another case of the lead-out divot

[bLouChipBlacksmith]

I didn't want to hijack another thread so I hope ya'll don't mind that I'm making another for a somewhat common problem.

I definitely learned from the referenced thread where Jim Colt was cited on trying perpendicular lead-in and lead-out. A modified version of that technique made the most improvement in my case. I can still use some improvement but I have minimized the divot sufficiently for now.

12gaHR, 45amps, 1900mm/min, 3.7mm PH, 1.5mm CH, 72psi, UPM105 torch w PMX45 nozzle (new), PrimeWeld Cut60.
grbl 1.1i, Sheetcam w custom grbl scpost.

This was the problem that triggered a 1/2 day of research, testing, and tweaking...

Test case #12 I believe is the most improved divot, best case.

So then I proceeded to cut some 'divot' tools for an upcoming golf trip with some buddies.

I'm ok with the result, but I'd like to get rid of the dang lead-out divot if that's possible.
Is it possible without high end CNC plasma controller as Mr. Colt described ?

Thanks in advance for any thoughts on improvement.

[adbuch]

move lead in.jpg (59.71 KiB) Viewed 192 times

[robertspark]

don't use grbl

M5 will not occur in sync with motion... ie there will be a delay when the machine is decelerating to a stop and then run M5

you can try a lead out.... but it won't work with single line letters... and if you have THC you will need to disable it... which grbl won't do in sync with motion (is it will decelerate to a stop, then toggle the THC, then accelerate, complete the cut and and lead out, decelerate to a stop then run M5)


you may be able to set the plasma up as a laser in grbl as that will toggle outputs in sync with motion (m64/m65 gcodes with linuxcnc or mach4 or M10/M11 with mach3 and uccnc (controller dependant with mach3 and 4 whether this functionality is supported)

[robertspark]

this may be of help to try

https://docs.lightburnsoftware.com/Comm ... ange%20RPM.

[bLouChipBlacksmith]

@adbunch- agreed. I should practice that more often for sure. Thanks for the tip.

@robertspark- "don't use grbl". I'm beginning to realize that.
"M5 will not occur in synch...". Understood.
"try a lead out..." already in use in the problem case, experimented with several styles and lengths; understood about single line letters, not the problem in this case.
"set the plasma up as a laser in grbl...". grbl laser mode $32=1 already enabled, and already using spindle power "S" word to enable/disable THC with continuous motion. But M5 as you noted still does not drop spindle enable until motion stops, M7, M8/M9 same behavior. Using M7/M9 also in THC control.
re. linuxcnc, m64/65 codes, mach4, uccnc etc. - I wasn't aware of continuous motion synchronization with those control stmts, good info to know, thanks.

All good feedback, thanks.