Advanced Search

Some things like this
www.aliexpress.com/item/1005005042915668....product.mainProduct

You should be able to find a LM358 low signal amplifier to bring it back up to near 24v for a Mesa analog input or simply use a Mesa THCAD-5 to read the voltage output which would be much more accurate.

This is Linuxcnc!

QTplasmac should hold its own against your integrated systems, possibly exceed. The $70 Mesa THCAD2/5/10's  are built for the dirty plasma environment and can handle 500 volt over voltage indefinitely. They are individually calibrated and are connected to an encoder input on a  Mesa card. Linuxcnc converts the frequency in software and controls the torch height internally.
Here is a walkthrough I did

And the cut recovery features


Data types are described in the docs here linuxcnc.org/docs/2.9/html/hal/comp.html

• TYPE - One of the HAL types: bit, signed, unsigned, or float. The old names s32 and u32 may also be used, but signed and unsigned are preferred.

I think the xml uses S32 and and U32


 

Its not really a big issue to have 0-10v. You can add a THCAD-2 to an encoder input to protect the electronics.
Did you get something done for your xml file? If not, just started with one pin and do it progressively.

Hello everyone!

I'm using a THCAD2 with QT/PlasmaC. I've verified the voltage conversions are correct in my configuration and started to run with the THC active. I noticed that it's moving up/down (I see the THC become active and up/down indicators flip on an off) but never gets to my setpoint (78v). Furthermore, it seems to actually be controlling to 100v (which it seems to achieve quite well). 

I'm using (I think) M190 Px to define the material parameters (configured in the GUI where I've set the voltage). Is this the wrong place to define the setpoint for the THC or is something else wrong?

Thank you!

rodw wrote:

Its been a while but the Everlast confuses the THCAD. The THCAD thinks the everlast's resistors are part of the THCAD's scaling. Do this:
​​​​​Do not use any scaling resistors
Set the Everlast divider to 16:1
Tell the THCAD/Qtplasmac you have a 24:1 divider.
That gives you a 240 volt full scale reading which is perfect!

I found this out after testing with 2 x 48 volt power supplies in series to give 96 volts and measuring with a multimeter.
There is a link in the QTplasmac docs to a calculator to determine the settings to use. Use it to confirm your calculations
​​​​
 

Actually  the divider resistance is always a factor but its normally low enough that is does not cause a significant
error. It must be that the Everlast has a  quite high resistance divider. You can calculate the scaling change by
measuring the divider resistance at the output of the divider  (with the plasma power unplugged) and ideally
the plasma output shorted to ground. This resistance adds directly to the THCADs input resistance (100K)
so changes the scaling factor to nominal_scaling_factor * (100K+divider_output resistance)/100K
 

brickhousemb wrote:

Yea I have followed your thread to get there I am now.  I put in the numbers in the QrolasmaC GUI this time instead of in the hal and for some reason works. 

This is normal and the expected place to enter the THCAD calibration data.

Glad you are up and runnng

Yea I have followed your thread to get there I am now.  I put in the numbers in the QrolasmaC GUI this time instead of in the hal and for some reason works.  I also ended up flipping my input to the THCAD wires and it now reads positive.   I did both at the same time so it was too variables and I'm not sure. Actually switching the wires did anything but at any rate it's working. 

I've only done very short 50 mm cuts So I don't think the ArcOK has time to engage.  I'll dial that in soon.  I appreciate the help. 

THCAD-10 and plasma wired and ready for the configurator and the first cut. Most probably next weekend if all goes as planned.

Its been a while but the Everlast confuses the THCAD. The THCAD thinks the everlast's resistors are part of the THCAD's scaling. Do this:
​​​​​Do not use any scaling resistors
Set the Everlast divider to 16:1
Tell the THCAD/Qtplasmac you have a 24:1 divider.
That gives you a 240 volt full scale reading which is perfect!

I found this out after testing with 2 x 48 volt power supplies in series to give 96 volts and measuring with a multimeter.
There is a link in the QTplasmac docs to a calculator to determine the settings to use. Use it to confirm your calculations
​​​​

I am running a clean install of QtplasmaC using a Mesa7i76s, a THCAD-5 for hyersensing, and a THCAD-2 (set to a TCHAD10) for THC.  I have an Everlast 102i and I have hooked into the divided voltage pin outs.  I have set the jumper in the machine to 16  The hypersensing works, however, but for the THC I am running into the negative volts issue when cutting (around -24) which also leads to the ArcOK not working.  I understand that due to the everlast's design, the 1/16 read differently.  The best I can figure it reads 1/24 after it is run through the THCAD.  Based on this, the manual, and my specific THCAD I have run the following numbers: 

Full Sacle kHz 900.4 0V Divided kHz 37.51
0V 99.2 divided 4.13
Divider 24
Frequency per Volt 3338.33
Volt Scale 0.009586
Offest 4.133.33

 I think I need to update my hal files to include these new numbers.  I have attached the applicable files.  I can run a backup and post those files tonight (if needed).  I would appreciate any feedback on whether I am correct to use the 1/24 to correct the voltage, and if my math is correct.  If it is correct which file would you update?  Should I make a custom hal file?

Thank you for the help.

Yes, to read the THCAD you need to:
1. Select the correct encoder
2. counter-mode for that encoder must be set to 1

For the rest to work the velocity pin of the selected encoder must be
connected to the appropriate qtplamac signal

Using PlasmaC or QtAxis PlasmaC or QtPlasmaC?
-22 should mean the THCAD is connected and sending pulses, the scaling and offset are not set in the settings tab.
Does -22 fluctuate?