As I've mentioned in a few other threads, I've been re-building a CNC Plasma table over the past few weeks, and at last today I got it back into the welding shop, where I can power up the Hypertherm and actually cut rather than emulate, as I had to in the 'clean workshop'

Well oddities abound - all excited I loaded up a sheet of 3 mm mild steel (zintec coated) and set it to cut out a 'bearing baffle' that I'd previously designed and cut on the table before I started updating it. It started off ok, measuring Z=0 and setting torch height, but when the actual plasma should have been initiated, movement stopped and the Hypertherm 45 displayed an error light showing 'loose consumables'

In the guts of the machine torch is a micro-switch that closes when the consumables are tightened onto the torch - I buzzed out the cables to the plug and everything was fine - closed when tight - open circuit when loose. Nothing to do with the CNC controller, this must be an internal issue with the Hypertherm - maybe something I disturbed when I remade the plasma voltage cable :scratch: . So I pulled the torch and cutter off the table, opened up the Hypertherm, and buzzed the switch all the way through to the main board - no issues. Put it back together and back on the CNC table, and tried again - same issue :bang:

There is a button on the screen allowing me to trigger the torch on and off manually. Out of curiosity I clicked it, and initially there was a very weak release of compressed air - I did this several times then the torch fired up, no errors and full air discharge. I've come to the conclusion that the air solenoid in the Hypertherm was perhaps sticky and has now cleared, but why that should make it report loose consumables is a mystery :scratch:

After that everything worked as it should - very odd. So here is a video of it cutting, and a picture of the result. Not surprisingly the cut quality hasn't improved since I did the re-build, but that's due to worn consumables and non-optimal cutting conditions controlled by program. Also there is an issue with the cut not ceasing fast enough at the end of a feature - I need to tweak the post processor to stop the arc marginally before the end of the feature to eliminate the nick it's producing.