position error in multiple axes

[bthoron]

We've been wrestling with a difficult problem for several months, The X and Y axes (and possibly the Z) are losing position. We can see this when the machine jogs home after a partfile, it comes home to a slightly different location than it left. Over time, this error increases from a few thousandths, to -3/16" in the x and +1/8" in the Y

We have spent a fair amount of time with Tom and Dave troubleshooting this issue (whose help I've truly appreciated), but we're stuck.

PRS 48x96 (2007/8 version) standard.

-The dust collector is off.
-The spindle can be on or off, unplugged or not.
-Vacuum on or off.
-We have used 3 computers to replicate the problem, with everything from windows xp to the most recent update of 10.
- We have used the recommended settings from SB.
-The USB cable is short, doesn't run parallel with power cables (and we've installed a new one).
-We switched power supply circuits from one leg to another, from one panel to another. Used an isolation transformer on the power circuit. Both the computer and Drive box are on the same circuit.
- The pinions are tight to the tracks and do not appear worn.
- The symptoms are worse doing parts with lots of curves, simply cutting rectangles with the shopbot controller doesn't seem to have a perceptible problem.
- the problem occurs air cutting and cutting materials.


I can't think of a logical place that noise would interfere with the communication cables or stepper cables that we haven't looked at.


The attached image shows the resulting error after an hour of running the part file.
http://www.talkshopbot.com/forum/attachment.php?attachmentid=29261&stc=1

Thanks for looking at this, I hope that somebody has had this happen and worked through it and can share the solution.

Ben Thoron
The Old Globe

[bthoron]

BUMP! I'm revisiting this problem-
It still exists all all axes- the same as before.

Thanks

Ben

[dmidkiff]

I'm not qualified to give you much help, but since the tech team hasn't been able to solve the problem, I will give it a stab. If the amount you are off is the same each time, I would look at tweaking the unit value. Maybe some one else will chime in to verify or shoot me down. Run it by tech support first as I am just throwing something out.

[robtown]

I'm not qualified to give you much help, but since the tech team hasn't been able to solve the problem, I will give it a stab. If the amount you are off is the same each time, I would look at tweaking the unit value. Maybe some one else will chime in to verify or shoot me down. Run it by tech support first as I am just throwing something out.
When you run those hour long files... do the parts come out correctly?

[srwtlc]

Not so sure unit values would be the issue, as it should return to the same 0,0, but sizes would be wrong.

What kind of hour long files are you running, 3D/v-carve? Are your ramp settings set to default or are you using some more aggressive settings? With a standard, you can lose steps if direction changes are too abrupt. If you 'soften' the ramp settings, you can alleviate that. This is mostly noticeable with 3D files, but could show up on v-carving or a lot of short quick raster movements.

If your communications were bad, you'd be seeing other errors, but it would be a good check to see what your packet et values are. Read up on the diagnostics for v3.8.XX under help/notes....

If you drop the motors from the rack, does the gantry/y car roll freely all over, no binding in the y car from side to side. When putting the motors back in place, just firm hand pressure is sufficient, no clamping force of the motors to the rack is needed.

[dmidkiff]

Scott is the guy that can help you sort this out. He's been helping folks on the forum for a long time. I'm glad that he and Rob caught my error.

[bthoron]

Thanks guys:
THe cars run smoothly.
values are the same as we've run in the past.
The files are typical profile cuts, inside and out.
I'll look at the ramp values, just in case- the moves are pretty short. 2" and less.
I'll double check the packet info- didn't seem low the last time we checked.

Ben

[knight_toolworks]

did you check to make sure the pinions are tight on the shafts? though usually that would not cause lost position so much as a sloppy cut. but I found one evenalter I had checked several times. when Check when the machine was cold I found one. worn pinions cause sloppy cuts usually.

[robtown]

Thanks guys:
THe cars run smoothly.
values are the same as we've run in the past.
The files are typical profile cuts, inside and out.
I'll look at the ramp values, just in case- the moves are pretty short. 2" and less.
I'll double check the packet info- didn't seem low the last time we checked.

Ben
If your problem was that the x and y position changed cumulatively during long cuts, I would expect your profile cuts, especially the ones with inside cuts and outside cuts near the end of a long cut, to produce bad parts.