I have one of those Parkside things with the built in compressor, as my existing tiny Bambi compressor was only required to operate power drawbars.

One requirement for a CNC plasma source is to implement a pilot arc starter. Although the internet is full of people connecting the pilot earth wire to the ground connection on an HF machine, that seems to be a bit of a bodge that damages the consumables and robs the arc.

A better scheme that is used on the more professional machines is to disconnect the pilot arc when the main arc is established to the workpiece. Of course, in these machines it is straightforward to control the pilot arc using the HF enable circuit but to modify a non pilot arc machine in this way requires a bit more surgery. There are add-on circuits that detect the arc voltage of current but it seems to me that the required control signal is already present within the unit.

Naturally I have wasted no time extracting the main PCBA from mine. I'm going to fit a second relay across the HF circuit relay and bring a couple of wires out. Usually there's also a power resistor of around 5 Ohms to limit the pilot arc current.

Yes, there are a few risk areas, such as whether my servos will run fast enough, whether I have enough noise immunity in my machine controls, is the Parkside thing up to the job etc.

I've come up with a concept where the workpiece is placed on a grid that is mounted to and in front of the knee. This will avoid water slooshing about and keep sparks and water off the machine. The torch will be mounted off the table so that X and Y movement of the table translates directly to torch movement.

The Newker integrated controller on this machine is a 4 axis milling controller, so I've modified the Fusion post processor to work as a plasma controller. The height movement between piercing and cutting isn't so simple, mechanically speaking, so I may use a solenoid and further modify the post processor to use a macro (relay output) instead of a Z move.

I don't have any room left in the workshop to house a proper plasma cutter and I can't easily move the Bridgeport out so it can be sold off, as I seem to have a 3 tonne CNC lathe blocking the passage. Besides, I need to get several more jobs started that will drag on for a few years. You know how it is. So this is really an experiment and another interesting area of learning. If it goes well, there's got to be the danger of suffering another serious accident involving a mouse. Again, you know how it is.