I printed a 30mm thin walled test cube with my new white PLA via Octoprint. As it came out perfect (digital calipers confirmed the filament was 1.72mm not the 1.75 listed) I then went to print two supports for my re-spooling design. I thought I could trust it to get on with it. Monitoring from outside the office OctoPi showed lots of spaghetti on the bed not a proper print.
Back in the office I cancelled and restarted the print. It homed, armed the z-probe then started probing in midair rather than touching the glass! I cancelled the print, powered everything off, restarted and it did the same. I changed from the Pi to PC and Pronterface, it did it also.
Switching to terminal entry G28 homes okay, G29 probes midair, so it’s nothing the print host is doing. Yet once it’s done jumping around way above the bed it moves to the correct x,y,z location to do the z-probe reset correctly.
My Marlin Configuration.h has: #define MANUAL_Z_HOME_POS 232.3
Yet now when I home it the LCD and M114 report Z is around 410!
If I home it, then try to use Pronterface to move Z down 10mm I get lots of nasty noise and it looks like it’s trying to move up not down.
I can’t understand how it thinks the height is 400+, tries to probe or print way above the bed but is able to get the probe arm and reset movements spot on every time. Any ideas please?
I’d check wiring (due to the nasty noise comment and it moving the incorrect direction), and if I had a sane config file, I’d reflash the firmware.
The nasty noise can be caused when a wire to one of the stepper motors gets cut, or is impartially connected (or even disconnected) - causing the stepper driver to only drive 3 out of 4 coils.
Especially if you’re getting a weird Z height like that. Also, how fast is this all going? If it goes too fast you might be skipping steps.
That fixed the bad move down from home. But not the probing in the air. That I’ve traced to a bad z-probe microswitch. I went to take the switch, pin and spring off and the arm fell off!!
If you see my other post with photo it was a bad combination. The pin wasn’t 100% straight, the microswitch falling apart and the pin would stick, half deployed half not. Shortly after starting the in-air probing I guess the vibration would make it drop the last few mm, but by then it was too late, the bed height was recorded many cm too high.
