How does CM handle retract height at the beginning and end of a job vs during the gcode file execution? I had a few crashes when machining a 1.875" thick stock on top of a 0.7" thick MDF board. Heights given because the Z-stage could be jogged up and clear the stock just fine manually, and also cleared just fine during the actual job. You can see two scars here where the bit cut the stock at the beginning coming from the tool change position, and near the back when it exited at the end of the job.
For some reason in that condition it does not retract high enough - is this some sort of soft limit? I saw some similar questions here but none exactly describing this situation. Is the only “fix” to move my router higher? I’m unclear if CM should have enough information to avoid this scenario (it knows the tool length and zero height, but not if the zero is the actual top of the stock).
It should retract to the safety/retract height set in the file — please post the .c2d file, generated G-Code, step-by-step notes on how you are securing your stock and setting zero relative to it, and notes on how you are managing any tool changes.
Could the machine have bottomed out against the bottom stop and lost steps?
Gcode and CC file attached. I am securing my stock with double sided foam tape at a good [0,0]. I let the BitSetter do its probing routine, and only changed the bit using the Tool Change prompt. I zeroed the Z manually in CM by touching off the top of the stock, then retracted manually and parked the head away from the stock. I then started the job.
The machine prompted for tool change (none needed) and then re-measured the tool. It went back to tool change position like normal, started the router (via BitRunner), and then rapided to the first path, crashing along the way. Everything else during the job looked good, it cut to proper depths, no grinding or bad stepper sounds, and no crashes when rapiding between pockets.
I am near the height limits as I had about 3-4mm of clearance between the collet nut and my stock top at the deepest operation (measured before starting the job, and confirmed visually during).
I raised my router about 5mm in the holder without changing the tool length, and was able to complete multiple jobs with the same stock thickness. I’m still assuming it was hitting some soft limit since it still had +Z travel available (before pressing homing switch). @WillAdams Let me know if anything jumped out from the files.
Your stock thickness is 1.875" and retract height 0.25" and you’re cutting as deeply as 1.6" — was there 1.6" of negative Z movement available before bottoming out?
If I’m understanding your reference frame properly…
Yes, all the motion took place near the top of the Z-stage, it was not close to sliding off at the bottom, nor was the bit close to hitting the spoilboard.
