Offset measurement goes bad every few prints about klipper_z_calibration

[TitusLabs]

Yes, it should be "fine, thats it"...

First, I would change the order to: QGL, G28, clean nozzle, calibrate_z

But, it sounds that there is something wrong in your system. You need to check everything! Is there a loose pulley on the z drives? Is your probe ok? Do you use an original Omron switch? Are the probing positions correct - especially on the switch? Do you probe slowly multiple time and use median as samples result? You can also use PROBE_ACCURACY and PROBE_Z_ACCURACY...

[Silikowski]

my z idlers are fine as far as i can see.
probe accuracy tests seems ok too:

probe accuracy results: maximum 6.988750, minimum 6.972500, range 0.016250, average 6.978750, median 6.977500, standard deviation 0.006175

i dont have an original omron switch cause they were not in stock somewhere.

probing positions are fine. i tuned them by hand and checked the probing Multiple times.

im gonna change my starr order as proposed and get back to you

[Silikowski]

Well i replaced my start_print macro with the new order and followed a few hints from the voron discord. it looks better now (so far) - im gonna let it run a while and see if stays fine consistently