MK2/S XYZ calibration failing - Other ideas from PINDA probe failing
Good evening,
Here is a brief description of the problem:
I have a i3 Mk2/s kit which I acquired in April from Prusa Research. The printer worked appropriately until approximately July. During the beginning of a print the mesh bed leveling failed, and after that I was unable to complete a successful XYZ calibration. I contacted Prusa Support where it was decided that I had a faulty PINDA probe and another one was sent. I have now replaced the PINDA probe and the problem is continuing. I am able to successfully do a Z calibration and a mesh bed leveling, however the XYZ calibration is not able to get past the initial 4 points in 3 calibrations. During the mesh bed leveling the probe falls dead center in the initial point (bottom left) circle, on every subsequent circle the probe is falling to the left of the center of the circle.
Attempted fixes:
A.) As stated, I replaced my PINDA probe, which made no difference. All of the tests with the pinda probe check out that the probe is working appropriately. Self-check is passed on all points
B.) I had been having a problem with the print head not triggering the x-axis end stop, which I thought might have been part of the problem. I was able to determine this was caused by the cable management off the print head (the bundle was ramming into the z-axis and the control box) where the motor could not generate the torque to overcome this. I moved the cables around so that this was no longer a problem and the x-axis ends stop is now triggering reliably.
C.) I had thought this may have been due to a skew problem on the y-axis, as I was getting a skew of 2 degrees for some reason. After moving the 10mm rods and nuts, and re-seating everything and remeasureing everything I was able to have the lines on the print bed parallel to the z-axis aluminum. XYZ calibration measured skew is now 0.12-0.15 degrees.
D.) I did a factory reset on the printer as well as update my software on the computer to most most updated versions of pronterface, slic3r, etc. From there I reflashed my printer to the most recent firmware and reran everything with the same results.
E.) I was able to successfully run a mesh bed leveling, which is where I found that the pinda probe was landing to the left side of the measurement circles on any points right of point 1 (bottom left). I ran an additional XYZ calibration from Pronterface and am copying the captured g-code from the calibration at the very bottoms of this thread.
In the end, it is appearing as though my printer is treating the print bed as though it is smaller than it actually is. If the calibration points are determined as a percentage of the total size of the print bed, and the print bed size is incorrect in software (even by 1cm in the x-axis and 2cm in the y-axis) then I think that might account for the problems that I am having, but I'm happy to hear any solutions from anyone else. Thank you
- Sean
[INFO] RECV: correction angles: -0.02409, -0.17438
[INFO] RECV: X vector, adjusted: 1.00000, 0.00042
[INFO] RECV: Y vector, adjusted: -0.00304, 1.00000
[INFO] RECV: center, adjusted: 0.46549, 0.80207
[INFO] RECV: echo:busy: processing
[INFO] RECV: Difference after correction:
[INFO] RECV: point #0measured: (112.60000, -1.54000); measured-corrected: (114.47145, -1.15001); target: (114.00000, -2.00000), error: 1.91
[INFO] RECV: point #1measured: (218.66000, 96.90000); measured-corrected: (215.17304, 96.89279); target: (215.00000, 96.00000), error: 3.49
[INFO] RECV: point #2measured: (112.81000, 194.87500); measured-corrected: (113.87492, 194.85066); target: (114.00000, 194.00000), error: 1.07
[INFO] RECV: point #3measured: (10.81000, 96.81501); measured-corrected: (12.17330, 96.80745); target: (12.00000, 96.00000), error: 1.36
[INFO] RECV: Iteration: 2
correction angles: -0.00693, -0.14770
X vector, adjusted: 1.00000, 0.00012
Y vector, adjusted: -0.00258, 1.00000
center, adjusted: 0.40595, 0.98862
echo:busy: processing
Difference after correction:
point #0measured: (112.59000, -1.46000); measured-corrected: (114.41108, -0.99760); target: (114.00000, -2.00000), error: 1.88
point #1measured: (218.60501, 96.86250); measured-corrected: (215.15841, 97.01492); target: (215.00000, 96.00000), error: 3.45
point #2measured: (112.85000, 195.36250); measured-corrected: (113.90581, 195.00299); target: (114.00000, 194.00000), error: 1.12
point #3measured: (10.82500, 96.83251); measured-corrected: (12.15847, 96.99037); target: (12.00000, 96.00000), error: 1.34
Iteration: 3
correction angles: -0.00224, -0.14006
X vector, adjusted: 1.00000, 0.00004
Y vector, adjusted: -0.00244, 1.00000
center, adjusted: 0.38501, 0.94938
echo:busy: processing
Difference after correction:
point #0measured: (112.58667, -1.37500); measured-corrected: (114.38989, -1.04617); target: (114.00000, -2.00000), error: 1.83
point #1measured: (218.58001, 96.85500); measured-corrected: (215.15031, 96.95808); target: (215.00000, 96.00000), error: 3.43
point #2measured: (112.83666, 195.20166); measured-corrected: (113.91078, 194.95442); target: (114.00000, 194.00000), error: 1.10
point #3measured: (10.87000, 96.84334); measured-corrected: (12.15034, 96.95014); target: (12.00000, 96.00000), error: 1.28
Re: MK2/S XYZ calibration failing - Other ideas from PINDA probe failing
Check on the belt-/chaindog beneath the bed. This can be shifted a bit forward and backwards.
Maybe it came somewhat loose, had that myself and big problems with the calibration.
Once properly seated and fastened to the back of the bed everyting was ok.
Second, I would recommend a complete firmware reset (read the FAQs), I seem to remember, that even
with a new calibration the firmware something like "remebers" the old points and is searching them there while
doing the new calibration. No idea whether this is correct, but a reset, too, for me did the trick.
Idid not check on your above stated measuremants, but if you suspect the firmware assuming a wrong bed size, I also would try
flashing the CORRECT FW anew.
Takes just a few minutes, and all the calibratin settings and live-z adjust are to be done, anyway.
Happy printing!
Re: MK2/S XYZ calibration failing - Other ideas from PINDA probe failing
Alexander, Thank you for your reply
The y-axis belt looks fine and the tension is appropriate. The attachment points for the y-axis belt also are secure.
One of the fixes I attempted was reflashing the firmware as well as doing a factory reset on the printer, neither one appeared to have any effect.
Re: MK2/S XYZ calibration failing - Other ideas from PINDA probe failing
I found the problem.
The problem was probably caused from my prior problem of the print head ramming the z-axis during homing. The set screw for the cam gear on the x-axis motor had become loose, not enough to allow the motor to turn freely, but enough so that there was 2-3mm of play in the motor movement that was translated to the belt. As soon as I corrected that problem, the printer was immediately able to complete the XYZ calibration successfully.
Hopefully this helps someone else in the future as well.
-Sean