Plasma Torch Height Control

Hello all and a big thank you for such great software. I used it with success for wood routing. I have come unstuck now that some fairly tricky configuration is needed for my first plasma setup.

Due to a low budget I am running with a standard un-modified(so far) live cd 2.5.1 installation hooked up to a cheapass chinese 4 axis TB6560 Driver Controller Board from ebay.

The plasma cutter again is an entry level Cut 40 with no signals out like arc ok or a voltage divider. I may be able to get help to retrofit this circuitry but I need help with the software side.

I hope someone can direct me to some dummies guide to intalling THC into hal. I keep running into dead ends in the git repository 'no such project'

Please someone tell me what I need to do to get the component to display in axis. When I can see the interface for THC then I will be very happy indeed.

I have had great results 25% of the time but if the torch doesn't stay +-1mm from the work it all goes nuts. Ultimately I want dynamic voltage monitoring as the method but I am happy to use a switch/probe approach in the short term.

Thanks again.

Thanks John for the prompt response. Do I understand correctly that the coding is specific to the mesa thcad card only? Can you imagine it would be possible to use an arduino instead or is that a crazy idea?

If not then is there a way to install the thc comp in the simulator to see it and play with it before buying the mesa board?

Failing that can you suggest any points I need to consider when installing a switch to work with your touchoff routine (in particular, how to get around the wait for arc-ok which my unit cannot currently provide - its a pretty dumb plasma unit)

Thank you

The THC comp should be easily adapted to any voltage reading device. Because its used in a real time feedback loop its voltage input scheme is real time so may not be suited to USB/Serial port interfaces. The THCAD is just a very well isolated V-F converter that is read by either the hardware or software encoder component (using the encoder velocity pin which reads frequency)

One thing to remember is that both the plasma voltage tap and plasma GND are very noisy so if you use something like an Arduino, make sure the Arduino runs on a floating power supply and that any connections to the PC are isolated

Thanks fellas. Some good points raised there, and the modified touchoff routine.
What do you mean by floating power supply (not regulated?) sorry fairly new to this stuff.

I expect I will be posting again soon when my 2nd parallel card arrives and I try re-jigging Linuxcnc to talk to it. I plan to affect some kind of arduino bob that can monitor float swith for thc and ultimately monitor tip voltage whilst taking over the driving of the Z axis from the other driver board. Sure a timer is needed for reference for stepper pulses. I hear people use a 555 to keep the arduino running at a particular frequency.

Are there any 'Dont Do's' to selecting a suitable volts to freq system/circuit and I'm guessing the voltage divider needs to scale it down from 100V to about 5V before the vfd? Am I anywhere near?

Lifes too short for the amount of questions I have haha

By floating I mean the Arduino and its power supply have no electrical connection to the PC. All data goes over an isolated link to the PC. Not only is plasma voltage up to 300V, the ground reference on the plasma side is very noisy electrically so you don't want any direct connection between the plasma power ground and the PC via your Arduino or you may 1: kill the Arduino, 2: kill the PC, 3: have endless noise problems.

Hi All,
I am using the arduino code as supplied by Jeff Epler to read the tip voltage. I would seem to work well at this stage. I am also using the touchoff code that JT supplied as well as his THC component.
I have a problem with the Z axis being very sluggish and difficult to control. I have seen this with every THC component that I have tried. Is this supposed to be common or is there a problem somewhere? Maybe something with the z pos being intercepted by the component?