Good cuts, THC working good, Z loses machine coordinates?

[ShawnM]

As Tom pointed out, it’s still and open loop system. Acorn has no clue what the motor position is, it only knows where it SHOULD be.

That said, your drives are more than likely not seeing ALL the step pulses being sent to it. Lower the step rate to 100,000 and retest. Most of the Chinese clones can only handle 100000 step pulses regardless of what they put in a manual. The X and Y may also be missing pulses and not as noticeable as the Z is under more of a load trying to directly lift a load at the rate you have it set at. The X and Y have gear reduction so not the same as the Z axis. Lower the max speed of Z and lower the step rate to 100,000.

The drive isn't alarming out because if you send it 100,000 pulses from Acorn and it only SEES 90,000 pulses and the drive moves those same 90,000 pulses the drive does see anything wrong and does not alarm. Now the drive is not in the same position as Acorn thinks it is.

[BoenBuilt]

Thanks for the write up, I understand this 100%, but I don't think you guys are getting what I'm trying to explain.

If the drive was missing input pulses because it can't keep up with the step rate, the machine coordinates in Acorn would still be the same as the travel limits set in the wizard, being that the loop between Acorn and the drive is open~

From a fresh startup, machine homed, jog the Z up and the machine coordinates have a Z of 0. I do a test cut with THC enabled, jog the Z up to top of travel soft limit and now the Z machine coordinates DRO reads like +.009". Do another cut and jog the Z up and the DRO reads +.5. Continue to do cuts jogging the Z up between them and it'll keep getting larger. All of this while you can watch the nozzle keep a perfect cut height. If the drive was missing an inch worth of steps during a cut you'd know it.

How can the machine coordinates ever get outside of the travel limits set in the wizard if Acorn is just counting pulses and sending them to the ether?

As Tom pointed out, it’s still and open loop system. Acorn has no clue what the motor position is, it only knows where it SHOULD be.

That said, your drives are more than likely not seeing ALL the step pulses being sent to it. Lower the step rate to 100,000 and retest. Most of the Chinese clones can only handle 100000 step pulses regardless of what they put in a manual. The X and Y may also be missing pulses and not as noticeable as the Z is under more of a load trying to directly lift a load at the rate you have it set at. The X and Y have gear reduction so not the same as the Z axis. Lower the max speed of Z and lower the step rate to 100,000.

The drive isn't alarming out because if you send it 100,000 pulses from Acorn and it only SEES 90,000 pulses and the drive moves those same 90,000 pulses the drive does see anything wrong and does not alarm. Now the drive is not in the same position as Acorn thinks it is.

[Joey]

Maybe I missed something simple in the setup. We are using rtelligent closed loop steppers, which will reliably run the x and y at 1000 ipm. 10mm lead ballscrew on the Z. Grounding seems good, ohmic works reliably. Home switches on all axis, Y and N slaved, auto square works. Pid config trace for the divided voltage only shows a ripple from 0 to 3 counts.

Z axis coupler is a tight, diaphragm type with sleeve retainer loctite, it is not slipping, and even if it was machine coordinates for the Z would stop at 0 when you jog it all the way up.

After a few cuts with the THC enabled the machine coordinates for the Z get messed up, and you can't jog the Z back up to the top of the travel range (soft limits)

Are we not supposed to have soft limits on the Z? I have it set to +0, -7.5" just like I would on a router.

So far everything else is working great.


I reinstalled CNC12 today, set everything back up in the wizard, and ran a new torch calibration.

Thanks for any help.

I was incorrect about Z Soft Limits not being Affected by THC. I can replicate with my machine when THC is used the Z Machine coordinates will walk. For now no Z Soft Limits with THC.

Thanks for pointing this out.
Will have an Updated build for THC and Z Soft limits out soon.

[BoenBuilt]

Thank you!!

I'll get rid of the limits and drastically lower the Z fast jog rate for safety.

Maybe I missed something simple in the setup. We are using rtelligent closed loop steppers, which will reliably run the x and y at 1000 ipm. 10mm lead ballscrew on the Z. Grounding seems good, ohmic works reliably. Home switches on all axis, Y and N slaved, auto square works. Pid config trace for the divided voltage only shows a ripple from 0 to 3 counts.

Z axis coupler is a tight, diaphragm type with sleeve retainer loctite, it is not slipping, and even if it was machine coordinates for the Z would stop at 0 when you jog it all the way up.

After a few cuts with the THC enabled the machine coordinates for the Z get messed up, and you can't jog the Z back up to the top of the travel range (soft limits)

Are we not supposed to have soft limits on the Z? I have it set to +0, -7.5" just like I would on a router.

So far everything else is working great.


I reinstalled CNC12 today, set everything back up in the wizard, and ran a new torch calibration.

Thanks for any help.

I was incorrect about Z Soft Limits not being Affected by THC. I can replicate with my machine when THC is used the Z Machine coordinates will walk. For now no Z Soft Limits with THC.

Thanks for pointing this out.
Will have an Updated build for THC and Z Soft limits out soon.

[BoenBuilt]

Changed the steps per rev to 6400 on all drives, cranked acceleration to .25, set Z to 250ipm max, re did the torch calibration and it's cutting like a dream. Super smooth on the motions and THC.

Got Sheetcam figured out with the feedrate post processor for Centroid, starting to dial in the process.

Thanks again!

[Joey]

Great! Can you Create Report in the CNC12 Utility menu and throw a fresh report up so i can check out the THC Plots?

[BoenBuilt]

Great! Can you Create Report in the CNC12 Utility menu and throw a fresh report up so i can check out the THC Plots?

I will as soon as soon as I can. Is there anything I have to do to make it create a thc plot in the report or does it do it by default?

[Joey]

When you calibrate or have a Job interruption a Plot is automatically created.
You can turn on Plots for every cut in the CNC12 Utility menu.

[BoenBuilt]

I spoke too soon, finally had time today to try and run some proper parts. It won't cut for more than a few minutes with THC enabled until the Z machine coordinates start to walk again. Have to shut the entire machine down and reboot and re-home in-between parts.

Still on 5.02 and didn't yet try the custom macro. Was any of this addressed with the latest update? I didn't see any mention in the release notes.

Somethings not counting right when the Z is handed over to a THC PID loop and back.

We will try 5.04 next I guess, and if it still walks maybe go back to 4.xx? Must be a recent issue after looking through the forum.

When you calibrate or have a Job interruption a Plot is automatically created.
You can turn on Plots for every cut in the CNC12 Utility menu.

[Joey]

THC Affecting Z Machine coordinates and soft limits will be Addressed in the Next Build.

Your Calibration Plots Looked good from the First Report you posted. I could not find a calibration report in the latest report. I Think in only found 1 plot from your Latest report that had THC Cuts. The rest looked like just Torch float trips?

Did you turn enable Plots for every cut in the Utility Menu?
To enable Plot captures at every cut your UTILITY menu should like the attached snip.