Problem found: rod length was incorrect. The config-override contained an incorrect length (+8%). Since I use 6 factor calibration this value isn't adjusted. It was probably a side effect of a 7 or 9 factor calibration experiment.

What I learned: On a delta, if both you x and y are off by the same amount but your z calibration is fine. Take a look at the L parameter of the command M665 in BOTH config and config-override.

I suddenly have the same problem with my Delta. Z is fine but X and Y are 5% too small. I noticed it today running 1.19 and is still a problem running 1.20 firmware. Microsteps are set correctly.

What could be the underlying cause of this? De printer calibrates using 6 factor calibration and 10 points to 'after 0.038'.

I also checked the gcode, it uses the expected X and Y dimensions.

I found my error… I changed the steps from 88.89 to 88,8889 with a decimal comma. This truncated the fractional part resulting in 1% lower setting than intended.

Recently I noticed my prints came out a bit too small (It used to be spot on). I have recalibrated the z-probe, delta, remeasured my rods etc. But to no avail. Next I measured the z-travel of the carriages. I noticed the carriages travel consistently only 99mm when I move the head down 100mm. And 49.5mm when moving 50mm down. I cannot measure beyond 100mm but when I move the head down 200mm en up 100mm again it is again at 99mm from the starting position.

My printer uses 1/16th steps, 0.9 degrees motors, GT2 belt and 36 teeth pulleys. I have always used 88.89 steps/mm (validated using the on-line prusa calculator). I am running the latest firmware 1.19.2 (2017-09-01).

What could be wrong?