THCAD-10

It has been several months since I last posted to the forum related to my brother's 4 x 4 plasma table...It has been running great and he has thus far cut many different parts on it and different types of materials (CRS, copper, S/S)...To date all of his cutting has been done without use of the THC as that was the last feature that was on the list to be completed to finish the project...We went through all the setup files relating to the THC as best as we understood what was necessary and gave it a try...I will summarize our findings and ask for guidance as the THC is not working as it should...

The present system is a MESA 5i25, 7i76, THCAD-10 stepper system...All mechanicals working properly including limit switches and Z-axis probing...We have borrowed the software logic code from John Thornton's setup and modified as necessary to work with our MESA components...All good now with the exception of the THC...Jumper settings for the THC are set to Unipolar and 32

Experimenting with the THC for the last few days here is what I have found...When doing a test cut (6.00" straight line) with the THC enabled, the machine starts and completes its probe, moves to pierce height, starts arc and moves to cut height, starts the cut and in a very short distance starts climbing until the arc terminates...PYVCP spin box settings for THC are set to 0.2 velocity tolerance, 118 volts requested, and 1.0 volts tolerance...Also discovered, that merely activating and deactivating the THC check box raises and lowers the Z-axis within approximately 1/2" to 1.0" or so...Observing the LED indicators in the pyvcp panel when the THC is enabled, only the Velocity Status indicator shows activity...No activity on any other LED indicators...

What I conclude from the testing so far is that there is some type of communication with the THC enabled having the effect of raising the torch, although as described above, causes loss of arc before completing test cut...IMHO, this leads me to believe I am missing proper I/O connections in the software files, but as I copied, pasted, and modified as best I could understand the THC portions of the software, anything in error, necessary or unnecessary, not properly connected in HAL files or otherwise, remains outside of my comprehension...I am attaching the files for the machine and would appreciate any input...

Mike

Hi Mike,

I see you are using the THC and THCUD component in your hal file.
I assume you are using Axis with a side plasma screen, because of your pyvcp description.

from John Thornton's setup and modified as necessary to work with our MESA components

Modified code is alway's difficult. One mistake digit (upper or down letter) can crap the whole code & process, that is my experience so far.
Your concluding issue is very difficult to locate for me at this moment.

But the good new is, i am developping a new component in c, special for plasma an external offset branche. Maybe some day you get it.

thanks for the comments Grotius..
As I was again reviewing all the HALfiles I realized that I was attempting to load and use both the THC and THCUD and removed the THC in favor of the THCUD and made the appropriate (I hope) changes in the rest of the HAL and Postgui files as set forth in the THCUD.comp file...You are correct as far as using Axis and side screen from JT's setup...

I will look forward to your new component, I read some about it in one of the other posts...I will anxiously await its completion...

Mike

Hi Mike,

THC and THCUD loading together is not a problem at all, but for your application you need only one THC component.

The THCUD component is i think a update of the THC component and is nice to have in linuxcnc available because there are limited options available.
The THC and THCUD component is written by an software expert.

But the real software expert is the Toma hal coder. He is like a professor.

Coding in c is much easyer then coding in hal.

Its probably not much point trying other things until you test your revised config. However, using halscope while cutting to observe the commanded voltage and actual voltage is very insightful. It would be helpful if we knew:
1. Which Plasma cutter
2. Are you using divided volts or raw volts as input to the THCAD-10?

One problem could be that your commanded volts is a long way from the actual volts. eg. commanding 100 volts but cut volts is actually 90 volts so an immediate unexpected correction is made when THC is enabled. That is why I wrote the torch sampling component shared elsewhere in this section.

The other concern is that if you are using divided voltage from the plasma cutter that the internal resistor network is seen by the THCAD-10 as belonging to it and thus the input is further divided than you expect.

Hi Rod,

In the new THC component, compatible with external offset branche. You can not blame the arc input, divided voltage or not.

So far as i can see, the EO (= external offsets) compatible component has a destination height output value.
The (pre processor) component for mesa connected to this new THC2 component, here the mesa stuff must do it's work correctly.

So for sure. For parport applications. It will work better then Mach3 for ever. As linuxcnc platinum boarder i am sure about
we have a golden future over here if we only publicate .SO files.

Rod, please make the Jacuzzi fertig.

Grotius wrote:

The THCUD component is i think a update of the THC component and is nice to have in linuxcnc available because there are limited options available.
The THC and THCUD component is written by an software expert.
.

THCUD.comp is designed to receive the up/down commands from something like a Proma external THC.
THC.comp is designed to replace the external THC and use the Mesa THCAD card to measure torch voltage

So in your case with a THCAD-10, you should be using the THC component.

From all accounts, the THC.comp solution is a proven working system so don't get confused with other talk about external offsets. (eg. Ignore us)

Grotius: "THC and THCUD loading together is not a problem at all, but for your application you need only one THC component. The THCUD component is i think a update of the THC component and is nice to have in linuxcnc available because there are limited options available."

I have updated the pertinent parts of HAL in accord with changing over to use the THCUD component now...At least as far as I understand it is newer and revised...

rodw: I used the HAL meter last night when I was testing...In response to your questions: 1. Plasma cutter is Hypertherm 45 and;
2. I can't remember right now if voltage is raw or divided...
Would using raw voltage be the reason for the torch almost immediately move away from the cutting path?

Thanks rod and grotius for your responses...

mike

Rod,
I just learned something from your post, that is, what is meant by an external THC as opposed to the MESA style THCAD-10...I had just assumed (based upon my limited knowledge) that there was no differences, as they were both used for THC control...

Thanks much for that input...I have change my HAL and related files back over to reflect use of the THC component...

Mike

Hi Mike,

You can see directly if it has to do with THCUD or THC.
in hal it always show's it's full name. So don't worry about that.
But only do this : #loadrt THCUD will be sure everyting is out of function on that component. (the # sign)

If you don't have opened the Powermax45 and it has a cpc port, it's for sure a 1:50 divided signal. If you open the Hypertherm,
you see a swich box to do 5 other settings in bit format. Scale 1 to 20 for example. But standard scaling (voltage divider settings) 1 to 50 is perfect.

Raw arc with xp45 is max ca 300 volts. You don't want this near your mouse.

(eg. Ignore us) Yes, but we are here and making it perfect within time.

@Rod,

I remember, the THC is only for Mesa. THCUD is for up and down signal.