Just because the voltage supply is a 36V, doesn't mean interference from the plasma won't cause a spike on it, however I'd expect that to of damaged something by now.

The stepper motor fault detection will be a reasonably sensitve system, which is what is probably causing the fault. I doubt there is an actual fault, but noise on the stepper wires may cause the driver to think it's dected a fault. Or even just noise in general may be causing problems in the stepper drive. Have you tried putting a additional capacitors on the power inputs to the drives?

By plasma wiring, I mean the torch cable. Is it routed near the stepper/control wiring?


Cable sheilds should only ever be connected at one end!
By connecting them at both ends, you create ground loops which are far more likely to cause noise issues. I'll try and dig out a link to the usually recommended earthing/grounding guide.