Replies: 2 comments 1 reply
-
With a close-loop commutation (stepper + calibrated encoder), the microstepping is not really a thing. The angle used for activating the phases is interpolated and the more accuracy there the better. The commutation accuracy is up to 163microsteps (32768 encoder resolution / 200 steps). In practice, even if the encoder had less resolution it still work good. I think I tested with 2^11 resolution limit and I couldn't tell the difference in operation. That equates to 0.17deg error. But it makes sense right? Because that is only 10% of a full step - much bigger than 1.5%. Whether we command 90deg or 80deg electrical angle (load angle), the generated torque will be similar. It only when we start increasing the error of load angle even further, the torque would change a lot. |
Beta Was this translation helpful? Give feedback.
-
Right, the step accuracy is irrelevant for the close loop control scheme. The 2^15 resolution results in an accuracy about |
Beta Was this translation helpful? Give feedback.
-
For a low cost 1.8 deg stepper motor, the step accuracy can be around +/- 1.5% (+/- 0.027deg) of full step. With the microstep of 256, the accuracy per microstep will be +/- 0.027 / (1.8/256) = +/- 384%. It may lead to some strange behavior under heavy load or near stall.
I think up to 32 microsteps can be a reasonable choice for low cost steppers.
Beta Was this translation helpful? Give feedback.
All reactions