Hi! I'm using the Duet system with a linear stepper motor, so right off the bat - I know this is outside of your expected operational envelope. Any help you can provide, would be fantastic; the toolset has been exceptional in many ways thus far, and I would love to continue using it on this project.
The linear stepper motor in initial tests was giving me poor repeatability, 0.1mm+ in open loop testing. I consequently added a quadrature linear encoder from Acu-rite (the glass slide type you use on lathe DROs). The 1HCL can read it, and I'm attempting to tune the set-up as if it were a "T2" quadrature shaft encoder. In principle this should work, although it is a bit peculiar as there are about 12700 counts/step.
Unfortunately, when running the tuning command M569.6 P1.0 V1, DWC feeds back that the tuning has failed with a measured backlash >0.5step. I've disassembled, and reassembled the system, tuned the air gap on the forcer, and adjusted the encoder to be as parallel as possible with the platen. The absolute lowest backlash I've seen thusfar is about 0.4 step. Understandably this is rather high, I suspect the source of the backlash is coming from either a) a stick slip condition in the bearings rolling on the platen, and/or b) excessive cogging between the poles. Notably, linear stepper motors often use an air bearing setup to maintain the air gap between the platen and the forcer; the set-up I'm using has standard roller bearings.
Are there any other tuning profiles or configurations I can try to get this working? Or is there a way I can increase the backlash allowance?
Thank you!