Hello, i upgraded to v5000 few day ago and since I have constantly getting random Z dives on G91 G00 Z0.0000 (Incremental move for switch offset distance comp). Sometimes it works sometimes it just dives till it triggers the Z safety switch.

ESS + mach4+ warprunner THC + touch sensor . I use Sheetcam for code gen

I have a touch probe and once it registers a touch on (>>>Probe/Ref<<<), it will randomly either continue the code properly or decide to dive. This is with THC on or off , makes no difference.This code has been running for the past year flawlessly on version 4809. 

Code example @ line : N0320
%
N0010 (Filename: ptest.plate.tap)
N0020 (Post processor: WarpRunner Mach4 v005.scpost)
N0030 (Date: 17/10/2022)
N0040 G20 (Units: Inches)
N0050 G53 G90 G91.1 G40
N0060 G64 (Constant Velocity Mode - normally used with Plasma)
N0070 m2020 (HC_Z_MAX_VALUE=1)
N0080 m2020 (HC_Z_MIN_VALUE=-0.1)
N0090 m2020 (TMC3IN1_AD1_DELAY_ENABLED=1)
N0100 m2020 (TMC3IN1_AD1_DELAY_VALUE=2)
N0110 m2020 (TMC3IN1_AD2_M62_M63_ENABLED=1)
N0120 m2020 (TMC3IN1_AD3_VELOCITY_ENABLED=1)
N0130 m2020 (TMC3IN1_AD3_VELOCITY_PERCENT=90)
N0140 m2020 (TMC3IN1_AD4_VOLTAGE_ENABLED=1)
N0150 m2020 (TMC3IN1_AD4_VOLTAGE_BUFFER=800)
N0160 m2020 (TMC3IN1_AD4_VOLTAGE_CHANGE_PERCENT=5)
N0170 m2020 (TMC3IN1_AD4_VOLTAGE_RESPONSE_PERCENT=0)
N0180 m2020 (TMC3IN1_AD5_VOLTAGE_ABOVE=15)
N0190 m2020 (TMC3IN1_AD6_VOLTAGE_BELOW=14)
N0200 G4 P0.1
N0210 (Part: ss.test.plate)
N0220 (Operation: Inside Offset, round, T9: Plasma,1mm kerf, 125ipm, 30.Amp, 18GA Alum)
N0230 m2020 (TMC3IN1_TARGET_TIP_VOLTS=90)
N0240 m2020 (TMC3IN1_TARGET_BAND_VOLTS=1.5)
N0250 m2020 (TMC3IN1_LINEAR_RESPONSE_BAND_VOLTS_ABOVE=25)
N0260 m2020 (TMC3IN1_LINEAR_RESPONSE_BAND_VOLTS_BELOW=15.2)
N0270 m2020 (ESS_DELAY_AFTER_ARC_OKAY=0.11)
N0280 G4 P0.1
N0290 G00 X0.7802 Y2.2514
N0300 (>>>Probe/Ref<<<)
N0310 G31.0 Z-10  F19.685
[b]N0320 G91 G00 Z0.0000 (Incremental move for switch offset distance comp)[/b]  <<<---- dives here
N0330 G90 (Back to absolute move mode)
N0340 G92 Z0.0 (Z work coordinate set to 0.0)
N0350 m2020 (HC_WORK_Z_ZEROED=1)
N0360 G4 P0.1
N0370 G00 Z0.1894
N0380 G4 P0.0011
N0390 M62P3 (>>>Torch ON<<<)

UPDATE:

So I went back to 4809 and run the same code did about 60 pierces and 0 issues.
With v5000 I got the weird random dive at least once every 3-5 pierces.

Over on the Warp9TD forum, Andy is suggesting there may be an issue with 5000 and is recommending 4809.