3.3rc1 Polar Kinematic issue.

Phospherus

@dc42 This is an update to the Polar issue we spoke of in the other thread.

I am getting a waviness on straight lines with the 3.3 Betas and the RC1, Firmware 3.2.2 is the newest that I do not see an issue.

The following is with the same Gcode file.

This is with the latest 3.3rc1 that DC42 linked to on Dropbox.

This is 3.2.2 just to show that it is not a mechanical issue.

These were done back to back for comparison and on Genuine Duet Ethernet 2.

dc42

@phospherus please send M669 without parameters to check the segmentation settings.

Phospherus

@dc42

M669
Kinematics is Polar with radius 0.0 to 120.0mm, homed radius 0.3mm, segments/sec 100, min. segment length 0.20

I checked both firmware versions this output above is the same.

Phospherus

@dc42

Are these ok for segments/sec or do you see an issue?

dc42

@phospherus thos segmentation value look OK to me. In the first image you posted, what is the approximate length of each wave; and what was the printing speed? It looks to me that the firmware was not using the configured segmentation settings but was using much larger segments instead.

Did you try 3.3beta1, and if so, did that version show the same problem as the later betas?

dc42

@phospherus, I think I have found the bug. I will fix it in 3.3RC2. Meanwhile, I think a workaround would be to send M669 T0.2 (later than your main M669 command) or alternatively add parameter T0.2 to your existing M669 command.

Phospherus

@dc42

Thank you for the info. Will reload 3.3 RC1 and try the fix and will let you know.

Phospherus

@dc42

Confirmed that issue is fixed with the provided work around. Thank you