics of the later versions of this lathe and they have the motors at the back so I think thats probably what they have done.
Hood
-
Hey Hood, is that the 3016 at the spindle ? 5000 max rpm ?
Curious, what is you planned ratio .... motor to spindle ?
Thanks,
Russ
:)
-
It is indeed Russ :)
Plan on keeping the ratio as is, 2:1 so 2500 max although I may limit it depending on how it sounds or if it sounds good may put it 1:1 :o
Hood
-
I've got one of the later ones with the x-axis motor at the rear.
Mine had had very little use, and I got it for a very good price from somebody who bought it from auction but couldn't get it to work.
I retrofitted mine with a SmoothStepper and C23 that I had doing nothing, and I bought a couple Gecko G251s to go with it.
One weird thing was the X-axis ball screw is imperial (3/8 x 10tpi), and the Z-axis is metric (12mm x 4mm)
Still not used it in anger yet, as I've had a rather hectic spring/summer so still not got past playing with Mach's built in wizards.
Full details of the woo's I had with mine can be found at http://www.mycncuk.com/forums/showthread.php/2619-Conect-121-Lathe-The-toy-collection-grows! (http://www.mycncuk.com/forums/showthread.php/2619-Conect-121-Lathe-The-toy-collection-grows!)
-
I have already seen it :)
Looks good :)
Where about are you, think its somewhere in Lothian if I recall.
I am up the coast a bit in Carnoustie but my workshop is in Arbroath.
You said your X ballscrew was 10TPI, is it definitely that as mine seems to be metric 2.5mm pitch.
Have the mounting brackets all worked out for the X motor, not sure if I have Alu to make it from, plenty steel but might be too heavy.
Heres some pics of what I plan.
Hood
-
I'm about 10miles east of Edinburgh.
The X-axis is definetly imperial, as it threw me for a while as to why I couldn't figure out the pitch in mm, until I dug out the micrometer and checked the diameter, at which point the penny dropped.
Althought I'm pretty sure I had to tweak the Steps per unit quite a bit from the calculated figures to get it accurate. If I get a chance, I'll check what figures I ended up with (I'm of on holiday next weekend, and I've got quite a lot of stuff still to do before I go!).
-
I'm about 10miles east of Edinburgh.
Bloody hell that must be in the sea ;D
Would be 800 steps per mm for the X if they were metric or 787.4015748 if they were imperial, thats assuming there is no gearing and I think you have Geckos with 1.8 deg stepper.
Hood
-
Bloody hell that must be in the sea ;D
Not quite!
Far enough away to have a hill between me and the sea :D
Just checked the settings, and X is 800.8, and the Z 500.
I must of had one of those stretchy rules the day I measured the X ::)
-
Bloody hell that must be in the sea ;D
Not quite!
Far enough away to have a hill between me and the sea :D
Not like me then, this is from the office above my workshop :D
Just checked the settings, and X is 800.8, and the Z 500.
I must of had one of those stretchy rules the day I measured the X ::)
Or maybe as they say, should have gone to specsavers ;D
Hood
-
Got a few bits done, have worked out how I am going to do the X axis and just waiting to get a motor for it.
Have the Z Axis bearing and oldham coupling housing made and fitted and axis tested.
Have fitted an old cabinet I had lying around to the end which will house the servo drives, SmoothStepper etc.
Going to use a NCR PC I have, got a few from a friend a while back and use one on the coil winder and this ones been sitting waiting for a project. Its a 15 inch touch screen with the computer inside the case. Its a Point Of Sale unit so has 12v and 24v USB as well as the standard 5v, I am thinking I will have enough current from the 24v ones to supply the I/O for my servo drives.
Few pics of progress and heres the axis getting tested.
http://www.youtube.com/watch?v=Kjtg5FATmAU
Hood
-
Nice, and looking good.
Interesting little PC ya got there. That's quite convenient for sure-
Dave
-
Yes, they are not the fastest, think 2 Gig Celeron with 512Meg ram but the one I have on the coil winders works fine with the SmoothStepper so this should as well.
Not sure where to put the computer though, in the pic its on top of the sliding door and it seems fine there, door actually slides better with the extra weight, the other option is make a wee arm to suspend it from over the headstock end. Might go for the latter as that will allow it to be a bit higher and have a keyboard below with a MPG.
Lots of I/O on the computer as well but sadly wont use much of it on this build.
Hood
-
I was hoping to make some progress on the lathe last weekend but unfortunately my 2 week old Tig welder packed up so that had to be put on hold as I am making the control console from 3mm Alu and the mig is just not the best for that. I have at least got the drive cabinet wired and fitted so thats at least some progress.
Boris from PoKeys has kindly sent me his new key panel and a USB and Ethernet PoKeys to test out and I will be using them on this retrofit. Here is a link to the key panel. http://poscope.com/product.php?pid=29 It looks much nicer in the flesh as well, I was very pleased when I saw it.
Below is also a pic of the control cabinet and also a render of the console I plan to make showing where the various parts will go. I still have to get a 60mm MPG as the spare one I have is 80mm and too big, that will have to wait though as the £2000 I have had to splash out on a decent tig has dented the funds a bit ;D
Hood
-
Looks good.
Although I'm now more interested in how well the PoKeys keyboard works!
I was looking at them at the start of the week, as now I've actually used my Conect for a few jobs, I'm already fed up moving between computer and lathe to tweak things.
I know I could add a few normal push buttons and wire them into the BOB, but I think I've only got room left for another 3 or 4, and I'd quite like to add an MPG aswell.
-
Not used it yet I am afraid, once I test it out I will let you know :)
Hood
-
Not had a chance to do much to either the Bridgeport or the Conect but I do have some progress. I managed to get the servo drive cabinet fitted, made the housing for the integrated computer/touch screen and have a hinging plate in the back for the SS/BOB/Pokeys etc. Also have the PoKeys Keyboard fitted and the trackball. Have to wire all the top electronics up and also fit the hard drive to the back of that hinging plate. Will be adding some switches to the front just above the keyboard.
I got a small servo that was going to be perfect for the X but unfortunately the encoder was cracked so I will have to see if I can manage to fit an encoder, not looked but I think shaft sizes will be different but it should fit inside the current housing on the end of the motor.
Heres a few pics of what I have done to date.
Hood
-
You do nice work there, Hood. The keyboard looks very "clean" and I would imagine it will take a bit of getting used to. Happy with the quality?
Thanks for sharing-
Dave
-
Hood,
Looks good! Going to be nice little machine... hmm... somehow it doesn't sound right - small machine in conjunction with Hood ;D
Did you go with servos just because you had them lying around? Besides, looks like an overkill to use a 400W servo on the Z of this little lathe?
Dan
-
You do nice work there, Hood. The keyboard looks very "clean" and I would imagine it will take a bit of getting used to. Happy with the quality?
Thanks for sharing-
Dave
Thanks Dave, yes keyboard is very nice indeed and a lot nicer in the flesh than it looks in pics. Quality seems to be very good as well, only time would tell how it would stand up but does seem robust.
The functionality isnt quite there yet, most of it does work via Brian Alpha plugin but there are a few things that are not quite right. Brian is wanting me to re-deign the key layout doing away with some of them and it will also be made to function as a standard keyboard as well with a function button, so it will be the only keyboard required :-) I just need to get time and get my head around what is in Brians thoughts as his mind is hard to read ;D. The other thing is Brain loves the way Fanucs work and I dont, I hate having to press 20 keys to do a task when one would do ;D
Hood
-
Hood,
Looks good! Going to be nice little machine... hmm... somehow it doesn't sound right - small machine in conjunction with Hood ;D
Did you go with servos just because you had them lying around? Besides, looks like an overkill to use a 400W servo on the Z of this little lathe?
Dan
Yes Dan I am taking bad with such a small machine but the build quality does seem to be there so its not too bad :D
Was originally going to go steppers, it had them on and I have spare G202's from the Bridgeport. But its just the slow accel and noise of steppers that gets me after being used to servos.
I had most of the drives and motors already and the 400w on the Z is overkill but better that way than not enough I suppose.
Hood
-
Hood,
Main reason I don't like large motors on small machines is I am afraid the motor would have enough power to damage the machine should something go wrong. But you're using a servo and can limit the current/torque to a value you like.
Would you share your E-stop diagram?
What size is the lathe? What are the travels?
Dan
-
Yes thats the beauty of servos :)
Not done the E=Stop yet, I tend to do things as I go as I find I work faster that way than trying to plan things out. Any time I have planned things I have seen better ways when actually doing it :) I will more than likely do them similar to the other machines, E-Stop, limits, drive faults etc all tied to the drives enables and sending a signal to Mach. As I wont be using a PLC on this it will be done with mechanical relays. The EStop will also kill the power to the drives but the others would just take the enables away.
Machine is a Myforld ML10 bed and headstock, about 350mm between centres, X was about 50mm but because I have fitted the motor to the back instead of the front I can get more, as much as 100mm I think, I am toying with the idea of a gang tool setup so the extra X will be handy. Swing is about 165 mm and over saddle prob about 115mm, the taper in the tailstock is No 2
Hood
-
Thanks for the info on the Estop, Hood. Does it work well on your other machines? I don't feel comfortable about merely cutting the power to the drives. They still have enough bus voltage to move a bit I think. I cut power to motors and switch in braking resistors. Depending on the resistors values the motor stop pretty much instantaneously.
Sounds a quite capable lathe.
Dan
-
The mills and big lathe stop instantly, maybe the weight/friction of the axis doing that but as the enables are also taken away from the drives the motor is not being powered. The bigger AB drives have internal shunt circuitry but the micros dont and require an active shunt if you want to use one, I have them on the Beaver but not the Bridgeport and there is no difference in an E=Stop, both Instant.
May be different on this machine as it will have less friction but I dont think it will be an issue, time will tell :)
Hood
-
Well not much progress, got a wee Yaskawa servo for the X axis, the wiring had been chopped off right at the exit so I opened to find out the pinout and thats when I saw the encoder disc was cracked in three places :(. Ok not so bad as I had a spare encoder with commutation tracks and it was for the same amount of pole pairs as the Yaskawa was :) Had to make up a new mounting plate, shaft sleeve (to make the shaft a bigger dia) and a new housing for the end, got that done and the encoder fitted and lined up and the motor ran. Lots of noise when it went above a few hundred RPM so decided it was the bearings, replaced them and all was quiet but next thing was I noticed the front shaft was bent. Managed to get that straightened fairly well and finished getting it dead true by turning it down, that was OK as I had to sleeve it anyway to fit the coupling. Got all that done and the motor is good :)
So now I have to make up mounting brackets and bearing housings and then it will be fitted.
I have decided to put things on hold for a while though as I plan to put the ESS on this lathe as well and want to use the breakout board I have been working on, have it designed now with a lot of help from friends who know about these things :) It is going to be three separate boards.
One will be 12 differential outputs and 5 differential inputs, they will also be capable of being used as single ended I/O if the need arises.
The second board will be 12 out and 5 ins, all 24v and each being capable of being sinking or sourcing.
The third board is similar but it will be 4 out and 13 in.
Reason I am making them as separate boards is so I can choose which to use depending on the needs of the machine, some will need more Ins than outs and others the opposite so I can use two boards the same on port 2 and 3 or two different :)
Screenshot of of the pcb software rendering of the boards as they will be made before I chop into three separates. Surface mount soldering will be new to me but I have been practising, just might need to get some specs :D
Hood
-
Oh and the reason I am making the boards wider than they need be is because I want them in Din rail mounting trays. the trays come in 108mm x 37mm sections and I couldnt quite fit them on to the 37mm wide sections so had to double them up to 74mm. No big deal for me but was a bit annoyed as I almost managed ;D
Hood
-
Hood,
Thanks for the PM heads up on this project. I have a few Q's
I understand the differential outputs, but what is the purpose of differential inputs? Are these foe encoder feedback or something similar?
Can you make these boards capable of being plugged directly onto a smoothstepper?
Incidentally, I know you have the ESS, so I'll sneak in a Q; is the ESS dimentionally similar to the original? i.e. a BOB made to fit the Original will fit the ESS?
How many axis will you have on a single board? Shall I just count the diff outputs and make an assumption?
Edit: I suppose it would be good to ask if you have alreay produced the boards and/or is tweaking a possibility. I am thinking about combining several of my boards into a smoothstepper compatible BOB and there is no sense in re-inventing the wheel. Typically I don't use OP work, but I have seen some boards that you have built in the past and they are high quality.
-
The diff inputs are for encoder inputs however I will be using a PoKeys for them now. I will be using this board along with a pokeys as it doesnt just have to be for the ESS. It means I can have a PoKeys with Diff I/O and also 24V I/O. That is one of the reasons I didnt design it for the ESS to plug into. The other reason I didnt do that is I can mix and match as the board will be cut into three separate boards before I populate them. I can then choose which boards I want. For example I may want port 1 with the 5v diff and port 2 with 12 out 5 in 24v and port 3 with 4 Out 13 In 24v or I could have the same boards on Port 3 as 2 or even have all ports using the same board whether that be 24v or 5v diff.
I will still likely put the encoder index pulse for the spindle in via the ESS and may also put the spindle encoder in if Greg gets Z sync'd to spindle.
Incidentally the diff I/O board can be used single ended if required, meant I had to add a fair few components to the Inputs but likely will be worth it for the flexibility it will give me as I plan to use it on all my machines and as said also with the Ethernet pokeys.
The Diff board is 12 out and 5 In so potentially 6 axis Step/Dir.
Yes the ESS is exactly the same footprint as the USB one, difference is Port 3, on the USB its 6 Inputs (3 being Diff) but the ESS has a full parallel port worth of I/O but none diff.
Any board currently made for the USB SS to plug into would be fine with the ESS as they all just had the two sockets for port 1 and 2. so if using one of them you would be able to add a ribbon to port 3 and take it to a single BOB or whatever.
Hood
-
Oh also MPG will likely be put in via the ESS so that would be put in as differential.
Hood
-
Thanks for the info.
You mention 24V inputs and from that I am assuming that you do not have isolation on the I/O.
What I did was isolate everything so it does not matter if the signals are 5V, 12V, 24V or actually any voltage up to the opto's rating. I this way, you do not have to have different components or different boards and the drives are 100% protected.
-
The 24v's I/O are isolated, changing resistors would be needed to change the inputs voltage, could be done but as they are primarily for me and a few friends we all use 24v so no need. Outputs are solid state relays so any voltage (within relays specs) could potentially be used on them I would think.
Hood
-
Starting again on the wee lathe.
Have decided not to bother making the breakout boards at the moment as I have purchased a CSMIO/IP-S, a CSMIO/MPG and a CSMIO/ENC.
This controller has all I wish for, 24v I/O Differential Step/Direction, Analogue I/O etc
It is overkill for this wee lathe but I decided to buy it and test it out on the wee lathe, if it works well then I will likely swap it over onto the big lathe and maybe put the ESS onto this one.
Heres a couple of pics, first is the back of the top control panel which houses the all in one computer/screen and on the back door I have the CSMIO controllers. The second pic is just a view directly down onto the back door so that the components can be seen. They are very professional looking and my hopes are high for it :)
Hood
-
They look very professional indeed. Can't wait to see how it works for you.
Dan
-
Got a wee bit more done the last few evenings. Made a panel up dor the switches/MPG/Pots/leyboard.
Wired up the rotary switches and MPG and tried them out in simulation and it seems to work well but without an axis hooked up I cant say for sure.
The MPG operation is more like a standard machine in that it has x1 x10 x100 positions, I actually have them set to x10 x100 x1000 in the config as I use metric units. The MPD motion is actually done in the controller rather than via Mach but the DROs get updated that fast that you would not know.
Waiting for the pots to arrive so I can get the Spindle and Feed Overrides wired up and tested.
Anyway so far I am impressed with the controller and the additional modules.
I am not sure what colour I will paint the actual housing for the control/computer, will probably see if I can get a match to the green of the lathe, I think the black panel and computer will look ok with that.
Hood
-
Well done, nice and compact, my friend-
Dave
-
Looks good and sounds good. Will be interesting how the MPG works. I have never been able to get the MPG work well with the SS nor the ESS. I want it to work in Velocity/Step mode like on commercial controls but it neither as responsive nor the step size is correct.
Dan
-
Thanks Dave, its a small machine and it needs to be compact, I am not used to that ;)
Dan
Step/Velocity mode has never worked since the day it was introduced as far as I am aware. I use Velocity and Multistep modes and quite like them. I have no need for accurate steps when in Velocity mode as I simply use it for jogging, turn fast it moves fast, slow its goes slow. When I want to sneak up on an edge I change to 0.01mm multistep mode and its perfect.
Not sure how I will get on with the MPG working this way as I dont think I will be able to get full rapid speed with the MPG and I am used to that.
On the scorpion I convert the MPG to 24v and use normal inputs, if I feel the need I could do the same with this controller I would imagine.
Time wil tell.
I found a pot last night and hooked it to one of the analogue inputs and it worked well. It seems to be filtered so that it steps in 5% increments but the one thing that wasnt quite so great was when at zero it did not stop motion fully. This used to happen with Mach when using Mach to do the actual overrides but Brian sorted that for me a while ago so maybe they can adapt the plugin to do the same. Its not a huge deal but I have got used to setting the FRO pot to zero and it not moving at all until I wind back up again.
Hood
-
Hooked up a servo motor/drive to test out the spindle control.
Had some weird problems at first such as having to tell the CSMIO that the encoder counts were double of what they were and also tell the drive the rpm/v were double. It turned out to be the drive that was not right for some reason as when I commanded a 1000rpm the CSMIO analogue output was measured at 2v and when I looked in the drive it also agreed that the command input was 2000mV yet the rpm/v was set at 1000 so in reality the motor should have been doing 2000rpm instead of the 1000 it was doing. I shut the drive down and restarted and this time things worked correctly and I had to set the encoder back to its correct ppr in the CSMIO config and also the rpm/volt to 500 in the drive.
So after getting that sorted things worked well, the speed seems linear and fairly accurate although there is a difference of 10 to 15 rpm from what the drive reports and what the CSMIO reports. This may or may not be an issue when threading, time will tell.
One other thing I found out is the Encoder input on the CSMIO-Enc module seems to be limited in frequency, if I asked for a spindle speed that took the pulses too high then the RPM would not display correctly, it seems to be around the 300KHz mark that it is limited to reading. To test I divided the drives buffered encoder outputs from my drive by 4 and all worked well.
I have also changed the port from being the feed override to the spindle override in the CSMIO config and it works very well with nice smooth transitions in speed control so that is looking good as well.
Hood
-
I think I missed something. Aren't you using the Ultra3000's? Why analogue? And isn't it the CSMIO/IP-A which is the analogue controller?
Dan
-
Dan,
the CSMIO-IP/S is the one I have but the only method of spindle control at the moment is Analogue 0-10v.
I am hoping they will allow an unused axis to be configured as a step/dir spindle in the future but at this moment in time its not an option.
Bit of a shame really, as although I dont know about writing plugins, I wouldnt imagine it would be too difficult, well thats what everyone says when they dont know whats involved ;D
Hood
-
Ah... that's a shame. Somehow I missed that part.
Dan
-
Was testing out the threading in simulation to see what it looked like but kept getting Resets whenever it came to the threading code. Couldnt understand as I had previously run some code from Mach that seemed to work. After a while the penny dropped, I had M4 for my code but the previous Mach code had M3. I changed my code to have a M3 in it and it ran fine so that was definitely the issue. That forced me to make up a relay board to invert the 0-10v analogue output from the CSMIO so that my spindle would actually reverse and sure enough the code ran fine whether there was a M3 or M4.
There still seems to be the delay in pullout at the end of the thread but it did seem faster sometimes more than others, suppose it will only be on a part that I will know for sure if its ok or not.
I also thought I would disable the servo drive so I could rotate the motor by hand and that way the CSMIO would see the encoder and hopefully thread. It seemed to track very well even though the revs were jumping around all over the place but again only cutting a real part would let me know for sure how good it was tracking.
One last thing I did was use the spindle override whilst a thread was being done and again the Z seemed to track the rotations nicely.
Hood
-
I'm glad to see I'm not the only one taking my time getting a fully functional machine.
I finally finished the gang tooled collet holder for mine today.
Made the collet holders from some silver steel I had lying around and I'm surprised how well the standard motors handled the threading (vid here (http://www.flickr.com/photos/mc_mtb/7818370534/) if you want to waste a minute or so!).
I'm interested to see how your's handles all the upgrades you've done. I doubt I'll do anything but running repairs to mine, as I could really do with something a bit bigger, although a bigger stepper on the Z would help with drilling...
-
Ha ha yes its taking a while but I put it on hold until I decided the way I was going to go.
What I am doing is way way way overkill for this wee lathe but it saves me putting the big lathe out of action while I test out the CSMIO control. if it works well then I will either buy another or put the one from this on the big lathe and the ESS on this.
Gang tooling looks good :) And the threading is excellent too :)
Is it the stock steppers and drives you are using?
Are you using the parallel port?
Hood
-
All the mechanics/motors are the original, however the original controls were dead to the world, so it now sports an original SS connected to a CNC4PC C23 with a couple Gecko 251s and a KBIC.
If I was to change anything, I'd swap the SS for a ESS, and use a PMDX board, however I already had the SS and C23.
I would like to add some extra functions via PoKeys, however I'm trying to resist spending anymore money on it until it's made me some!
-
Wow, mc your ER gang tooling attachment is great! I couldn't tell from the pic, does it bolt down to the carriage or clamp in a QCTP?
I'm in the planning stages for a twin QCTP attachment for my little ORAC. Could you take a few notes and post a "Tools & Offsets For Dummies" how-to thread when you start setting up Mach for yours? Pretty-Please? ;D
-
The holder drops into the QCTP, and I've added a support (pic here (http://www.flickr.com/photos/mc_mtb/6350728231/in/set-72157625154378941))for the far edge to rest against. It's only intended for drilling/reaming, so it shouldn't see any up/down forces, just end on forces, unless I somehow manage to issue an X move with a drill in hole.
To mark the original holes in the bar (the Z-axis didn't have enough power to sink the centre drill into the steel), I used G52 offsets as per this thread (http://www.machsupport.com/forum/index.php/topic,20153.0.html). I still need to get stub drills in the required sizes to try things properly, and figure out how I'm going to code the gang tooling. Actually, I'll go and update that thread!
-
Thanks for the great info mc; that's really given me some hope & ideas for the future of my ORAC!
What was the limitation drilling steel, spindle power or Z-axis power?
ps: Sorry for the thread drift Hood.
-
It was probably a combination of the Z-axis lacking power, and a not very sharp centre drill, but it wouldn't do any more than mark the steel.
-
Hi Hood,
Have a question for you. I know you usually use PLCs on your machines. So it made me wonder why were you after a controller with lots of I/O and being 24V? Wouldn't an ESS/PLC combination make up a much cheaper solution? And possibly lots more I/O?
Dan
-
PLC is relatively expensive
PLC = £255.07 thats 20ins 16 outs
and each extra I or O or analogue module adds much more to the cost and then there is a limit of 4 extra modules.
Analogue module = £77.43
Extra 10 In module = £57.87
Extra 8 relay outs = £59.87
So not that cheap and add on a breakout board or two for the ESS as well
ESS gets almost unused as its 5v I/O is not any interest to me except for MPG and Step/Dir, thats why I was thinking of making up the 24v breakout board.
Step/Dir is single ended, I want differential so I have to make up a line driver.
Making things how I want means lots of add on boards where the CSMIO is all integrated.
Thing I like about using a PLC however is it is easy to do all sorts of fancy things in the ladder and fairly easy to do it where doing things in a macropump or brains gets a bit messy.
I keep hinting to Brian that Mach4 should have ladder logic but whether it will or not only time will tell.
Hood
-
OK, I see now. Thanks. Although I would have to check to be sure, but I think I could get here a Delta PLC for at least half that price.
Dan
-
Even at £200 for a PLC plus ESS of £120 and a BOB of say £120 (for 2 port one) then another single BOB maybe £50 then that would be almost £500 and no 24v or diff on the ESS and no real integration between the lot.
For a machine that you are happy with 5v I/O no differential I/O etc then the ESS is an excellent route. If only the Step/Dir spindle was sorted as the rest of the niggles that the USB SS suffers from have been cured with the ESS.
Hood
-
Little bit of progress, got the lower half of the housing made up, had some 3mm steel left over from a job so tigged up the part where the keyboard sits out of it. I dont normally have steel as most of my sheet work is either 316 stainless or 5083 Alu.
I extended the front of it out to the edge of the coolant tray, that allowed me to put a better slope on the keyboard, previously I felt it was too steep. Have given it a coat of primer and hope to get some paint to match the original green but if I cant I may just paint it grey or white. Should manage to paint it in place so I can now get on wiring things up.
Anyway a couple of pics of progress.
Hood
-
Coming along quite nicely Hood, looks great !
I'm anxious to get a performance report on the spindle servo arrangement.
I have an identical set-up waiting on the shelf for a retro here.
First inclination is that the 3016 is borderline but hoping for better.
Nice work .... as usual.
Russ
-
Very nice, Hood! Do you cut your sheet metal parts yourself?
Dan
-
Its just a wee lathe Russ, the original motor was a 1/2Hp DC motor and the 3016 is 1Hp so it should be plenty if the original was. If not the orig motor was 2000rpm and geared 2:1, I have kept the gearing at 2:1 but the 3016 is 5000rpm so I could gear by 4:1 and still have more than the original speed.
Time will tell though as you say.
What size lathe are you planning on putting it on? This is only about a 6 inch swing lathe, its based on a Myford ML10.
Hood
-
Yes Dan, got an angle grinder that makes short work of it :-D
Hood
-
Hood,
I have 2 of these. McLean #500
5C collet w/pneumatic closer.
1 hp DC, 1/1 (roughly)
The current control is just timers and relays for the cycle. Very basic and simple.
Will neen one axis with live tooling and C for the spindle.
The 3016 reduced 1.5 to 2 / 1 would be OK for speed, just looks a bit dinky.
I'm probably under estimating the servo. :)
Thanks,
Russ
-
Yes Dan, got an angle grinder that makes short work of it :-D
Hood
Ah.. OK. It just looked so good that I thought you acquired a Plasma CNC and didn't tell us ;)
Dan
-
Russ,
looks a solid wee lathe and with it having a 1HP at the start you should gear the 3016 so its around the same max RPM as the original motor was. That way you will still have the same HP as before and plenty torque I would imagine.
Hood
-
Dan,
No room here for anything else, wouldnt mind a plasma but its not likely to happen. I often mill bits out though, the Beaver has 800 x 400 travels so its not too bad for things like that.
Hood
-
Hood, as somebody who's removed the steppers, how were the ballscrews attached to them on yours?
-
It was just a drilled out piece of steel with rollpins through the motor and ballscrew.
The backlash take-up was relying on the motor bearing and a single thrust which was pretty poor IMO.
Hood
-
Cheers.
Sounds just like mine, but I never went any further than taking the ballscrew cover of the Z-axis. From what I remember, mine never looked like it had any substantial bearings anywhere, so I guess they never changed that part of the design.
I'm weighing up a couple upgrade options, and it looks like modifying or replacing the ballscrews needs factored into the options.
-
No need to modify the ballscrews really, well as long as they are ok. I used two thrust bearings, one at the motor end and another at the tailstock end then just used an oldham coupling to connect screw to motor.
Hood
-
then just used an oldham coupling to connect screw to motor.
Does it not introduce backlash?
Dan
-
No, ballscrew is held between the two thrust bearings so it is held tight.
Hood
-
How is your ball screw supported at the tail stock end?
Mine is machined and simply sits/runs in what would be the hole for the original leadscrew. It doesn't look as though it's secured in anyway, but I guess I could drill/tap it and come up with some method of adding a thrust bearing to hold the entire screw under tension...
-
I made up a wee housing to accept the ball thrust bearing and it is simply held there against the end bracket by the ballscrew pusing against it, the other end is the same so the the ballscrew is kept captive. I just use the half of the oldham coupling to keep the preload on the bearing, I nipped it up then lichtly tapped untill I was happy then tightened the capscrew fully. It should be good enogh to keep it from moving but if not I will drill the coupling and screw for a pin.
Hood
-
Got the lower cabinet wired up with the exception of the main power. Have the wires run up to the control and hope to get them connected soon.
I got some 2 pack paint that, although not a perfect match for this lathe (actually closer to the big lathe) it is good enough and came out relatively well. Need to redo the back panel as there are a few runs there as the paint was setting when I painted it so applied it too heavily, nothing a sand and re-coat wont solve I hope.
Hood
Edit, just looked at the pic and the paint looks a lot lighter than it is, if you look at the big lathes control panel in the background you will see the true colour as I gave that a coat with the paint that was left.
-
Got the spindle motor tested through the CSMIO today and it worked fine and was nice and quiet. Had to do a bit of jiggery pokery with the encoder output of the drive and what I told the CSMIO it was, seems like the ratio in Spindle Pulleys is not getting seen by the CSMIO.
Tested the spindle up to the full 2500rpm (5000 motor) and again it was fine and quiet, tried at 10rpm and a few others in between and all was well.
I then thought I would do a bit of testing for Russ so went and got a brush and set the spindle to 500rpm (1000rpm motor) and shoved the shaft between the chuck and bed, took a fair amount of force to slow it down so I think it will be fine at the 2:1 gearing.
Problem now is the chuck has tightened up that much I cant get the bleeding thing off ;D It is now also rubbing on the steel plate behind it so will have to remove the keyboard/switch panel so I can get a spanner on the spindle flats to get the chuck removed. I will then have to machine the back of the chuck a bit so that it cant rub on the steel. Dont think it would ever normally tighten up that much even under heavy cutting but might as well do it just in case.
I am actually thinking of making a new spindle as I am not a fan of the threaded spindle noses as if I ever want to use M4 then there is a chance it could undo.
Hope to get the axis moving this weekend but not sure if I will get the time.
Hood
-
Good news, the ol' Broom Handle Jam" method of torque verification.
(same as I use here btw)
Thanks for the test result Hood, anxious now to see a heavy cut vid.
Regards,
Russ
-
I too don't like the threaded spindle on my 9x20 and had the same problem with the chuck tightening up while I was tuning the servo. Took some serious effort and fair extension to the keys to unthread it. Have also been thinking of making a new spindle for it, but haven't yet decided how I want it. If you have ideas would be interesting.
Dan
-
Not going to be many heavy cuts on this lathe Russ ;D Well certainly suppose things are relative, what would you call heavy cuts?
Dan
was thinking of a D spindle nose at first but then thought an A or B would be better or rather easier, same taper but no cams to worry about.
Hood
-
Does your's have flats on the spindle?
Mine doesn't, and I was under the impression any flats would of been on the original pulley block/back gear.
-
Yes its just behind the front bearing housing, heres a pic of it and a close up so you can see the actual flats.
Hood
-
Hood,
And will you just make a new A type spindle nose and permanently attach it to the current spindle? Or will you make new the whole spindle with the nose?
Dan
-
I will just make a complete spindle, thats if I ever get time to get round to do it ;D
Hood
-
Managed to get the Z hooked up and tested, have it set at 10m/min and its working well.
Also tried out the MPG via the MPG module and I am very impresed with the precise, smooth motion even at the x100 setting.
Heres a quick vid of the MPG being used.
https://www.youtube.com/watch?v=Xz62bOHCc_M&list=UUNEcFI4fdD5kV938uchW3sg&index=1&feature=plcp
Hood
-
Feels better than with the SS?
Dan
-
Yes Dan, by a long way. The SS/ESS does things the same way as the PP and in multistep mode it is sort of clunky, the CSMIO is very smooth just like you were commanding a feed. One other major difference is the PP/SS etc will buffer moves so for example if you have it set to multistep and 0.1mm increments you have to be careful you dont wind too fast or the axis will continue for a while after you stop, set the steps bigger and its even worse. The vid shows me at first in steps of 0.01mm then 0.1mm then finally 1mm, there is no buffering at all and the movement is very smooth between clicks.
Hood
-
Some nice progress here, Hood, looks great-
thanks for sharing-
Dave
-
Thanks Dave, just waiting for a motor for the X coming from your side of the world so wont be much more progress until it arrives :(
Hood
-
Thanks, Hood. That sounds really good! Got two Cincinnati machines (a mill and a lathe) awaiting retrofit and I am quite convinced now that the CS Lab controller is the way to go. Although the delayed pullout in threading would a be a problem on the lathe. Hope it gets sorted by the time I get to it.
Dan
-
Yes, looking good so far and hopefully it continues the same way. I like the SS and ESS but sadly Greg just didnt seem to want to go for more industrial type controllers, may be a mistake on his part, maybe not, time will tell but I think the Mach user base is moving away from the hobby only world and getting into production/industrial sized/style machines. There will always be the market for basic controllers that take place of the parallel port but the future, at least in my mind, is industrial quality and spec controllers.
Regarding the threading, not managed to do any testing with axis yet as I am waiting on a small motor arriving from the USA for the X. On screen its really hard to tell if it does pull out or if its delayed, some times it looks like it does, others not so only real axis tests will show it up. If it has the delay then they will need to get that sorted before anyone will take the controller seriously for a lathe and it would be a shame for that not to happen as so far it does look like its an extremely accomplished controller with industrial qualities.
Hood
-
Here's a little thought provoking issue for you.
With the (pretty heavy) X-axis motor mounted on the rear, will there be noticeable movement/vibration issues when using a front mounted cutting tool?
I'm probably over thinking it (I'm stuck in a **** hotel and never slept much last night!), but with such a large motor on the rear, it's going to cause the whole carriage/cross slide assembly to tilt towards the rear due to any play/flex. When you then take a large enough cut with a front mounted tool, it's going to counteract the weight of the motor, and the assembly will tilt to the front.
The only reason I've been thining about it, is I want to upgrade the x-axis motor, which involves a heavier motor, and moving it further rearwards to allow room for a bearing block and a proper bearing block.
-
The motor I am putting on is an Allen Bradley Y-1002, it is 40mm square and 70mm long and weighs 0.5Kg so I dont think I have to worry about anything like that. If the gib is adjusted properly then even a heavy motor should not affect things anyway.
Hood
-
Hi Hood,
Have you had a chance to test the threading yet? Really the only thing I am waiting on before I buy one. I will go with the analgue version though as I mentioned to you earlier.
Dan
-
Afraid not Dan, things happened too fast with the Chiron, had to get rid of two machines and get the Chiron in all within a week and the workshop is still a mess so cant get in to work on the Conect. I have the X motor now but as its a 4500rpm motor I am going to redesign the X mounting and make it belt drive with 2:1 reduction. Will be a week or two at the earliest before I get to that stage.
Hood
-
OK. Please post when you have some tests made.
Dan
-
Managed to spend a few hours on the Conect today and got a motor mount almost made up. I have temporarily used some spacers to seperate the two plates but will be machining a piece of 40mm thick Alu to make a complete housing, well thats assuming I can find some at the right price ;D If not then I have some 20mm thick that will do, will just have to make it two part.
Hope to get some time in the next day or two to get the motor wired up and I can then do some testing of the CSMIO for threading.
Hood
-
Here is the first tests of the conect running both axis. I am simulating a thread in this video and whilst it is running I am overriding the spindle. The Z seems to track the spindle speed closely but only actually cutting a thread would show whether it is correct, I still have toolposts etc to alter before I can do that. I think there is still a delay on the X at pullout but again only cutting a thread would show that definitely one way or the other. If it is I hope that can be sorted as having an annular groove at the end of a thread is not always the ideal situation, fingers crossed ;)
http://www.youtube.com/watch?v=HC1PK6uD26U
Hood
-
Hi Hood,
Thanks for the video. Z axis tracking seems good, but it definitely stops at the end of each pass. However, I can't tell if it stops after the X pulls out or before.
Dan
-
Dan it is hard to tell even in the flesh. Sometimes you think it has pulled out others you think it hasnt. I am tending to think the latter unfortunately but even if it is I am sure it can be fixed relatively easily. I talked to Brian about it a while back and he said they had done it in the Galil no problems so the resync with Mach was after the X pullout so it should be possible with the CSMIO as well with just some alterations to the plugin. Hopefully I will get a test done soon cutting some real threads and will have a definitive answer to what its doing.
Hood.
-
Is it something that Brian has to take care of?
You mentioned this to CS Lab a few months back so I wonder why they haven't taken care of this yet. Especially when you say it's a simple one.
Dan
-
No its in the plugin it needs done.
Never said it would be simple. I just said I think it should be relatively easy but as I know absolutely nothing about plugins thats why I said think it should ;)
I did another run today and held my finger against the X so I could feel the movement. It definitely pauses before the retract, what actually happens is Z completes then a pause then X and Z move to the clearance amount then Z continues alone back to the start. So sadly its not perfect yet. You could cut threads with it as long as an annualr groove was acceptable at the end but for the majority of the threads I do it is not ;(
Hood
-
Did you contact CS Lab again?
Dan
-
Not yet Dan, too many things going on at the moment to concentrate on this, I like to be able to devote time to situations like this if there is any testing needing done. .
Hood
-
Managed to get the toolpost altered today and tried cutting a thread, it worked well. I then re-ran the cycle again but this time I varied the spindle override as it was moving and the Z tracked the spindle nicely following in the track of the previously cut thread.
Heres a vid.
http://www.youtube.com/watch?v=sKp343Ubf7U&feature=youtu.be
Hood
-
That's good Hood. The bits from CS labs are looking very promising.
Brett
-
Yes fingers crossed :)
Hood
-
Very nice! Have you ever tried this with the ESS? If I do this with the ESS it faults out here. It tracks nicely at times but at times it would fault out - mostly if I increase the SRO.
Dan
-
Dan, I cant try it with the ESS as the spindle override is disabled due to the problems with it and the ESS. If I enable it then the spindle every so often has its pulses momentarily dropped. With it having a heavy chuck there is a violent decell/accel which I do not like as it puts a lot of mechanical strain on the machine, so it is unlikely I will ever have spindle override enabled again whilst using the ESS.
Hood
-
Hi Hood,
A quick question regrading MPG with the CSMIO. Does it work in Step-Velocity mode as well? Or just multistep? What is your step size at x1 ?
Dan
-
Dan,
it works in a totally different way than you are used to with Mach, it is more like other Industrial controls.
You just have x1 x10 x100 modes, for metric you choose x10 x100 x1000 in the plugin but I will just refer to as x1 x10 x100 to avoid confusion.
When it is at x1 then each click of the MPG is 0.01mm, x10 its 0.1mm an x100 it is 1mm.
With Mach or Smoothstepper etc you have Velocity, multistep or step, with the CSMIO you dont have any mode.
When in Velocity mode with for example the SS the click of the MPG is not an accurate move but with the CSMIO it is, so it is very easy to move to exact positions and unlike Multistep mode it is very fluid in movement.
Hood
-
Thanks a lot for information, Hood. Just what I needed to know.
Dan