Help wiring THC

[beefy]

Well, let's forget the THC trying to set the voltage for now.

First you need to sort out this initial torch height problem. Do a simple cut that reproduces this Z offset problem but leave the cut plate in position on the table after it's cut.

A simple way to check how flat your plate is relative to the XY rails is get a piece of flat/square bar (anything flat will do). Put it on the plate where you zeroed the first pierce point (not on top of any slag) and zero the torch on top of the bar. Remove the bar. Now jog the torch to another location and put the bar underneath. On a perfectly accurately made table without slag on the slats, and with a flat steel sheet that has not warped with the cutting, you should be able to slide the bar under the torch and confirm your Z zero point is still the same. If you can't get the bar under, the plate is higher at that point. If the bar is loose under the torch, the plate is lower at that point. You can see exactly how much difference by sliding a thinner or thicker bar underneath and measuring the difference. Does anything correspond to the difference in the Z zero position your are experiencing.

[Spoolnms]

My slats arent perfect but nowhere close to that bad. That 1/4" off was with z in about 4" to the right of where the cut started. My slats arent out 1/4" within 4" lol. And when checking z it was off. At first i thought my z was losing steps but when testing and moving all around up down, side to side, torch on and off, when i zeroed it it was right where i had set it.

Only thing i can think of is z offset is being over written somehow somewhere.

[beefy]

OK, interesting.

Off the top of my head, 3 things I can think of that would alter Z zero position:
1). What we've just talked about, the actual height of the plate, but now based on what you've said that doesn't appear to be the cause.
2). Lost steps
3). Gcode is resetting Z zero point.

UCCNC should not be altering Z offset. Once you manually zero Z, the position in UCCNC should be maintained, i.e. whether it's jogging, or THC up/down moves, or gcode Z moves, UCCNC should keep track of EXACTLY where Z is.

Is this problem repeatable, or does this changing offset vary. If it's accurately repeatable, it seems something is commanding this change. So use the same file and do this cut 2 or 3 times and see if the stuffed up offset is the same or different each time. Have you checked the gcode to see it's nothing wrong in there.

[beefy]

And when checking z it was off. At first i thought my z was losing steps but when testing and moving all around up down, side to side, torch on and off, when i zeroed it it was right where i had set it.

Could you explain this one a bit better. Do you mean:

1). You manually zeroed Z at start of cut at top of material.
2). You manually moved/jogged X, Y and Z all over the place.
3). You jogged back to the original start point at 1). and moved Z to top of material and the Z showed zero in DRO, thus confirming no lost steps.

It helps if you are very specific when describing what you did. It's more open to interpretation when you say things like, "when testing and moving all around up down, side to side, torch on and off". I mean how did you test exactly, how did you move it around (manual jogging or running gcode). The fact you say "torch on and off" implies it was gcode doing the moves, but that is me taking a calculated guess. Likewise when you say "when i zeroed it it was right where i had set it", do you mean you jogged the torch back to the start point and checked Z DRO reading was as you originally set it at that location.

[Spoolnms]

Sorry ill rephrase. Upon zeroing the z offset to the top of work piece, i manually mjogged the machine all over the place, as well as manually turning the torch on/off (thinking maybe interference) and the z offset remained at zero where i had set it. Ill check the gcode but the post processor im using is mach3, wouldnt think it would reset zero but ill double check.

Ill double check and do what you recommended and cut a piece then remeasure zero and report back.

[GaryLa]

Is it possible the "Write offset on homing" in the Configuration Menu under Z Axis is doing this?

[Spoolnms]

I doubt it, as I do not home the machine (only at first machine startup)

[GaryLa]

I doubt it, as I do not home the machine (only at first machine startup)

A friend of mine kept getting his Z set to the value in that field. Perhaps he was homing. But I'd suggest perhaps you clear that value, or check to see if it's the same value as the issue you're having.

[beefy]

Post the gcode file that gives this issue so we can check it's nothing in that file that is doing it.

[jcoldon]

what you want is called IHS initial height sence if you don't want a torch slide then get an ohmic senc unit like the one from robot 3t
if you don't want to use any thing then lower the torch till top of steel then zero the z if your cut settings are correct if work
I made a uccnc post processor for uccnc plasma for a torch slide and ohmic input and smart probe and thc on and off features it on my Face book site cnc plasma table builders you can down load it