Microstepping in relation to deviation?

[jackatom74]

My build is a 1m delta and I am stuck at .045 being the lowest deviation I can achieve. I am currently running 1/32 microstepping with interpolation on .9 NEMA 23 motors. Would there be any benefit to dropping to 1/16 with interpolation.
(On another thread I read about adjusting belt tightness, and plan to do that tonigt)

One other question, can anyone explain deviation and how it is calculated?

[Phaedrux]

Interpolation only works with 16 microsteps.

[dc42]

With x16 microstepping and 0.9deg motors, the theoretical resolution is 5um using 16 tooth pulleys (200 steps/mm). That's quite a lot less than the 45um deviation you are getting, so using higher microstepping is unlikely to help. I get around 25um deviation with this setup, using linear rails and Smart Effector.

Deviation visit the root mean square average of the height error at the probe points.

[jackatom74]

@dc42 I am using 20 tooth pulleys and have set my steps accordingly, is there possible benefit to me using 16?

[Phaedrux]

You get to use interpolation

[timcurtis67]

@phaedrux said in Microstepping in relation to deviation?:

You get to use interpolation

To avoid confusion, I believe @jackatom74 is refering to a 16 tooth pulley. Not 1/16 microstepping.

[Phaedrux]

@timcurtis67 I guess I shouldn't browse the forum before coffee.

[timcurtis67]

@Phaedrux Yeah I do that all the time...

[jackatom74]

@dc42 said in Microstepping in relation to deviation?:

Deviation visit the root mean square average of the height error at the probe points.

Thank You DC42, Something clicked in my head and I went back and looked at the the repeatably of my trigger height and started messing with the smart effector sensitivity. Correcting this fixed everything. I am around 20um deviation now, Also thank you to everyone else who lent me their guidance.