Arc Ok Problem

I am getting back into plasma cutting on my table, I am using QTPlasmaC and for some reason the Arc Ok does not seem to work.
I can see the Arc Ok In signal in Hal Show and have check the Hal file and the signal is shown attached screen shot.  It will fire the torch 3 times before stopping.  I changed the Mode from 2 to 0 no change.  I checked the perimeters and the Arc volts hi and lo are as thee manual stated.
In the perimeters should the THC be set to auto start?  did not see that until disconnecting the Plasma unit to put thing away.
I cut fine some time back when I started using the table of routing. 
Thanks for the help

thanks
I am using the THCAD-10 so I will switch to the mode 1.  I had manually changed in Hal from mode 2 to 0 now will change to 1.
For clarification when running Pnccof should I select mode 1 at that point.
Not having much experience with LinuxCNC but trying to learn.
I am working with another guy starting to build a Plasma table and may use LinuxCNC. Mesa seems to be out of stock on the THCAD-10. Your answer to me hints a Proma THC could be used.

The THCAD-10 and THCAD-5 have been replaced by the THCAD2
 

thanks I will pass that along

The mode 1 change did not show a difference, I turned off auto volts and that made a big difference. But not with a normal cut plan it fails with a material list error #1000000. It fires the torch and errors out. Looked at the G-code there is material info lines but nothing about a list number. I changed the material list to the one posted for Hyperturm tool list for the metal thickness. Still fails I am using SheetCam and LinuxCNC PlasmaC post processor. Have a customer wanting cut parts that I cannot cut because of this problem.

Can the GIU Axis be configured to run Plasma????

Thanks I just found that myself, hope that gets me going to cut the parts then I can dig into why QTPlasmaC is failing. I did use it some time back so was surprised when I gave me a problem
Have fun
Tom