Looking at the parameters we see the following:
The drive is still getting the last valid position command with 200,6 mm, but already gets a new position command in the next cycle with 15 mm. This seems like it is the end position after homing but the homing seems not to be really finished. Getting from 200,6 mm to 15 mm in one NC cycle of 2 ms would need a velocity of 5 568 m/min which is out of the limits --> Therefor error F2037 shows up.
Please create an oscilloscope reading in the drive using the signals with time base 2 ms:
- actual position (S-0-0051)
- actual speed (S-0-0040)
- command position from NC (P-0-0047)
- diagnostic message number (S-0-0390)
and trigger on the error F2037 with Pre-Trigger set on 80%.
I reported this issue a while back. The issue was confirmed and was to be fixed in RM21.11 at least for ctrlX DRIVE. I did request it to be fixed for IndraDrive too, but don't know what if anything was done for either drive.
I believe the issue is due to the C0600 Command taking over the Position Control and once finished it turns it back over ot the ctrlX CORE. So the CORE last had the prior Position Command value from before the C0600 Command and when it gets control back after C0600 completes it wants to instantly repopulate the Position Command with the old position hence the big jump which causes the error.
I did some tests with 1.20 and IndraDrive. Drive controlled homing works fine when using MotionInterface SetupMode.
SetupMode is using ML_AxsHoming internally. So assume it would also work when you use ML_AxsHoming directly.
I had a look to the scope.
You see that right before the C0600 is finished the control is switching the command value (P47) from a value close to the actual position to -260mm. This jump in the commanded position is the reason for the F2037.
If you are able to do an other oszi plot, then please add the status information S135 and P115. Then we have the possibility to get some more drive status information
There is an E4071 also present, that indicates an problem with EtherCAT communication (parameter channel). For further diagnostic the parameterfile is need to get the detail diagnostiv for this warning.
Did you try to use the command ML_AxsSetIpoPosFromActPos (see online documentation) before deactivating C0600? That will use the actual position of the drive and write it into the interpolated position of the motion.
Did some test for other customers yesterday that did following:
Thanks @CodeShepherd but this problem occurs while the homing is active for me. It is a fault during the homing so I am not able to call the function in the meantime because the homing does not complete.