Mk3 Z Calibration Fail & Workaround
This may not be the solution for everyone with this problem, but I share my experience in case it's helpful to some.
A long time S3D user, I started using Prusa Slicer (NO this is not a prusa slicer issue!), in order to do what I could about making my bed as level as possible (using this post from the knowledge base https://help.prusa3d.com/article/TPGip0OmaP-bed-level-correction) So the Prusa Slicer told me that there was a firmware update for my MK3 (my machine shipped with 3.5.2). So I installed 3.7.1 in order to take advantage of the 7x7 bed sensing. After struggling with that for a few days I decided to recalibrate my Z axis... and that's when the problem started.
Z calibration was unable to calibrate. The extruder moved up to the top of the frame... I was reminded to remove the metal plate and to put a piece of paper on the bed at which point the extruder would descend ultimately pressing against the plate and giving me a "Calibration Fail". Checking online most people seemed to suggest the problem was probably the PINDA probe, and while I readjusted and readjusted the probe there was no joy. I checked for obstructions and my head being too low and could not change the outcome.
Someone suggested I do a clean install of the latest update but installing it, doing a factory reset ( https://help.prusa3d.com/article/SYvbQ66IXF-factory-reset) and then re-install the latest update. This had no effect.
Given that the only thing different in the behaviour of my machine (from when it worked until it was broken) was the firmware I decided to downgrade to the 3.5.2 firmware. So following the factory reset clean install procedure I installed the 3.5.2 firmware that I had downloaded from the Prusa site.
No joy.
Finally, I just did a factory reset and nothing else... no firmware modification... with the machine booting up with what I presume was the 3.5.2 firmware that is in the ROM of the controller.
Success!
So goodness knows what the difference is between the 3.5.2 hard coded on my machine and what's on the Prusa Website... or what the problem with the updater is, or perhaps what the bug in the 3.7.1 firmware is... but after a couple of very frustrating days my printer is working again and I'm not highly motivated to update to the latest firmware but I wanted to pass on this workaround in case it matches your symptoms.
RE: Mk3 Z Calibration Fail & Workaround
Your symptom seems similar to what just happened to me but for me it was not resulting in the 3.7.1 upgrade. I’ve been running 3.7.1 for a while. However, I just had a print failure where the print got unstuck from the bed. I was not there so cannot comment on what exactly happened. I found my print behind the printer. But when I tried to restart the print, when homing, the nozzle hit the bed (and complained about possible debris). Retried Calibration and hit the bed again. The Pinda had moved, so reset the proper height with the ziptie trick. Still hitting the bed.
If I raise the Z axis and place an Allen key, or anything metal, the sensor triggers. So currently I’ll also inclined to say my Pinda is out of whack...
Louis
RE: Mk3 Z Calibration Fail & Workaround
Anything that keeps the X-Axis from homing properly will cause this. Usually the offending culprit is rotation of the cable ties on the rat tail coming out of the extruder. They hit the EINSY case and stop the X-Axis from homing properly.
Other culprits are nice looking fan grills on the cooling fan; or wrong screws holding the fan. Anything too "tall" here will also keep the X-Axis from homing.
RE: Mk3 Z Calibration Fail & Workaround
No issues with the X-axis homing. Not sure either how that would prevent the Z-axis from homing...
RE: Mk3 Z Calibration Fail & Workaround
No issues with the X-axis homing. Not sure either how that would prevent the Z-axis from homing...
Very simple, when calibrating Z the PINDA must sense metal. So it positions the extruder X steps off of "home" to align with metal embedded into the heat bed ... so if X home is not where it is supposed to be, Z cal fails. But you also mentioned it is FW dependent... missed that first time around.
As for why flashing new firmware didn't work, I suspect a deranged value somewhere in EEPROM that needs to be cleared using FACTORY RESET with DATA CLEAR.
If you don't select DATA CLEAR when doing the reset, corrupt EEPROM stays with the firmware upgrade; and corrupt EEPROM data has caused all sorts of "this cal doesn't work" issues. Just search the forums to see them all.