Looking for some advice on an intermittent x axis fault

[Norvic]

Hi I'm a furniture maker from Norfolk and long time lurker. I've been using CNC routers for about two years now.

I've got a problem with my x axis losing position or possibly binding I've had trouble actually pinning down what's causing the issue. It can happen after 10 minutes or 8 hours which has made troubleshooting difficult.

The x axis randomly loses position in both the positive or negative direction, usually by less than the distance it was suppose to move during that piece of g code. I've had no issue with the Y or Z axis and it can happen when moving just on the x axis or when multiple axis's are moving at the same time

It hasn't happened during jogging. How often it happens and by how much it loses position seems to be independent of feed rate, acceleration or step and direction pulse length settings in mach3(which is a genuine license). I've tried running it with the spindle, extraction, water pump and everything else in the workshop off and it still happens. The frame is grounded on each axis and and the spindle cable is grounded at both ends. The few times I've been there to actually see it happen, the motor makes a high pitch noise then carries on.

I've tried swapping over the stepper motor, driver and all the wiring between the stepper motor, driver and BOB. As well as using different pins on the breakout board. After swapping over the wiring the issue seems to happen a lot less often for a time then goes back to being random. I've marked the coupling where it attaches to the ball screw and motor to check for slipping and checked the balscrew for any damage.. A fresh install of mach3 and changing the kernel speed also didn't help. I've tried turning off micro stepping which is set at 10 for the other axises.

I set up a bit of test gcode that move it 120mm left to right for an hour. I tried this with different direction pulse lengths which caused the x position loss being 120mm repeatably in either direction but after restarting the system it seems to go back to being random.

The machine is one I picked up off eBay, I was planning on building one myself but this was pretty similar to what I had planned and cheaper than I could buy the components. The bed is about 700mm x 800mm. Uses uc400eth CP0-10v BOB, 4.5Nm steppers and CWD556 drivers from CNC4You and Hiwin HGR25 rails and 5mm pitch ball screws. it worked perfectly for the first few months, cuts within the 0.02mm accuracy of my veniers on aluminium although it's mainly used for wood. The seller has been very helpful. He was only selling because his shed was built on a slope that its slowly falling down.

I'd like to swap to axbbe and UCCNC but I don't want to spend any more money until I've identified the problem. My next step was going to be shielding the stepper motor wire but I seem to be running out of things to try so I would love some advice on what I can try next.

Thanks

Sam

[JohnHaine]

.....The few times I've been there to actually see it happen, the motor makes a high pitch noise then carries on.

....
Sam

Hello Sam, how annoying this sort of thing can be! My mill X power feed is done with a stepper motor and I get the high pitched squeal when and if it stalls. There is a tight spot in the table towards one end and if I try to drive the table too fast that will cause this. It sounds as though you have tried different feeds so may not be that.

One other thing that can happen is that a stepper can have a fragment of swarf in it which gets caught in the "teeth" and jams it. This may go away but then come back in a different place as the bit gets "circulated". I've had this once on a new NEMA 8 motor from a far east supplier. Take the stepper off and disconnect the wires, then try spinning it with your fingers. You may need to try for quite a while but if you suddenly feel the motor jam you've found it. There's no cure really, better to scrap the motor and replace.

[Norvic]

Hello Sam, how annoying this sort of thing can be! My mill X power feed is done with a stepper motor and I get the high pitched squeal when and if it stalls. There is a tight spot in the table towards one end and if I try to drive the table too fast that will cause this. It sounds as though you have tried different feeds so may not be that.

One other thing that can happen is that a stepper can have a fragment of swarf in it which gets caught in the "teeth" and jams it. This may go away but then come back in a different place as the bit gets "circulated". I've had this once on a new NEMA 8 motor from a far east supplier. Take the stepper off and disconnect the wires, then try spinning it with your fingers. You may need to try for quite a while but if you suddenly feel the motor jam you've found it. There's no cure really, better to scrap the motor and replace.

Hi John, yeah it is a little frustrating. Thanks for your suggestions. I've tried swapping the stepper with one on the y axis already which didn't work so I think I can rule out the motor itself. I can turn the ball screw all the way fairly easily and I can't see any damage to the screw but considering how random it is maybe something worked its way into the ball nut which is causing it to stall. I also noticed this morning I can hear the ball bearings circulating which I can't on the other axis so it might be the next thing to investigate.

[mekanik]

Hi Sam
I have no experience of operating a CNC device, but in the early days i followed all of the build logs on the forum and it was recommended that screened cable only be grounded at one location only(usually your electronics cabinet) spindle especially.

The frame is grounded on each axis and and the spindle cable is grounded at both ends.
Regards
Mike

[Norvic]

Hi Mike
Yeah from what I've read it seems like quite a contentious issue whether you ground on one or both end but since the position loss happens even with the spindle and vfd off it seems unlikely to be causing the problem. Thanks for the suggestion though

[ZASto]

Hi John, yeah it is a little frustrating. Thanks for your suggestions. I've tried swapping the stepper with one on the y axis already which didn't work so I think I can rule out the motor itself. I can turn the ball screw all the way fairly easily and I can't see any damage to the screw but considering how random it is maybe something worked its way into the ball nut which is causing it to stall. I also noticed this morning I can hear the ball bearings circulating which I can't on the other axis so it might be the next thing to investigate.

It seems that your ball nut is dry. Try to lubricate it.

[Norvic]

I tried lubricating it directly and a new ballnut and I'm still getting position loss. Although the new ballnut sounds completely normal so it seems unlikely to be a mechanical issue unless I'm missing something else.

[JohnHaine]

When you swapped the motor did you try the old x motor on the y axis? If there is a bit of crud it might not show up for quite a long time.

[Norvic]

Yeah I swapped them, I've had no issue with the old x axis stepper on the y axis so pretty sure the motor isn't causing the issue

[m_c]

Have you tried dropping the ballnut/carrier of the x-axis, and running the gantry back and forth by hand?

It could be dirt has built up in the linear rail trucks, and causing them to bind occasionally.

[Norvic]

I've tried that twice on separate occasions and as far as I can tell it moves freely.

Would you expect the binding to happen more often at higher feed rates and acceleration if that was the case?

[m_c]

If it was a bit dirt occasionally jamming the bearing tracks, then it could happen completely randomly, just depending on when the bit dirt jammed a bearing.
Have you tried greasing the trucks?

I've not seen any pics of the machine, but you could try twisting the gantry in various directions as you push it, to see if you can make it bind in anyway.

[Norvic]

The bearing blocks were greased last week when I did the ballnut without any luck.

I'll upload some pictures of it tomorrow but the x axis is a piece of 120 x 60 mm aluminium extrusion with 8 mm walls.
The rails are positioned on the top and bottom with the ballscrew on the back with 20mm aluminium plates forming a box around around which seemed like an unusual design so I did try to see if it would twist when I first got it. I couldn't get it to twist, rack or alter shape at all. I've also made sure everything is square.

I'll try it again tomorrow just in case and also check each block when they're disconnected from everything else.

[Norvic]

Hi Thanks for your help.
I checked the linear rails and the wiring again thoroughly but still couldn't find anything out of the ordinary.
I got my head around how to set up UCCNC with my breakout board and I've had it running all week without an issue, the longest it had gone without position loss before was about 8 hours. So I'm still not exactly sure what was causing the problem but at least at the moment it seems to be fixed.