Following Error When Homing

I am having trougble with getting a following error allarm when homing the machine for the first time after opening Axis. If I open Axis and click on home all, my 1st 2 axis home fine (w and z) then the 2nd two start to home (x and y). Almost always x (joint 0) will give a following error alarm and stop the homing and somtimes y will as well. This happens just before the axis reaches the home/limit switch. I can then repress the home button and all axis will home and run fine and not give any errors, I can rehome as often as I like without a problem. But if I close and reopen Axis, I get a following error the first time I try to home every time. I am using analog torque mode servos, homing to index after touching the limit switch.

Any ideas what the source of the problem is and why it only happens the first time I try to home after start up?

Yes, one possibility is that you dont have index enable wired to the PID component
so when the encoder count is cleared you get a big 1servo thread thump in the D term

My config was done with PNC config. Is there someplace I could find a good example hal file to compare this with?

I dont know if the examples use this (as it only affects homing via index) but it is documented in the PID component manual:

pid.N.index-enable bit in
On the falling edge of index-enable, pid does not update the
internal command derivative estimate. On systems which use the
encoder index pulse, this pin should be connected to the index-
enable signal. When this is not done, and FF1 is nonzero, a
step change in the input command causes a single-cycle spike in
the PID output. On systems which use exactly one of the -deriv
inputs, this affects the D term as well.

I don't think this is my problem as I do have a line in my hal file that is:
net x-index-enable <=> pid.x.index-enable

You will probably have to halscope the index_enable, homeswitch, and following error to track this down

Has this situation ever been resolved? I have more or less the same problem. Homing works for the first axis (Z), but X and Y generate following error at first try. Subsequent homing attempts always work.

this is in the .hal file:

I have the same problem we discussed it here:

forum.linuxcnc.org/27-driver-boards/3562...e-homing-only#121831

I did some tests but in the end I gave up because the system after homing is perfect.

Can you post your hal file?

No, I never was able to resolve this problem for this machine.

The problem for me is that I chose an unusual configuration where I am communicating with the Mesa cards and running the PID loops in a fast floating point base-thread, rather than running them in the servo thread. This is defeating the mechanisms built into Linuxcnc that allow for the encoder count to be reset to zero without setting a following error when homing to index.

If I could run the servo-thread at the speed of my floating point base-thread, and do away with the base-thread, the problem would go away. But on this machine because of servo-tuning issues I want to run the PIDs as fast as possible, and this configuration lets me squeak out a few more kHz without receiving real time alarms on this PC. So I'm just living with having to home the machine twice after restarts of Linuxcnc.

If you are having a problem with following errors when homing to index caused by the encoder count being reset, on an ordinary config. The problem is usually a problem with the order components are addf-ed to the servo-thread. But there are numerous other reasons you may be receiving a following error, that may be unrelated.