Z differences when resuming long programs

Ever since I started trying to do very long programs that required multiple days (20-80 hours total), I’ve noticed problems with Z-height changing after resuming a file. When it happens, which is almost always, It’s always significantly lower than when I stopped it, sometimes nearly 1/16”.

I am running the Elite Journeyman with stiffy, Masso, QCW, Redline spindle, and using the tool setter to automatically measure the tool on startup.

As the file runs, over 3-4 hours the Z height seems to raise up very slowly, to the point that it’s not cutting all the way through the material any more, and you can see a very gradual height gradient in the material.

I have the braking motor for the Z axis, but I’ve noticed that it always drops by maybe 3/16” every time the e-stop is hit or the machine is turned off. I figured this would be compensated for by the tool setter, so I ignored it, but maybe that’s part of the problem.

I first thought humidity / temp changes might be an issue, so I put a dehumidifier and AC unit in the shop, and started to keep it very consistent, and also swapped out my MDF spoilboard for expanded PVC, but none of this has had any effect on the issue at all. Both spoilboard and material are always surfaced before running the files as well.

Results of most recent attempt: https://www.youtube.com/shorts/iwlNHznV-8c

After installing the PVC I started this long program, ran it for 6 hours, by the end of which it wasn’t cutting through the material anymore (but weirdly, only at the edges..). Then resumed it 6 days later and the height was back down, and cutting through again. After 4 hours it wasn’t cutting through. The next day I resumed, had another height drop, though not as significant, and ran it almost until the end.

Then I tried resuming the file immediately after shutting down to see what happened, and the height dropped right away, so now I’m pretty thoroughly confused.

Any ideas? Just replace the Z motor?

1 Like

Is it always on one side ? That picture sure looks like you have an issue with your X axis.

Can you reverse the toolpath from right to left and see if it does the same thing?

Are you keeping the machine on for this carve or powering off and using the jump to line feature?

Do you have a breaking motor on your z?

What firmware is on your masso?

What are your masso dip switches set to?

1 Like

That’s just where the program starts near X/Y 0, and it continues at a 45 degree angle from left to right. It seems to be a function of time, not axis so wherever it begins is always where it’s the deepest. The height drop always occurs exactly wherever I start it.

I could try starting with the origin on the opposite side, but I’m pretty sure I would just get a reversed gradient.

1 Like

I can try that, but I think it will just be a gradient in the opposite direction. I have the braking motor for z. It drops a bit on e-stop or power off, but I thought the tool setter would compensate for that.

I’ll have to check on the dip switches, but everything was set per the video instructions, and I never seem to have any problems with short files.

I think my firmware is 5.10.

I always power off, reload the program, and use jump to line when resuming, usually picking 10-20 lines prior to where it stopped.

1 Like

Do you also home on bootup?

Ensure you are on 5.12 (latest), dip switch 1,4,5 on/up and 2,3,6 down off.

Then reload the defaults here:

That way we’ll be certain you have everything is as it ships from the factory today and has a good baseline to start with.

1 Like

I do home on bootup.

I am on 5.10 which I was using since I didn’t think 5.12 was good with the toolchanger, but since I’m not using that anymore, I will go ahead and upgrade to 5.12 tomorrow when I have some time in the shop, and will reload the defaults then too.

I checked my Z dip switches, and they were set according to this part of the motor resolution video (1, 4 off, 2,3,5,6 on) is that obsolete now?

1 Like

Great! Yes, you’ll want to have them like this: dip switch 1,4,5 on/up and 2,3,6 down off.

1 Like

Ok that’s for all motors right?

Correct, all motors should be the same

1 Like
2 Likes

Ok thanks. I’ll set everything up tomorrow, and try again from there.

2 Likes

Ensure you are on 5.12 (latest), dip switch 1,4,5 on/up and 2,3,6 down off.

Ok, upgraded to 5.12. Just wanted to confirm per the link you provided, that the dip switches should actually be 1,4,5 off/down, and 2,3,6 on/up. Is that correct?

1,4,5 off/down, and 2,3,6 on/up

This is correct.

2 Likes

Ok great, thanks. I’ll set those and load the config, and then try a file going from right to left and see what happens.

1 Like

Ok, I did everything suggested and made a basic test tile to run from right to left. Part of why this is so tricky is that it only seems to be detectable after the machine has been running for around 4 hours or so continuously. Did this test last night, stopping and resuming 6 times over the course of about 3.5 hours, and noticed almost no height change at any of the stopping points. One slight change in the middle at about 1hr 38min run time, but only about 1/32” or less. Very minor, and no detectable changes at any of the other points where I stopped and resumed.

Then today I ran a similar test, but let it run for a little over 4 hours before I stopped and resumed. And here we have the same problem I’ve been dealing with, a significant drop, only this time in the opposite direction, as I expected from a right to left toolpath.

Any ideas?

Do you use a UPS in your operation?
Is it possible you are getting a voltage drop from your utility provider?

1 Like

No UPS. Don’t think it’s the utilities since it’s constantly repeatable.

1 Like

Here’s some dial indicator action. Seems very conclusive as to what’s going on. Should have tried this a long time ago…

Looks like it’s constantly slowly gaining height in z as the program goes along until stopping and restarting the program which initiates a new tool touchoff which brings it right back down to the actual zero.

1 Like

What are the indicator units?