UM3 X/Y Calibration pattern repeatability

[Matioupi 1]

Hello,

despite the fact that I need to manually push the filament with current 5.2.8 firmware on my UM3, I just tried the following :
doing 2 X/Y calibration patterns in a row. I mean : I did one calibration pattern, up to the end, entering the X/Y offset values, and immediatly did a second one (after cleaning the glass plate).
I ended up with 2 very different set of values :
X:5 / Y -7 the first time

X : -5 / Y : - 17 the second time
What could be wrong here ? Nothing changd between both prints (same material, same printcore, a few minutes between both prints)

Edited May 6, 2019 by Matioupi

[Matioupi 1]

I went a little further with testing today :

Initial state was X : 5 Y : 17
I ran a first calibration test :
result : X : 7 Y : 13

I ran a second calibration test :

result X : -5 Y : -18
While going back to the menu to check if the values where properly taken into account, I saw in the "change offset" that signs looks like they were not taken into accoun (or badly).
At this stage, I went for a full printer reset (factory + cura connect)
After reset, I manually entered the last test results : -5 / -18
I ran a third calibration test :
result :
X : 7 Y : 13
I checked though change offset, this time, the correct values are displayed.

I ran a forth calibration test
result : X : -5 Y : -18
I ran a fifth calibration test :
result X : 7 Y : 13

At this stage, it seems clear that there is something wrong as the calibration oscilates between 2 set of values, so I went for the following :

I manually changed the offsets to 0 / 0 and ran a sixth calibration

result is X : 2 Y : -5
 

I can see that 2 = 7 -5 and -5 = 13 - 18

At this stage, I would think that there is some kind of bug in the firmware... something like "relative vs. absolute" offset management.

 

[CarloK 85]

There was a bug where the XY calibration pattern would apply the last set offsets to the patterns where it should have reset them to zero during the pattern printing. This was fixed in the v5.2.11 release of today.

[Matioupi 1]

thank you for the confirmation... i think this was the biggest bug for me in 5.2.8

[ultiarjan 1,184]

8 hours ago, CarloK said:

There was a bug where the XY calibration pattern would apply the last set offsets to the patterns where it should have reset them to zero during the pattern printing. This was fixed in the v5.2.11 release of today.

 

Seems the v5.2.11 release notes are incomplete then...

https://ultimaker.com/en/ultimaker-firmware