Encoder Error

[Greg82]

My oak was replaced and running the same tests on the new oak showed all 4 axis ports working perfectly. I was told after that they didn't find a problem with the oak I shipped to them

[thelinuxman]

Has there been any resolution to this other than a new board for Greg82?
I am setting up a new OAK controller with Yaskawa Sigma 1 drives and having the same issue. Whichever axis is plugged into #1 is fine, other axis will not go to commanded position & accumulate error on PID screen with every move.

[tblough]

Have you verified that the encoders are counting in the correct direction for the other axis? You may need to reverse the encode phasing on those axis.

[eng199]

Has there been any resolution to this other than a new board for Greg82?

Yes, we have discovered the cause of the problem.
Oak with IC28 labeled LC4D v0x0031 have an issue.

Position request outputs 2-4 may have unexpected short times between A/B transition. This will cause missed counts to servo drives with a slower input rate, such as Yaskawa Sigma II (200kpps) and DMM DYN4 (500kpps). Drives with faster input rates like Yaskawa Sigma IV (1000kpps) and Automation Direct SureServo2 (4000kpps) did not lose position during testing.

Velocity mode (analog) output is not affected.

The solution is to update IC28 to LC4D v0x0032. Reprogramming requires an RMA number to return to Centroid.

[cncsnw]

Approximately what was the range of production dates for Oak units with v0x0031 firmware?

[occustoms]

Yup i have this problem too. And i have the that ic28 label. Looks like i gotta call and get rma. Hopefully its a quick turnaround

[eng199]

Approximately what was the range of production dates for Oak units with v0x0031 firmware?

sn 1378 9-16-22 to sn 1446 3-6-23, some quantity less than 67 units
Some of the boards in that range did not ship and are now updated.

Yup i have this problem too. And i have the that ic28 label. Looks like i gotta call and get rma. Hopefully its a quick turnaround

The turnaround should be quick since the solution is known. If you email a report to support@centroidcnc.com, we should have all the information to generate the RMA and verify that the code needs an update.

[aamir]

I also have the same problem but i switch the Drive from Normal Pulse train to Hi speed pulse input in one of the chinese Dorna Drive that resolve my issue .

[m5white]

Centroid updated my firmware and that fixed the issue I was Having with Y, Z and A position errors. It was interesting that after I installed OAK the Y axis still had count errors but not enough to fault as a position error. I decided to redo some of the setup procedures by retuning the drives/motors then rerunning the PID Tune and that fixed that. The machine is now running well and the PID Error is now less than about 20-30 counts (this equates to around .0001 axis movement). I have tested all axis for repeatability with a .0001 indicator and they are showing no discernable error. When I move over about 100 IPM to 200 IPM I start to see a change of about .0001 to .00015 which I attribute to indicator bounce. If the indicator reading changes by .00015 at end of travel at 200IPM it still repeats to .0000 on my dial. Very impressive.

Thank you for solving and fixing this problem Centroid! This is why I now have retrofitted my lathe and mill (that had the position error problem) with OAK and just also finished the retrofit of my second mill with ACORN. Wow, was the ACORN retrofit easy.

[cnckeith]

Centroid updated my firmware and that fixed the issue I was Having with Y, Z and A position errors. It was interesting that after I installed OAK the Y axis still had count errors but not enough to fault as a position error. I decided to redo some of the setup procedures by retuning the drives/motors then rerunning the PID Tune and that fixed that. The machine is now running well and the PID Error is now less than about 20-30 counts (this equates to around .0001 axis movement). I have tested all axis for repeatability with a .0001 indicator and they are showing no discernable error. When I move over about 100 IPM to 200 IPM I start to see a change of about .0001 to .00015 which I attribute to indicator bounce. If the indicator reading changes by .00015 at end of travel at 200IPM it still repeats to .0000 on my dial. Very impressive.

Thank you for solving and fixing this problem Centroid! This is why I now have retrofitted my lathe and mill (that had the position error problem) with OAK and just also finished the retrofit of my second mill with ACORN. Wow, was the ACORN retrofit easy.

thanks for taking the time to report back.
good deal on the Acorn Retrofit.
if you can, photos (online album) of customer machine projects are always fun to see