Hi all,

We're experiencing issues with the plasma torch diving down into the sheet at random points in longer programs. We're using build 4809 with Warp 9 ESS 285 smoothstepper and Proma 150 THC. This is the start of a block of our code:

N8505 G0 Z0.2
N8510 G0 X-6.8795 Y111.7594
N8515 G31 Z-100. F80.
N8520 G92 Z-0.4
N8525 G0 Z0.
N8530 M62 P1
N8535 G4 P0.001
N8540 G1 X-6.7756 Y111.8194 Z0. F80.

At G31 the torch probes down (hence Z-100) until the switch is triggered. Then it should G92 the Z position the torch is sitting at to Z-0.4, and then rapid up to Z0.

However, at the G92 line instead of setting the Z to -0.4, we find it set to Z99. The next line is then read to rapid to Z0, causing the torch to crash down into the part.

We have tried changing the post processor to output arcs as R instead of I, J, and K based on the solution we found to a similar problem with our waterjet, where the Y axis was losing its zero mid-program. We have also tried reducing the THC velocity to 5% up and 10% down and reducing the rapid rate to 25%. None of these solved the problem.

We also tried disabling the use of sequence numbers (ie. no more N1234 preceding lines) from the Fusion 360 post processor. This actually made it so that the crash happened right away at the first G92 line, rather than at a random point further into the program. Hopefully that information can help.

Any advice is appreciated!