Homing fails 2.9.0 pre0-5518

Are you building from source or using buildbot debs?

This would be a good point to run git-bisect, if things definitely worked better with a previous version of 2.9.

I use builbot deb.
I was already planning to force older versions to see which version works.

joints are setup correctly and if you jog them they work as they should. Can see it in youtube video.

I'm not saying that your joints are set up wrong. Sometimes it just helps to find where the problem may be by adjusting the settings.

Aciera: good point, I did paranoid check.... just to be sure

Found out that FF1 does not make any affect anymore,
when this change has happened?

The complete history for the pid component is here:

master: github.com/LinuxCNC/linuxcnc/commits/mas...hal/components/pid.c
2.8: github.com/LinuxCNC/linuxcnc/commits/2.8...hal/components/pid.c

don't see any recent changes that would affect FF1 usage.

The only commmit that is in master and not in current 2.8 is:
bab0790f26 2021-11-11 07:41:40 -0500 Fix typos in src/hal0

Posting a tar or zip file of the files in your configuration might
allow someone to spot what is different

I found this
forum.linuxcnc.org/10-advanced-configura...2-9-pid-seems-broken

I think i have exact same problem.

I noticed that I have too much following error in my joints.
So I verified that FF1 does not make any change in my setup.

Aciera: Thanks for for idea for the paranoid check.

Problem solved.
"2.8+ and 2.9 have a bug fix in the PID component."

commenting out my "net n-vel-cmd => pid.n.command-deriv" got PID loops working again.

PID and error limits where so in my setup that jogging joint stayed under error limit, but immediately when homing started tighter folowwing error is used and joint gives following error.