401 Position Error Yaskawa

[mattm]

I keep getting 401 position errors on Z axis machine which I'm setting up. I've followed install procedure (inc. technical bulletin TB 295) and so far the X & Y seem to be moving pretty well. However the Z axis quickly throws a position error fault whilst trying to jog, it can manage small increments a bit better, but as soon as increments are increased or continuous mode is selected it faults 401. At first I thought this could be because the motor hadn't yet been tuned, but it runs well when jogging using SigmaWin+. I also checked the encoder count was going up when manually turned CCW.

I removed the motor for a bench test and still getting the same fault. I'm now wondering if this maybe something to do with the encoder resolution? The Z axis is 13bit encoder whilst the X&Y drives are 17bit. I noticed that the encoder counts are all 32768 - which seemed possibly too high for a 13bit encoder...? Again I tried 8192 instead but still the same problem. I have attached printout of Servo drive parameters.

There seems to be a significant position error accumulated in CNC12 (see pic) - after a cnc12 restart the max/min error builds whilst jogging until the first time the 401 position error occurs, then doesn't seem to change after I reset and start jogging again, though the 401 error keeps reoccurring. I monitored Deviation Counter in SigmaWin+ and saw it jump as high as 16 on occasion when jogging with CNC12, the higher the increment the higher the deviation it seemed. However, it doesn't move from 0 at all when jogging with SigmaWin+ does this signify a problem with cnc12 configuration?

Any advice would be greatly appreciated!

[martyscncgarage]

Matt, could you please provide more information: viewtopic.php?f=64&t=383
The more infomation you provide, the better we can help out. What kind of machine? All we can do is assume its a mill? Counter balanced head???
Remember, we can not "SEE" what you "SEE"

Marty

[mattm]

Hi Marty, apologies it's a CNC router and the Z is the vertical axis. I was hoping I'd isolated the problem from the machine itself by disconnecting the motor mechanically - it would normally be driving a 5mm ballscrew. The problem still exists when there's no load on the motor. Here's a picture of the cabinet before Centroid installation started and also of the Y axis carriage. I'll get some better photos uploaded soon. Many thanks, Matt

[martyscncgarage]

OK, since it only exists on the Z axis, have you tried connecting the X or Y motor to the Z drive and the Z motor to the X or Y drive and retest to see if the problem stays with Z or follows....

[cncsnw]

The maximum counts/rev for the 13 bit encoder is 8192.

You probably need to set Pn201 = 2048. I am not sure if there are any other changes you would need to make in the drive to make the position reference match.

In the Centroid Machine Configuration -> Motor Parameters you would set counts/rev for that axis to 8192.

[mattm]

I set changed the Pn201 to 2048 and counts/rev 8192 but still same issue. However I hadn't noticed yesterday (as I had the SigmaWin+ window over it ) that the Error column in the PID menu consistently builds to about 8180 at which point the 410 error occurs. Is it just a coincidence that it is almost 8192??!!

See video at link below that shows this - you can see how I drop down the jog multiplier as its getting closer to 8180 the fault occurs and I repeat once more. Also, should 8192 represent one full revolution of the servo? - the servo actually moves about one and a third of a revolution. Does this shed any light on what the problem could be?

https://photos.app.goo.gl/8FkkpRvgvfjYU8Sf9

[cncsnw]

In your video, it appears that you have selected Incremental jog, x100. That should cause the motor to move about 0.25 mm each time you press the jog key. You have your turns ratio set to 8mm/rev, so 0.25mm should be about 1/32 of a motor turn.

Are you telling us that just a few presses of the key causes the motor to turn more than one full revolution?

With servo power off (i.e. with emergency stop pressed), and with Pn201 set to 2048, how many encoder counts do you see in the Abs Pos column for one turn of the servo motor shaft?

Turn the shaft back to zero counts in the Abs Pos column, then release emergency stop. Select Incremental jog mode, x100. Jog one increment Z- (which should cause the motor to move and count plus, since you have direction reversal set).

What is the Abs Pos value after that one increment? What is the error value? Approximately how far did the motor shaft actually turn?

[mattm]

With servo power off and Pn201 set to 2048, there are roughly 8264 counts in Abs Pos column.

Moved motor shaft back to zero and did one Z- increment on x100 and the Abs Pos column is at 1040. The error value is -781. One increment moves the motor approx 45 degrees.

I also tried to catch all this on video - I'll send the links shortly once they've uploaded.

[cncsnw]

Also see what happens when you jog by x10 and x1 increments. Does it move under control, by proportionally smaller amounts?

If, before you get to the "410 ... position error" threshold, you turn around and jog the other way, does it move under control back to where you started?

[mattm]

It seems to move fine incrementally; ten x10 were equal to one x100 increment move of the motor shaft and ten x1 increments were equal to one x10 move.

I jogged it up close to the 410 position error theshold and back down again fine - and continued in the opposite direction until hitting the threshold which seems to be mirrored in both directions. Again I've taken videos of both theses tests (video 3&4), 1 and 2 are the first ones.

1. Abs Pos after one full turn of the servo with the servo power off:
https://photos.app.goo.gl/aLVzDP5Q28qxcdhQ6

2. Turn motor back to zero then Abs Pos value after that one increment x100:
https://photos.app.goo.gl/JK57xGSTxphqPtjX9

3. Moving in smaller increments:
https://photos.app.goo.gl/XDKJ1Vn2HGP4T1CYA

4. Jogging upto the 410 error threshold:
https://photos.app.goo.gl/TLCpG1wTLAdoiAaL8

Many thanks for all the help - apologies it was rather early in the morning when I did the first couple of videos!