Good afternoon (o;
Recently bought a plasma cutter with THC output and control input.
So what do you use as controller board/controller software for plasma cutting with THC support?
Thinking of grblHAL as it has THC plugins and runs on newer 32bit microcontrollers.
BTW: Is the THC voltage output normally filtered? Meaning no HF ignition spikes?
grblHAL is a good choice, I like this BOB- https://www.grbl.org/grblhal-breakout-board , but there are others. Also consider LinuxCNC, perhaps better choice but requires a computer with Linux OS as a controller and machine interface (jogging, gocde flow, etc.) But the other choices still require a computer usually for running a gcode sender to a MPU controller, not needed for LinuxCNC or Mach3/4, yet another choice.
re. filtered THC voltage- unfortunately noise (EMI) filtering is not always employed in the plasma power source. Hypertherm of course is the exception, they do it and do it well with a Pi filter design, one on each of the + and - wires. That is the essence of their CPC Voltage Divider kit/board. Other brands of plasma power sources are hit and miss if they filter those signals or not. Voltage dividing alone is not noise filtering, in fact it can worsen the effect of EMI since it lowers the signal to noise ratio, the signal being DC voltage the THC wants to see vs. the EMI noise picked on the wires due in part to EMI radio signals from the plasma arc. EMI also comes down the wire from inside the plasma source, hence the need to filter the signals before leaving the box. Filtering should also be performed at the THC and I suspect it usually is. Lastly, you can always throw a ferrite core (choke/inductor) on the shielded cable for arc voltage sensing at each end of the cable at the connectors. This can squelch EMI just enough to make the system work better or well.
re. HF ignition spikes- ha. Easy fix. Don’t use a plasma power source that uses HF ignition, use the ones that employ ‘blow back’ ignition for pilot arc.
I asked my supplier if the Weldinger PS 42 pilot Pro CNC has any filtering circuit on the THC output. No answer yet. Also not sure how the pilot arc is triggered…nothing in the description.
Guess I just have to hook up an oscilloscope and see what is coming ( the scope running behind an isolating transformer and the input divided ;o).
My thought is anyway to use the THC voltage to drive an optocoupler circuit…as the input voltage is high there isn’t much non-linearity to expect…and 1mA couldn’t bring a plasma cutter on its knees (o;
grblHAL is still my first choice…but probably gonna do my own PCB design so no unnecessary stuff is included. I already have a Linux laptop in my workshop with QCAD, SheetCAM and UGS to drive my milling machine.
Ordered some STM32H723 evaluation boards at Digikey which should arrive tomorrow to see how grblHAL and the THC plugin runs on it…and then take this as a base for my design.
You might chat up Phil Barrett’s, he’s the designer/builder/seller of the BOB link I referenced. He makes a THC board to go with it and grblHAL that already uses opto isolation for EMI mitigation, and he has opto isolation for digital and analog sense I/O and limit switches on both the THC board and the BOB as I recall. I have one, it’s pretty nice. I got the Teensy 4.1 model and I run grblHAL on it in my office disconnected to CNC machine, its just for bench testing gcode on grblHAL. The BOB is still in kit form and not in service for my machine due to having fixed my atmega2560 solution with 4 axis grbl v1.1i in the meantime of ordering/recieving Barrett’s; I got his BOB as plan B. If it ain’t broke, don’t fix it rule is in effect until I need Barrett’s solution.
I forgot to mention- the hand torch in the product link for the Weldinger PS 42 pilot Pro CNC seems to be a Tecmo IPT60, which would mean the plasma power source uses blow-back pilot arc method. Strange they don’t show the torch consumables, that would clinch it.
Good luck with your build. Seems like you have good background and the right tools for building it.
So this is the simulation output…took a while until I got used to LTSpice (o;
And this is the schematic for the simulation:
Interestingly simple circuit. Well done.
Based on your graph of output voltage vs. input, you’re going to want to invert the output.
The signal current and voltage isolation is clear, however does the opto coupler also as act as a low pass filter taking out EMI which is sure to be on the input signal? Depends on the opto coupling response time.
Regardless, consider adding a low pass PI filter on the inverted output, and also add a ferrite core at the other end of the output signal wires where it connects to the THC. These circuit precautions and EMI mitigations are needed since EMI is also present as transmitted radio waves from the arc jumping the gap while cutting.
Nice job.
Consider using the Proma THC 150 if you are using grblHAL, which allows integration of the THC with grblHAL handling Z motion to maintain arc gap.
That was just a first try how to convert the supplied THC voltage and convert it optoisolated…no filtering thoughts at this stage 
Gonna build up this circuit and log the voltage with my ADC USB device during cutting process.
You know: In theory there is no difference between theory and practice, in practice there is
