Hi after zeroing out my MPCNC and running the attached gcode (Ryans lowrider y plate) the machine starts cutting air above the work surface. The final cut into the workpiece is about 5.25mm of the expected 13mm cut.
I’ve tested the machine using manual controls and the z movement when used manually is correct.
I’m using ESTLCAM with Ryans recommended settings and a standard 1/8 bit also with the settings from the tutorial.
Any ideas?
Thanks!
Y_plate2 13mm (2).gcode (237.0 KB)
Are you zeroing the Z at the top of the workpiece?
Does your Z move 10mm when you command a 10mm move?
Are you skipping steps?
The gcode looks about right. It starts at Z=0 and steps down -3mm, -6mm, etc. The speed is a bit high, 630mm/min, which is over 10mm/s. That’s a bit high for a plunge. 3mm/s would be safer.
Yes zero’d at top of piece. Manual movement via display moves x in proper amount. No skipped steps as its cutting air at the time of the start.
If I rerun the script I get repeatable results always at the same depths. Its like there’s an 8mm offset set someplace.
My ESTLCAM settings are attached
Estlcam doesn’t matter if the gcode is right. The gcode shows this:
G01 Z-3.0000 F635
If that isn’t working, it isn’t estlcam or your gcodes fault. Did you try the test crown premade gcode?
You can do this separately. Put the bit on the top of the work. Send G92 Z0. Then move with G1 Z10 F635 to move up and G1 Z0 F635 to move down and it should stop right where you set the zero.
Jeff thanks for sticking with me.
Crown printed ok not sure about the depth but it was probably close. Ran this code:
G92 Z0
G90
G1 Z10 F500
G1 Z-10 F500
G1 Z0 F500
G91
G0 Z10 F500
G1 Z-10 F500
G1 Z0 F500
Which ran properly both for absolute and relative positioning. Up 10mm and down 10mm worked and machine returned to z0 properly.
I guess this is a good thing which means the issue lives in my Gcode/ESTLCAM and not in firmware / hardware right?
I am still suspect of the set up. How are you zeroing out the Z when you set up the job?
Tried 3 ways
Marlin Custom/Reset Coordinates
Machine Reset at top of workpiece
Through G92 Z0 in the gcode
It must be ghosts then. 
That doesn’t make much sense. It isn’t magic. If you can send G commands and it works manually, then doing that same thing in a file is identical.
I didn’t read the whole file, but at least the first cut looks fine. There isn’t anything weird, except the high Z plunge.
If I’m seeing things right it looks like the manual tests were at a speed of 500 and the gcode is calling for 635. That difference could be enough to cause missed steps. Try the manual test again at 635 and see if the results are the same.
I’ll try messing with the plunge rate tomorrow, I doubt its missing steps before it touches wood but I’m also not sure why my plunge is set so aggressive so I’ll fix and try it thanks.
Missed steps are not solely due to too much resistance to the cutter moving, they can occur if you try to move a stepper motor too fast - even without a load…this is what Tom is getting at.
Thanks I wasn’t aware of that, I’ve reduced the plunge speed of the bit. I was modifying the Fast Move speed previously.
Unfortunately this had no effect other than a slower plunge. Tool is still cutting air over top of the workpiece.
Going to try and simplify and cut some basic squares and see if I can isolate the issue.
So I cut out 4 2x2 squares at the following depths 5,10,15, 20MM
All cut to 5mm deep. The weird thing is that the tool rose 5mm with each cut so the
5mm started at 0 and cut to -5
10mm started at 5 and cut to -5
15 started at 10 and cut to -5
20 started at 15 and cut to 5
Z axis origin is set to workpiece top.
2x2.gcode (49.0 KB)
This is the gcode
The g-code looks fine to me including the feedrates. Have you checked the grub screws on the connector between the stepper and the lead screw? It is often the culprit in lost steps on the Z axis, but I would not expect it to be this regular. Have you check for other mechanical issues on the Z axis…the mount hitting at a certain height or the Z axis binding on any part of its travel? Have you verified that the length of the end mill remains constant?
If you have a display, look at the position reading after each square is cut. Without a display you can use the M114 g-code to get the position. If the position the machine thinks it is at does not match the physical position, then you are looking at a physical cause for the problem.
I also suggest putting a G92 X0 Y0 Z0 at the top of your g-code file. Given your description, it is unlikely to solve your problem, but it will eliminate some potential causes.
Yeah I don’t think it’s lost steps the measurements are too regular… I’m thinking about flashing the controller next maybe there is something going on there
@jeffeb3 and others …
I hang my head in shame… After pouring over settings and gcode for hours and learning much about my MPCNC I found the problem…
The machine was simply bottoming out. I recently switched to a thinner 1/4 spoil board to save weight which I thought was a brilliant idea; totally forgetting that I had designed the machine to bottom out at the surface of a 3/4 spoil board. I normally don’t do through cuts so this is the first time I’ve noticed.
The machine was bottoming out and losing steps then raising back up. My manual tests were done with a higher Z-0 because I didn’t want to unglue the workpiece thus they acted properly.
Anyway thank you all for your help. I really appreciate all of your efforts.
Thank you for sharing the confirmed problem. So many times we never learn the real problem and how it was solved.
Too late…I was wrong anyway… 
Glad you sorted it.