High frequency ohmic struggles

I am trying to get ohmic sensing to work using a mesa 7i97t, thcad-2(jumpers on low range and 1/32), and a 24v power supply(meanwell hdr-15-24).

I have tried rodw's ohmic comp and ohmic comp3, as well as try many variation in the hal file mentioned in the original ohmic post/thread on this forum.
While I have no doubt there are issues with my hal file, I can't seem to get any response from the encoder in halshow, so I'm starting to wonder if I'm even getting a signal.
I tried using a multimeter to test the thcad outputs, but I don't know what I should be looking for. Is there a good way to test this?

I have it wired up similar to the diagram in the manual. I am using a 370k(or maybe 390k, can't remember) resistor between 24v+ and the thcad.
Since all the controls are on the gantry(this is a retrofit), I connected the 0v on the power supply to the buss bar on the gantry which in turn is wired to the star ground on the frame. I also connected the thcad shield terminal to the same buss bar.

The plasma is a ht hsd130 which uses a high frequency start. I previously damaged a 7i97t(which has been replaced) starting the plasma. I wonder if I could have damaged the thcad, as well. 
How can I protect the mesa boards from the high frequency start? Would a relay between the torch tip and thcad be enough? I assume it would need to be a special relay.

Attached is a hal file I have been using.

 

Ohmic sensing on a HF start plasma torch is very tricky because of the 20 or so KV
Tesla coil used for starting the arc

I suspect you would need a large (say 1 UF 600V) capacitor from the probe to table
 to bypass the high voltage high frequency  spike from a probe strike to ground.


To check 7I97T encoder inputs, set the encoder to TTL mode, then ground the
A+ B+ and IDX+ pin in turn and see if the corresponding pins change in halshow.

Pin names would be like:

hm2_7i97.0.encoder.00.input-a
hm2_7i97.0.encoder.00.input-b
hm2_7i97.0.encoder.00.input-idx

Since replacing the 7i97t, I haven't had any of the ohmic wired up while the plasma is running. I have only tried to get it working with the plasma off, so the 7i97t encoders should be fine.
I was more referring to the thcad output.
Can measure voltage between f+ and f- with the ohmic circuit open and again with it closed to test the thcad?

For the capacitor you mentioned, would I wire the torch tip to capacitor and the other end of the capacitor to star ground. Then run a separate wire from torch tip to thcad? Or would I wire both the star ground and the thcad to the capacitor?

Do you have a blinking yellow light on the THCAD?

If you set  the THCAD frequency divider to 1X you can probably
measure a change in the frequency as a voltage change on the F+
or F- relative to ground. Not sure of the polarity but full scale
should be about 2.5V at full scale and maybe 0.3V at 0V

I would need to study a bit about the probe wiring.
 

There is a yellow blinking light. I'll try to see if I can measure anything on 1x with my multimeter.
Thanks for all your help and the awesome mesa products.

On another forum, I saw the os3t-01 from robot3t.com. It's an ohmic sensor that can supposedly handle hf.
Does anyone have any experience with this company or their products?

I think there are much bigger issues than ohmic sensing from what you describe, namely bad wiring/shielding/grounding in general, so my strong advice is:
-do not fire the torch before grounding things properly
-make sure the power socket and the machine are grounded and the shop has proper ground rods
-make sure the PC is grounded
-make sure the plasma source is grounded
-disconnect anything ohmic related from Mesa
-disconnect the THCAD from Mesa
-connect THCAD only to plasma source
-wire and test the new Mesa, make sure everything works properly, wire the Torch ON to plasma source
-power on the plasma source and cnc machine and start QtPlasmaC or PlasmaC or whatever you are using
--enable LinuxCNC, jog around a bit, then use the Torch PULSE on screen set to 1 sec (default) ONLY ONCE !!!!!
-did anything move/burn/make strange noise/screen blank/PC turn off or lock/Mesa reset or lost connection/THCAD smoking ?
If any of those is yes, turn OFF everything and go back to wiring/shielding/grounding (goggle Hypetherm grounding they have a nice general guide)
If all above are no, try the pulse again and again, waiting 5 seconds between, make the pulse time 3 seconds and test several times again
If all that goes OK, try to do some manual cutting (use F9 to probe and fire the torch then jog manually), do several short cuts
If all good, power down and wire the THCAD to Mesa, do the same procedure from the line with two - signs above
ONLY if all tests go without issues proceed to wiring and testing Ohmic sensing. Again test from line with two -