new day new problems

Back again with new problems,  some time I get good cuts and other time not at all.
First thing when starting up I get a material error see attached.
 
I can cut a straight line (7") no problem,  Add shapes to a cut and either stops about 1/3 the way through, or may crash the torch into the metal.
Today after a crash it gave my the restart option, I moved the torch a small amount for restart off the path.  The cycle resume button was flashing, when I pushed it the torch moves down touched the metal and retracted.  And I get a message about the probe.
 
 
I can run without the torch without error with multi-part G-code.
Just did an update/upgrade to V2.10-005.033 do not know if this will help with the problems I am having.

Ok this may have been the problem I have with the material list. I looked in SheetCam and I had many other then plasma tools that showed very high tool number. I will rerun the tool conversion in the morning with only Plasma tools.

You cannot have material numbers of 1000000 or greater in the material file.

The last error shows that you are using offset probing. You must have the probe height set to higher than the thickest material that you will cut. Probe height is measured from the Z axis lower limit, it is not the material thickness.
linuxcnc.org/docs/devel/html/plasma/qtpl...plasma:initial-setup

 

Thanks as usual for the info, you guys are great. I assume with the "auto Volts" off it gets the THC volts from the material list or G-code setting.
Question under the THC there is a box named Auto Activate, should it be on or off??

Redoing the tool list in SheetCam (plasma only tools) and then the material conversion in LinuxCNC corrected the problem on the message at the start about the material over 1000000.

Well the pest is back, different problem.
I have updated to version 033, and now my plasma table has a new problem.
When I fire the torch I get an error showing from one to all limit switches tripped.
Now I am running Proximity sensors which work fine under the home all at the start up. And when I try to test the Ohmic sensor I can see it functioning in Hal Show but when clicking on the Ohmic test button the display lightens up and the ohmic led does not light. I have attached a video showing these actions.
drive.google.com/file/d/1Fj7Qm9ozVSof0kO.../view?usp=drive_link
The only changes to the controller is the update to version 033.

I am not saying you are wrong but I have fought noise problems before on other tables I built.  Troubleshooting includes the "What was the last thing you did before the problem" .  With this it was the update/upgrade. 
I have used Cat 6 cables for signal with feed through panel plugs and shielded other cables. The shields are terminated in the controller only. The Plasma cable is completely isolated from all other cables I use a boom for that cable.  The attached video is from last month with an older version of QTPlasmaC without a single hick up. This all came from the latest update.
This is the 10th table I have built and I have used other control electronics and software, so I understand the noise problems that can happen. the other problem with the ohmic sensor came also with the new version, that is not noise related. As my other video shows the ohmic signal is input to QTPlasmaC Hal show but not through to where it need to be in the software.
drive.google.com/file/d/1mTytmUI8LUFoJ-7.../view?usp=drive_link