Firmware 3.14.1 - constant Z-Calibration Failures
I had this Mk3S printer that I haven't used for a year, and the other days I took it from storage and put it to test - all worked well, but it required some calibration. But as I was at it, and i saw there is a new fw release, I decided to upgrade - and boy I regretted it.
Now, no matter what I do, the Z-Calibration fails as soon as the print head descends above the bed.
I tried factory reset and full recalibration - everything passes successfully (even the XYZ Calibration), but the Z-Calibration fails constantly and I have no idea what to do next.
Please help me get out of this pitfall of despair. Thanks.
RE: Firmware 3.14.1 - constant Z-Calibration Failures
Check, whether the x-axis is parallel to the bed. I suspect the upper ends or the bed has become misaligned (y-rods). I assume, that bed and sheet are in good condition and clean. The trapezoid nuts should also run easily. when z is low, you can loosen the screws of the nuts a little move the nuts into the middle and retighten them. The plastic caps at the low end of the z-spindles must not touch the motor.
If you have the MK3S (without +) you probably still have the PINDA with the integrated thermistor, not the so called Super-PINDA. Please verify that you flash the correct firmware. The fact that the XYZ-Calibration suceeds indicates you probably have indeed flashed the correct firmware.
You might downgrade to 3.11 or 3.12 and try the z-calibration. Just to cross check that it is not a hardware issue.
RE: Firmware 3.14.1 - constant Z-Calibration Failures
Check, whether the x-axis is parallel to the bed. I suspect the upper ends or the bed has become misaligned (y-rods). I assume, that bed and sheet are in good condition and clean. The trapezoid nuts should also run easily. when z is low, you can loosen the screws of the nuts a little move the nuts into the middle and retighten them. The plastic caps at the low end of the z-spindles must not touch the motor.
If you have the MK3S (without +) you probably still have the PINDA with the integrated thermistor, not the so called Super-PINDA. Please verify that you flash the correct firmware. The fact that the XYZ-Calibration suceeds indicates you probably have indeed flashed the correct firmware.
You might downgrade to 3.11 or 3.12 and try the z-calibration. Just to cross check that it is not a hardware issue.
Thank you. I checked the cable on the probe (next to the plug) - 3 wires, and a QR code, which scanned gave me the number "5185-03164087111040745" however I could not make up any other info, and there is no label on the probe itself.
The frustrating thing is I printed with it the day before I did the fw update - and it worked well. Do you know if the tolerances are tighter in the 3.14.1 version, or is there anything else I should be aware of?
The general geometry is pretty much aligned (skew is well below threshold), the threaded rods clean, I lubricated them for good measure, I can easily move the gantry by turning the rods by hand...
RE: Firmware 3.14.1 - constant Z-Calibration Failures
Please try the old firmware as I recommended before and report back whether it works.
With 3 wires you have the Super-PINDA. You need firmware version 3.9.2 or newer. Its the same firmware for MK3S and MK3S+. A thermal model protection was introduced with firmware 3.12.0. On some printers the temperature tolerances seem somewhat tight so that a "thermal anomaly" warning is emitted even when the printer is fine. One of our printers does this as well on some jobs with higher temperatures even though we did the thermal model calibration.
What did you touch mechanically wise? Something at the z-axis, the y-axis, the height of the PINDA, the extruder, the upper end-stops, zip ties? Do you mean xy skew? For z-calibration xy and yz are pretty much (linear) independent. If the x-axis is skewed with respect to the bed height that would be a thing to correct. The bed has to be parallel to the upper frame because z-calibration bumps into the upper ends. The printer might even do this mid-print when it detected a crash. Therefore, correcting xz skew by turning a spindle by hand is not a good idea. You have to redo the z-calibration instead. If you touched the heights of the y-rods your bed might be skewed vertically.
The standard z-axis-tops do not offer any adjustments. Usually that is good enough. Nevertheless I added a vertical m5-thread to each so that I can fine tune the one that is higher. See attached file. Please do not descend the stop more than absolutely necessary, less than 1 mm. Otherwise the firmware might complain about a to short z-range. The thread is cut tight and I have a counter nut on top of each, so that it pulls the screw into the upper flancs of the thread.
Please note that pulling at the extruder cables or having another printer jerking next to this one can disturb it --- especially any calibration.