Torch Tip drags on plate in Dry Run or AVTHC off condition

Matt, you are correct that Dry Run and cutting with AVTHC disabled should still use the cut height from the selected Hypertherm chart.

Dry Run still performs the complete fire_torch touch-off sequence: it probes the plate, compensates for the floating-head travel, establishes Z zero at the plate surface, moves to pierce height, and then moves down to the chart’s cut height. It only skips firing the torch. With AVTHC disabled, the same sequence runs, but Z remains at the fixed chart height instead of being adjusted from arc voltage.

This most likely means the floating-head compensation needs to be calibrated. Open Edit → Machines, select your G7 machine profile, choose Subroutines, and expand fire_torch. The setting is called Floating Head Backlash.

With the torch output safely disabled, jog down until the nozzle just touches the plate and note the Z position. Continue downward in very small increments until the Floating indicator beside the Z readout lights. The difference between those two positions is the value to enter for Floating Head Backlash.

Also, if the machine uses only a floating-head switch and does not have working ohmic sensing, turn Ohmic sensing (off = floating head only) off in that same fire_torch panel. Otherwise, the routine may select the zero-valued ohmic compensation instead of the floating-head compensation.

After changing that value, run another Dry Run. The torch should touch off, lift, and trace the program at the Hypertherm chart’s programmed cut height rather than dragging on the plate.

My Floating Head backlash needed adjusted.

Now it works as it should with the Ohmic Sensing, but with the Ohmic switch toggled off. Still drags on the plate with Ohmic on or the Ohmic wire disconnected to force floating head sensing.

My machine has both Floating Head and Ohmic.

Matt, thanks for the follow-up. You were right—this was not just a calibration issue. I reproduced the same behavior on my own plasma machine and found a bug in the touch-off routine.

The controller was recording the correct ohmic contact point, but Z continued moving slightly while it decelerated. The routine was establishing plate zero from the final stopped position instead of the exact recorded contact position. On my machine that difference was exactly 0.060", which completely consumed a programmed 0.060" cut height and put the torch back on the plate.

I also found that the older routine could determine whether ohmic or the floating head triggered by looking at the inputs after Z had already stopped. Depending on the machine’s deceleration and switch travel, the floating switch could be active by then even though ohmic actually triggered first.

Scott has told me that ohmic works perfectly on his machines, and that can still be completely accurate. How visible this problem is depends on the Z acceleration, probe feed rate, floating-head travel, cut height, and compensation settings. On some machines the older floating-head compensation may also accidentally cancel most or all of the stopping-distance error. I don’t have one of Scott’s exact profiles and probe logs yet, so I can’t say which difference has been masking it on his machines.

I corrected the routine so it now uses the exact captured contact position and trigger source. Z deceleration after contact can no longer consume the programmed cut clearance. I tested the fix physically on my machine in Dry Run and confirmed that the ohmic input released and remained released while it traced the program at the commanded 0.060" cut height.

The fix is committed for the next JetCad3 update. Until then, since floating-head-only mode is working correctly after your backlash adjustment, I would leave the Ohmic Sensing option off. After updating JetCad3, you’ll also need to select the amber update shown for fire_torch and choose Replace with bundled. JetCad3 does not automatically overwrite installed subroutines because some users customize them.

I thought I should add to this and let you know that since we previously turned on OEM Connect for your machine profile for other stuff I helped you with, that allowed me to reproduce this with your actual machine parameters. Otherwise, without that ability, this would have been very difficult to nail down without a lot more back and forth here with the needed information. I'm pretty sure this was an obscure bug missed since June of this year but never effected anyone else enough for it to be noticeable.

Ohmic Sensing and Floating Head works as it should should.

Thanks!