Including "arc ok" requirement in the post processor

Hello,
A question and thread for those among us that use a plasma cutter with grbHAL as the software of their plasma cutter controller. I am in this case and it looks like my new plasma cutter has issue keeping the arc steady. (Especially at first cut. Once warmed, the generator keeps arcing in a more stable way) How would you have the postprocessor generate a gcode making the CNC wait for an arc OK signal before moving ? Is there an easy trick about that ? In the blow back ignition case (my case) I guess it would require to do M5 an M3 again to actually start the arc again, for those with other ignition, keeping M3 until the arc is enough steady before resuming to pierce and move may be enough.

Less about postprocessor, more about electronics, this question also leads to another one : how would you wire the signal ? In my case, I am using T41U5XBB, which have besides their probe, cycle/start, Hold, door and switch limit inputs, a few digital inputs. I do not know how to use them (but will ask to the designer) I think the probe input could be used as an Arc Ok input too, as suggested here
https://github.com/gnea/grbl/issues/254. In this case, a relay may be required to isolate bot outputs, the probe on the CNC from the arc ok switch in the plasma cutter. (Just a suggestion, not sure it would be more simple)
Does anyone has some idea about how to deal with this issue ?

check the doc here: GitHub - grblHAL/Plugin_plasma: grblHAL plugin for plasma cutters (THC)
It looks like $350=2 perhaps, with a Proma 150 THC ?

Barrett also makes a THC BOB that digitizes the arc voltage for $350=1 case.

Thanks. However this does not fully address my issue. I have not yet implemented my THC, even though I already got the THC board from Phill Barrett. But I would like to address the arc stopping issue. I thought I read somewhere about a code to stop moving the torch if the Arc ok signal is interrupted, so that the CNC could then be put on hold, and I could at least manually rewind the gcode to the same cut. Most of my first job M3 signals fail due to the plasma cutter being faulty. If the second cut occurs quickly after the frst one, the plasma cutter is hot and works fine? But this first strike issue is annoying,n I would like to be able to automatically stop aborted cuts in order to repeat them.

The gcode to fire the torch, then wait for Arc OK signal is most likely one in the same with grblHAL, but if not, should be 2 gcodes in sequence then. You’ll have to read about all that in grblHAL doc, along with using $350 register to set it up correctly. I don’t understand how you want to get this behavior otherwise. You’ve got to connect something to ArcOK, a signal set by your plasma power source. [ which btw, what is the brand and model of your plasma power source? ]
“Something” is Barrett’s THC BOB, which then connects to his grblHAL controller BOB. Post to his discussion board if necessary- phil-barrett/grblHAL-teensy-4.x · Discussions · GitHub

The ArcOK signal and associated gcode to wait may or may not solve your first pierce problem. You’re not describing it like I’ve seen the first pierce issue behave. In my case, with either a PMX65 and a PrimeWeld Cut60 unit, the first pierce was over cooked (too long) if I used a pierce delay sufficient for subsequent piercing which occurred with post-cut-airflow in the torch. So my first pierce adjustment is to subtract 0.1sec from my tool defined pierce delay.

You are right.
I had actually read, but forgotten, that the THC features, at least the one from Phil, enables both height consistency AND Arc OK check. I was focusing and solving first the arc stability without considering I could have two birds on a single shot.
It looks unfortunately lie I soldered badly my Teensy and shorted or destroyed something I can not see, because it does not allow me to flash it again. I had to buy a new one and have to wait for its delivery now. I will definitely flash it with THC plugin BEFORE soldering the pins. May that serve as an advice for the other bad solderers.

As for my plasma source, it is a rather cheap one, regarding industrial standards at least, it is a HeroCut 8000, from China. But I read good reviews about it.