Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: russh on July 07, 2016, 05:16:44 PM
-
Hello All,
I thought I would restart this thread after a misunderstanding regarding my Mach 3 license which has now been resolved with Tweakie.
As I originally wrote, I seem to be experiencing an issue with my machine that results in it loosing its positioning.
As an example, I have made a video (see below link) in which I set the work offset zero at the surface of the work piece. I then rapidly move the axis around a little, and re position the cutter on the work piece surface. Only to find that it shows up as not being as Zero in Mach 3
https://www.youtube.com/watch?v=gR1-5u9HwIs (https://www.youtube.com/watch?v=gR1-5u9HwIs)
Just to give you some detail, the machine uses NEMA 34 Stepper motors, operated via a JNC-40M control board connected to a new PC via USB. All the electronics (Input Transformer, Control Board, Motor Drivers, etc) sit in the same enclosure, all close to each other.
From what I have read (as Tweakie pointed out), the JNC-40M controller has a pretty hit or miss success rate of being stable with Mach 3, so Im wondering if this is the cause of the problem?
Before I go buying a new control board, I guess I would like some feedback from you guys of if this seems to be a case of losing steps, and if there is anything specific I can check (maybe with the electronics such as feedback, etc)?
Any help would be greatly appreciated
Thanks
-
Sounds like this is a NEW setup?
I have no experience with the control board, but lots with grounding issues ;)
You say everything is pretty close together, so control, power and sensor lines are pretty close to?
Did you utilize a star grounding point? If everything is so close together, it can cause cross talk in the lines.
My setup had everything really close together too, and gave me problems until I ripped the wiring apart
and re-wired the works, paying close attention to detail with grounding EVERYTHING to a single point.
I'm not losing steps anymore.
-
Thank you for your reply.
Could be a huge bonus having you help if you have loads of grounding experience, as that could be the first starting point.
The following might help
I am in contacted with a chap over in Australia (Im Based in the UK by the way), who also bought one of these machines (by the way, its not a new setup I have made, its an actual CNC bought from a Chinese supplier). I believe the machines are based around the mechanics of the old Makini CNC but with revamped motors, drivers and electronics (apparently!)
Anyway, this other chap has been having issue of a different kind, but possibly all down to the same design problem.
He was finding an issue with the spindle drive that would cause sporadic operation on the spindle speed at low RPM. He found that the problem with the spindle drive was cross talk between the output and input of the axis stepper motor drivers, they were sending noise back into the transformer, then into the motion board and on into the spindle driver.
Best I be honest up front, I am from an electrical engineer (Im a mechanical grease monkey). So if you could share any help and steps in baby talk.....that would be great. But it sounds like this could be a good starting point if you all agree.
Therefore, what do I need to look at doing? Do I need to separate components, separate wires, fit ferrites, etc ??
Thanks again all
-
Watching your video, it shows at the end that you reach the top of your work surface at -0.73mm. If you were to lose steps, you would reach the work surface before the Z axis reads zero. Normally losing steps on the Z axis results in plunging into the work piece. It appears as if you were gaining steps when Z moves up resulting you reach Z zero above your work surface.
I assume that moving Z up is the Z+ direction?
Gaining steps could be a noise issue as Mogal indicated earlier or a PC problem. Is your machine grounded to the mains power outlet, the wall socket?
-
Hi,
I think you are right, I am gaining steps more often that loosing, as when I have ran the test other times, even though there are times it looses steps (plunges down), most of the time it results in the tooling being higher (+) than the set zero point.
I need to examine the electrical setup in more detail, but when you say grounded to mains, do you mean to the mains earth or the neutral? (Uk has Live, Neutral and Earth which are Brown, Blue and Green respectively)
I will take some more close up photos of the electronics later to share, but if it helps for now, the following photo is of a similar machine with pretty much that same electronic setup as mine.
(https://s32.postimg.org/5d7urtred/DSC03427.jpg)
-
Yes, Europe has a green or yellow/green wire for ground or earth as you call it. Is your machine connected to a wall socket that has an earth connection or do you just use the 2 power wires Phase and Neutral?
A few pictures of your electronics might help. Is this problem only the Z axis or do the X and Y axis show the same behavior?
What steppers are you sing, what drivers and what power supply set-up?
-
Hi,
It connected to a wall socket (well a isolation switch), that includes the earth connection. The cable from the machine (that connects to the wall isolation switch) also has earth. If that earth cable is connect to the actual machine somewhere I can not answer at the moment, but I will check on this.
The issue exists in all the axis (should have mentioned that sorry), but seem more exaggerated in the Z.
I will get some picture of the power supply setup, drivers etc. But there are not labels on the NEMA 34 Stepper motors so I have limited information on them
The earth system of the mains supply to the factory is a TT (Terra-Terra) earthing system with a local earth electrode
-
Hi All,
I have taken some pictures of the electronics.
Couple of points I noticed
1) While the machine uses the JNC-40M drivers, the PCB control board seems to be made by the machine manufacturer, so Im not sure he they have done a spin off of it.
2) There are 4 wires going to the stepper motors, not sure if that helps
(https://s32.postimg.org/tr2nmb1k5/20160708_143436.jpg)
(https://s31.postimg.org/uzcajf8fv/20160708_143621.jpg)
(https://s32.postimg.org/ynpmn6dat/20160708_143454.jpg)
(https://s31.postimg.org/4z6ppv1iz/20160708_143657.jpg)
(https://s31.postimg.org/nwmc36i3v/20160708_143509.jpg)
(https://s32.postimg.org/9h24fcn85/20160708_143650.jpg)
(https://s31.postimg.org/cmj3ifymz/20160708_143747.jpg)
(https://s32.postimg.org/9l3r15vad/20160708_143736.jpg)
-
This does not look bad at all. I would start by untangling the wires for the steppers from the USB card, it looks like they are on top of the card. Secondly I would unplug the green connector blocks from the stepper drives and check for lose or bad connectors.
Also check the connectors on top of the USB card for lose wiring.
I would measure the voltage to the stepper drives, should be close to 70VAC but not higher then 80VAC. Measure for some time and see if it spikes up over 80VAC. Your stepper drives are Leadshine drives after I Googled the partnumber in the picture and these are good drives but don't like overvoltage.
Do you have access to another computer to run Mach3 on? Just to see if this makes a difference.
-
Thanks G21,
I will do check all the bits you mention
I've actually tried two PCs already with Mach 3, and the same issue occurs (one was an old one I had laying about but good spec, the other one was a brand new built from scratch job)
One out of interested question, as the steppers have 4 wires, does that me they are the better "position feedback" type
-
I had the same problem but mine was the acceleration and deceleration of the steppers in the Motor tuning. I had not allowed time for the stepper to slow down. This was more pronounced in the Z axis due to the weight of the vertical column overriding the steppers. Mach counts the steps then moves on, if there is overrun being as there is no travel feed back it continues with the next command. Just though I would throw in my thoughts.
-
I think the wires to the stepper drivers are on the right of the USB card???
-
I think the wires to the stepper drivers are on the right of the USB card???
Hi, what did you end up slowing your acceleration and deceleration down to?
-
Trial and error, before you do anything put some slow feeds in and see if you loose steps if you don't it has time to slow down so keep adjusting till its gives you the correct position (manual measured)
-
Ive actually had this issue for some time now, and posted quite a few months ago, the recommendation then was to play with the motor tuning, and while it did start to stabilise at low acceleration and deceleration (although still loosing steps slightly), unfortunately it was unrealistically slow
-
Not sure how the velocity V acceleration/deceleration works but I think you can maintain the velocity its just the amount of steps you use to decelerate. One other suggestion is to check you jib strips especially on the Z axis if they are to loose the head drops so fast it overruns your intended stop, I used a digital vernier clamped to the table and kept mechanically measuring the overrun until it stopped at zero. adjusting the jib strip helped with the fine tuning. Early hobbyist machines had a damper/assist shock absorber fitted to solve the Z axis over/under run. If your main problem is with the Z axis that could be your problem
Hope this helps
Jim
-
just checked and the acceleration/deceleration can vary from 0 to 0.25 seconds at the start and end of the axis travel whilst maintaining the velocity.
-
Ok just looked at your video, if its just your Z that's loosing position that's the problem. I noticed you have a dampers fitted, is it still working correctly and then tighten up your jibs. If you overtighten them the Z will loose steps in the Z+ too loose it will be the Z -
-
what are jibs?
-
They are usually strip of metal inserted between the column and the spindle head Vee slides. They take up the slack in the wear of the slides. My Machin has 3 grub screws you adjust to push the jib out to adjust the possible play in the slides. Don't overtighten as you might put a strain on your motors. Jim
-
This machine has no jibs, is all linear slides. would not hurt to check though but as OP mentioned, it happens on al axis.
Do you have access to a spare USB board in the machine to try? Tried different USB cable (shielded)?
What accel / decal settings do you have set now when this happens?
-
Thanks for all the help on this so far guys. much appreciated it
I will do all the tests you recommend next week and feedback. While Im spend the day looking over it, is there anything else you recommend checking/doing/applying etc while Im tinkering
-
Surprisingly nice looking construction. Ferrules on the wires is very good. Drives should have been mounted with heat sinks vertical but my Leadshine drives don't produce much heat. I see a toroidal power supply transformer which is good, but where are the filter caps located? Hard to see how the grounding is done. You should not be concerned about earth grounding. That's a safety issue if not connected but not a control issue. The way the grounding is done inside though is very important. All grounds from every device, circuit boards, drives, motors, etc. must all come back to one ground bus by individual conductors. If any ground wire is connected to multiple devices that is bad. It is hard to tell if any wiring is shielded cable. Cable shields if they exist should only connect at one end, typically at the panel ground bus. However if they are already connected to the case of a device like a motor that is grounded to the frame then they should be insulated from ground at the panel end. If cables for devices like limit switches, signal wires to drives and such are not shielded they should not be bundled with spindle or stepper motor wires but should be separated as far as possible. Last thing. The USB cable may have a shield connected at both ends. The PC has a ground wire to the power source, and the machine has a ground wire to the power source, making a rather non-obvious ground loop that often causes problems.
-
Hi Gary,
Thanks for the reply
The pictures I posted are of all the power and control electronics, so I would say there is no input filtering. Ive read that the design of this machine was based around chinese voltage which is around 220v, while the UK is more around 240v / 250v. So as G21 mentioned, the drives dont like over voltage. Could not having a filter circuit inplace therefore be part of this overall design error?
If so, could I get a filter PCB made up, would it just be a capacitor and resistor setup?
I also do not see any specific single gounding junction, so I will need see if I can get my head around that. When you cable shields, do you mean the flexi steel cable cable protection inside cutting area of the cnc or actual EMI sheilding. If the latter....there is no. And the flexi steel cable protection isnt well done (I.E: it dosnt ground at the ends as it still has the PVC wrap on it inside the connection)
I think the USB cable has a clip on ferrite at each end....is this enough
Would it be worth me buying a bunch of ferrites and putting them on then driver wires etc?
I will check on the PC grounding
-
Hello Rush,
What design error are you talking about? So far we have not yet determined if there is a design error or not.
A filter on the secondary AC power supply (after the transformer) could help reducing spikes and ripple on the AC. You have not answered my earlier question: what is the AC voltage on your drives? You need a multimeter set to AC and measure the power supply lines for your drives with the machine staic and with the machine running.
If you have access to a scope, you could see what how the power supply for the drivers looks like. Also the power supply for the USB card might need to be checked, this appears to have its own power supply on board so a clean AC input is desired although it appears the card has some capacitors on board. Checking this power supply might help also, a scope will tell you if you have a 'dirty' power supply or not.
Ground wiring is quite simple, every component needs to be grounded once and to one common point. Where this common point is or which part of the machine this is (could be the entire frame) is not important but there can only be one connection from the drives to this point. Since the drives do not have a separate ground wire, I assume the frame, or the metal enclosure I see in the picture is the ground and there is one wire from the metal frame to the yellow/green ground wire. All seems to be correct but is worth checking for lose or bad connections.
Shielded wires should have a ground connection only on one side of the wire, quite simple.
I would start with measuring the power supply lines to the drives and the USB card first. See what the AC voltage is and if you have a scope, see if they are clean.
The spindle drive is connected directly to the AC lines at 220V? (measure the voltage to be sure) If so, I would disconnect the spindle drive from the power supply lines and run the test again, this eliminates the spindle driver and spindle motor from the problem.
Let us know what you find.
-
Russ,
I didn't realize your drives were AC powered, I was thinking the trandmsformer was for a DC power supply which would have needed filter caps.
The shielding I was speaking of is cable shields. If your machine has cable carriers to the moving parts and they contain signal wires along with motor cables and neither are shielded that could be a problem.
-
Hi All,
Sorry for the delay in getting some data back.
So I have carried out a series of test, mostly concentrating on the Z axis, but the issue is present on the X and Y as well.
Firstly, I measured the voltage at the main input, this was 248 - 249v
Next, the voltage going to each driver was 68v (There seemed to be no drop under load)
So I stripped out the spindle driver and re-ran the position test. Work Piece Zero went from Zero to -0.7969mm (which I think is what you class as gaining steps??)
Next, I stripped out the motor driver that drives the 4th axis (not used it yet!!), as well as disconnecting all leads that go to the front panel.
Re-Ran the test and Work Piece Zero went from Zero to -1.4781mm (which in machine chords was -217.3669 started and finished at -218.8450)
Final Test was pretty much down to bare minimum. I stripped out all the drivers except the Z axis. Removed all power lead not in use. Disconnected X and Y Limit and Home switch (kept Z connected)
Re Ran Test and Work Piece Zero went from Zero to -0.8206mm
(https://s31.postimg.org/4f5fawrgb/20160721_205518.jpg)
(https://s32.postimg.org/dv1gtosad/20160721_214656.jpg)
Below are just some extra photos I took of the electronics if of any help
(https://s32.postimg.org/9cz61lc8l/20160721_213026.jpg)
(https://s31.postimg.org/6zt2ys6sr/20160721_212245.jpg)
(https://s31.postimg.org/5i0a1fdyj/20160721_214335.jpg)
-
By the way, I do have a oscilloscope.....but embarrassingly, I don't really know how to use it. Never the less, I gave it a go and connected to the power in of the motor drivers (with the neg lead to the neutral), then twiddled a few knobs and pushed a few buttons, and this is the best I could get??
(https://s31.postimg.org/ornzywvln/20160721_222908.jpg)
-
Don't like to keep going on but being a mechanical engineer and I'm sure any problems with the electronics would be a thread going through this design of CNC and highlighted on the web. You never said the make of your machine however I looked on the web and decided it was a Chinese SVM1. looking at the specifications it does not have slide adjustment as you told me earlier in your thread however the specification does state this.
GAS SPRING COUNTERBALANCE DESIGN
To counter balance the machine head to provide smooth Z axis movement and prevent the head from dropping when the machine is powered down, a heavy duty gas spring is installed. This also reduces the work the Z axis motor has to do when lifting the head.
I still think this is the problem with your machine as I recently repaired a similar Chinese machine with a new gas spring as he kept losing steps due to over run as the gas strut was not assisting the slow down and change of direction of the Z axis. The main problem the user was having was every time he put a new component in the depth was a little deeper than the last one. replacing the gas strut solved the problem.
Jim
-
Hi Jim,
This is the SVM-2 model, but has the same gas strut. Its based on (I believe uses the same castings) the old Mikini, but with updated screws and motors, and a complete electronics system overall as the old mikini had many a problem.
I understand your description of the issue you had, but Im gaining steps, not loosing. So rather than the Z ending up deeper and deeper after every run.....Mine is ending up higher and higher. Maybe the gas strut it pushing the head up to much, although that would be surprising.
The other point is this issue is also on the X and Y axis
-
Gas springs are usually a fixed beginning force rating. You must apply that amount of force to move it off the stop, then it increases about 10% over the full stroke. Because it uses a fluid it also acts as a damper keeping quick motions from happening as well. Some are adjustable, they come fully pressurized, but a bleed screw lets you reduce the force. So one that was too strong also might damp motion too much and cause lost steps. I'd lower the head on a bathroom scale and see what it weighs, then push the gas spring against the scale and see how that matches. If it is too far off it may very well be the problem.
I added a gas spring on my machine when the new spindle motor cause the head to drift downwards when powered off.
-
OK russh I've read somewhere that using the sherline 1/2 stepping option and changing the pulse widths on the ports and pins setup page helps control problems with stepping faults on Chinese machines.
I'm out of my depth on this subject but someone on the forum with that knowledge might be able to help, then again I might be barking up the wrong tree.
PS I'm also in the UK, a Brummie
-
Russ,
Since you have a scope, could you hook it up to the step input pulse on one of your stepper drivers? Just select a federate of lets say 1000 and see how the pulses look going into your step driver. Connect the scope to the pulse - and pulse + input on your driver. Take some pictures how the signal looks like, it should be a nice stable square wave. since you are having problems on all 3 axis, it does not matter which axis you select, would be good to check all 3 axis and show us how the pulses look like.
G21
-
Many thanks to all.
I will check out the gas spring as recommended.
Jim - I will do a little reading on your suggestions, also a fellow brummie (work in Great Barr)
G21 - Can I just confirm which wire is the step input pulse. (DIR-, DIR+, PUL-, PUL+, ENA-, ENA+)?
-
Russh,
The step input are the Pulse - and the Pulse +, Dir - and Dir + are the direction inputs which seem to work fine and the ENA - / ENA + are the driver enable signals which also seem to work since you can move your axis around.
Since you are gaining steps, the driver must think there are additional pulses coming from the Pulse inputs to be able to make steps. If there is noise or just a bad Pulse signal coming from your card, this could be the reason for irregular and uncontrolled steps. If the signals look nice and clean, stable without much fluctuation then I start to worry about your USB card in the machine. There are quite a few machines out there by now that use the same configuration as yours and run without problem so there must be something wrong with your hardware from the moment you got yours or the computer / USB cable to control the machine maybe the culprit.
Step by step and we will figure this out.
G21
-
I'm not that experienced with CNC setups but I had a simmilar problem with my Router losing/gaining steps. Turned out to be motor resonance or ringing caused by excessive current drive to the motor. My machine also uses Nema 34 motors and Leadshine HM860H drives, I had to reduce the drive current to less tha half what was set by the manufacturer and now all is perfect. I've no idea if this could be the cause of your problem but could just be worth a try.
All the best, Bob.
-
cncbouk,
You have a good point here, since we don't know anything about the motors the current settings could be wrong. In Russh current set-up according to the pictures posted earlier, the motor current is set to the maximum of 6A per motor and 7.2A peak. I attached a little screen-shot of the dip switch settings per the manual for his drivers.
Definitely worth trying to reduce the current and see what happens.
If all the signals to the drivers are correct and clean and power supply is good, the driver / motor has to be the problem as Russh already rules out mechanical problems if I am correct.
Give it a try and see what happens.
G21
-
Hi All,
Only just picked up the latest two posts so didnt try that out yet, but definately will.
Anyway, Ive spent this evening running a series of tests, checking over the electronics and also some mechnicals
Firstly, I have now stripped back all the electronics, such that I am now left with only the Z axis drive gear, the controller, and the power system. All other axis drivers have been disconnected and removed, the front panel connections have been removed etc.
So I thought I would start with a simply test, and that was to change the USB lead. I went and got a decent shielded USB lead and double checked the contiuity of the outer sheilding to the connector housing. Unforunately, this changed nothing.
I then went over all the remainging electrics, checking for any loose connections and tighting everything up to ensure all was so. I again ran a Z axis test and the same results.
Next, I took a look at the mechanical couplings from the motor to the ball screw. Ive checked these previously, but wanted to be doubly sure. I marke up a line between all the connections. I re-ran a test and the same result. Having checked the mechanical connection, the line was still intact along all connection points. So This rules out any slipage in the couplings.
So, out of ideas, I thought I would play around with the acc. and Vel. settings in mach 3. I started very very low and slowly worked back up. However, the test program I ran only moved the Z axis through a distance 70mm (reoccuring) so Im wondering if at the very low acceleration range if the motor every got up to the velocity?
Anyway, these are the results of those tests
Original Settings - Velocity: 6000, Acceleration: 300
Velocity: 500, Acceleration 20 - Good Result, Held Zero
Velocity: 2000, Acceleration 20 - Good Result, Held Zero
Velocity: 4000, Acceleration 50 - Bad Result, Lost Zero
Velocity: 4000, Acceleration 20 - Good Result, Held Zero
So it seems that if I slow the acceleration down to painfully slow, it stabilises. Im not sure this is a solution though, more an observation, as I know there are the same machines out there running fine at the orignal setting, but maybe this test can help to highlight some possible issue.
And Finally, an observation that I feel I need to mention, and Im very much wondering if this could lead us on right path to finding the issue. While I was messing with the Velocity and Acceleration settings, I forgot to compensate for this while jogging the Z axis downward. As such, I ended up crashing in to the workpiece. What I observed (which I assume is a common action) is that the motor and ball screw continued to rotate, and instead the slipping must have occured at what I think is called the Ball Nut.
What happened was that because the motor was still turning, Mach 3 thought the z axis was still travelling downward. Therefore, when retracted, the Z zero point had changed and was above the workpiece (same results).
Could this be a possible reason to whats happening durning a program run, could the Ball Nut be slipping?
The only thing I cant answer is how this would be happening on the X and Y axis as well, but concentraiting on the Z for now.
BVy the way, this is the signal from Pulse + pin
(https://s32.postimg.org/ep79tfx3p/24_EE3986615164_EDE9496_ECB91_FD9_CBD30_F4_E125618_E1_EA.jpg)
-
No way should the screw have continued rotating unless the ball nut managed to turn as well. So that could be the source of your problem. If the ball nut turns at all it will appear as lost steps. Check the ball nut mounting method.
-
The ballscrew should not continue to rotate after your head crashed and stopped. Are you sure the screw kept on turning or did you just look at the screen at Mach 3? Mach 3 will continue to run and the DRO's think the machine is still moving although the actual motor is stopped, blocked or whatever. With the head locked or blocked, the ball screw cannot turn much, maybe a few degrees due to the screw bending and the nut pressing the balls into one side of the grove.
Secondly, the signal you show on the pulse + pin does not make sense to me, did you hook up the scope correctly? Pulse + to scope A in and scope ground to pulse -. you should see a nice square wave and not what you see right now.
If you want to verify the motor turning with the head blocked, why don't you purposely block the head by moving it down to the (scrap) workpiece and see if you can slowly lower it more by jogging the Z down slowly! and watch if the motor is actually turning with the screw turning also? Since you already crashed it and saw something funny, this would confirm it. I think the couplers would slip or the motor would stall out but the screw wont turn more then a few degrees.
G21.
-
"Pulse + to scope A in and scope ground to pulse -"
This is not recommended as it could damage the line driver on the Jamen board, rather connect both channel 1 and 2 ground to ground and channel 1 to pulse plus and channel 2 to pulse -.
This will show the full differential output.
This link has some explanation on measuring differential signals.
http://electronics.stackexchange.com/questions/76514/how-do-i-measure-differential-signals-like-rs-485-or-dmx-on-an-oscilloscope
Kevin
-
Hello KevinRSA
You are correct, I not realize this was a differential input although I downloaded his driver manual. My mistake, let's see what RussH finds out.
Lowering the current is a quick and easy thing to try I think. Checking if the ballnuts are not loose in his machine is also quite easy.
Thanks for straightening this out and correcting my mistake.
G21
-
Ive taken a video of the ball screw action, let me know what you think.
Im just working on taking the oscilloscope readings
https://youtu.be/q4UkBnZKrwg (https://youtu.be/q4UkBnZKrwg)
-
Ive taken a video of the ball screw action, let me know what you think.
Im just working on taking the oscilloscope readings
https://youtu.be/q4UkBnZKrwg
-
You can see the ballscrew shaft moving up and down! You have a problem with the bearings at the end(s). The shaft should not move at all and even be preloaded against a stop so that it has no backlash at all.
-
I viewed that as play in the spindle?
If you look early on, you can see the spindle nose moving in relation to the spindle body, where its wedged on the wood???
-
I think what your seeing is the spindle sinking slightly in to the wood under the force. I can assure this is not play in the spindle. The ball screw slips and the head bounces back up
.
-
Can we just confirm a couple of things, your early threads said the loosing of steps is on all three axis. If this is still the problem it cannot be mechanical it must be electronic. As I said early on in this thread my problem with my machine was with the weight of the z axis motor and slides overriding the motor trying to stop it falling and the failure of the damper to help it back up. if you say slowing the slides down reduces the errors I still think its the acceleration and deceleration parameters on your machine or the pulse settings. Surely if its on all three axis they cant all have a mechanical fault even if its Chinese built. Pursue the electronic route and settings. Jim
-
Hi Russ, I can't see from the video that there is any problem there(especially not one that would create a 1.4mm error). I would go with Jim's advise and pursue the electronic suspects.
Like I said in my mail to you, I would expect to see a reasonably clean square wave on the pulse output, Once we have established that the pulse output is indeed good I would be keen to see if reducing the steps/rev has any effect.
Kevin
This is the Pulse+ wave form from a Jamen controller
(https://s19.postimg.org/tne022ycj/IMG_3410.jpg)
-
Bit of a delayed update, work got in the way of me running more tests.
Right I stopped stripping down the machanics and reverted back to the electronics as recommended. To confirm, this issue was noticed on all 3 axis, with the Z axis being the mostr prominent.
As you have probably seen from previous posts, I have stripped back the machines electronics right down to the requirements of only the Z Axis.
As such, I have testing the Pulse + signal as per Kevins photos. Below is the results of this test. The first image is the overall signal. The second image is a zoom in on the start of the signal, wasnt sure if that was class as undesirable noise and if it would cause any issues. The third image is just the setting of the osiciscope in case I had set anything wrong.
(https://s9.postimg.org/h3ayy38bz/20160810_100733.jpg)
(https://s10.postimg.org/95ul4d5zd/20160810_100819.jpg)
(https://s10.postimg.org/clnfiljl5/20160810_100703.jpg)
I also ran through all the voltage pre controller again, the following is what I recorded.
(https://s9.postimg.org/m4vnkhunz/20160810_181002.jpg)
-
Wow! 68V to your stepper motor controller? That seems a bit much. Most drivers operate in the range of about 24 - 48V. Using too high a voltage can create issues too. From looking at earlier pictures you posted, it looks like we have similar stepper controllers so you definitely need to look at that. Additionally, coming off of a transformer you are getting AC voltage. There should be some sort of rectifier at least and regulators, etc. involved as well. Recommended is to use a full DC power supply the puts out the required 24 or 48V DC.
Stephen "Highspeed" Kruse
P.S. I know I'm a late comer to the discussion and only briefly perused through the thread so forgive me if I've missed something or repeating information already supplied.
EDIT: I just saw the post that your stepper drivers are AC power rated, so this my comments may be totally irrelevant.
-
Thanks the the input mate, every comment helps trust me.
My drivers are rated between 36 - 8- VAC (although it doesnt mention a DC voltrage range of 50 - 110 VDC, so not sure if it operates in AC and DC)
http://www.leadshine.com/UploadFile/Down/MA860Hm.pdf
-
Ok, yes. Mine are the M542 and utilize only DC at up to 50V. Yours are rated quite substantially and can use either AC or DC looking at that document. So that eliminates this as a potential source of the issue. Just something I saw and am glad you had the documentation for it.
Stephen "Highspeed" Kruse
-
Hello RussH,
Your drivers are fine with this voltage, they are rated from 18 - 80VAC or 24V - 110VDC. Your mains voltage is a little higher then usual so your driver voltage is also a little higher but well within specs for your drivers. Pulses on the scope look fine, no worries about the ringing at the beginning of the pulse.
Try to reduce the current through the motors by changing the dip switch settings I posted earlier. Try with SW1: OFF, SW2: OFF and SW3: ON. This should reduce your current to 3.7A and peak current 4.45A.
Keep your acceleration low to start with and see what happens. If this works fine, increase your acceleration and speed back to the normal settings step by step until you feel OK with it.
G21
-
Thanks, I will give it a go.
Thought I would mention, I found a label on one of the steppers yesterday. They have a part number of 86bygh450b-06d-15j
Not sure it this helps at all.
-
Copied and pasted your motor numbers in ebay and came across another CNC'er loosing steps with these motors it appears his problem was mechanical, have a look at his thread although he doesn't say he has solved the problem
http://www.cnczone.com/forums/mach-software-artsoft-software-/154503-z-axis-losing-steeps-2.html
-
Hi Jim,
I took the motor and coupling off the ball screw today, and ran the spindle head up and down its slides 3 times by manually turning the ball screw. It was fairly easy to turn (could do so with just a finger grip), and didnt feel any sudden change in resistance during its travels.
-
Any update here? Been quite a long time now.
Curious if you figured it out.
G21