Mill atc configuration

Ah, yes, The problem here is that 'spindle.0.orient' goes through that convoluted logic I mentioned above. So change this:

# M19 trigger orient.enable; orient-pid enable; spinena; select analogout
net spindle-orient <= spindle.0.orient
net spindle-orient => or2.1.in0 #to orient enable
net spindle-orient => and2.0.in0 #to spindle is oriented

net orient-lock <= spindle.0.locked
net orient-lock => or2.1.in1 #to orient ennable

net orient-enable => or2.1.out
net orient-enable => orient.enable
net orient-enable => orient-pid.enable
net orient-enable => or2.0.in1
net orient-enable => mux2.0.sel


To this:


# M19 trigger orient.enable; orient-pid enable; spinena; select analogout
net orient-enable <= spindle.0.orient
net orient-enable => orient.enable
net orient-enable => orient-pid.enable
net orient-enable => or2.0.in1
net orient-enable => mux2.0.sel

I'm not sure if you stil have these lines but remove them if you do :

net orient-complete => and2.0.in1
net spindle-is-oriented <= and2.0.out


Also remove the components 'or2.1' and 'and2.0' from 'loadrt/addf'

Actually you'll probably want to leave 'or2.1' in the 'loadrt' because of the component numbering you would have to renumber all the ones with a higher index numer.

But after orient is complete spindle.0.is-oriented is low. Even if spindle stop exactly on orient position.

Yes, I think that is the intended behavior  but I don't have a system to actually verify this.

'orient.is-oriented' True -> 'spindle.0.is-oriented' True -> 'spindle.0.orient' False -> 'orient.enable' False -> 'orient.is-oriented' False -> 'spindle.0.is-oriented' False

If you need a signal that stays True after a completed orient then you can use 'spindle.0.locked' which stays True until M3,M4 or M5 command

linuxcnc.org/docs/html/man/man9/motion.9.html

Still play with orient pid but often after orient is complete spindle turn little and is out of orient.tolerance.

Depending on the inertia of the spindle and the spindle drive I would expect that to be a problem and I don't really know how to deal with that if the spindle drive cannot get the spindle to stop quickly enough (unless maybe with a mechanical brake).

I'm afraid I can't really help you with the PID tuning but maybe you get some inspiration from this thread:
forum.linuxcnc.org/24-hal-components/458...-solved-yay?start=60

Spindle-at-speed is still High
spindle.0.index-enable After start rigid tapping is set to High but after hit index is set to Low and dont go to High anymore.

Sounds correct.
I think I may have made a mistake resetting the encoder with the input-index. So remove this:

net spindle-input-index <= hm2_5i25.0.encoder.04.input-index
net spindle-input-index <= hm2_5i25.0.encoder.04.reset

and add this instead:

net spindle-index-enable <= hm2_5i25.0.encoder.04.input-index

 

Hm, not sure if this works but replace this:

net spindle-index-enable <= hm2_5i25.0.encoder.04.input-index

with this:

net spindle-index-enable <= hm2_5i25.0.encoder.04.input-index
net orient-enable => and2.1.in0
net spindle-input-index => and2.1.in1
net encoder-reset <= and2.1.out => hm2_5i25.0.encoder.04.reset

Don't forget to 'loadrt/addf' the 'and2.1' component