Non repeatability of motor motor RPM observed while tuning gains

[Joe Dinius]

Hello everyone and Happy Holidays!

I am observing a strange issue when I try to run the same motor configuration (including speed controller gains) against the same RPM step response reference value and I am getting very different responses.  My "clean" response looks like this:

There's only a slight initial transient (just prior to 4535s in the plot).  Now, when I run this same example again a few minutes later, I get this gross chart:

I requested the step at around 4574.5 and until about 4581 there was an annoying low RPM chatter.

Running the duty cycle at 20% shows a similar indeterminism; sometimes the wheels begin spinning immediately and sometimes there is a nasty low RPM transient that eventually (though not always) is overcome.

I ran through the Detecting Motor Parameters step multiple times and am not seeing any issues during it.

I suspect that I have a bad motor.  Does anyone have any other suggestions or follow-up questions?

Thanks,

~Joe

[Jaroslav Klapálek]

Hello,

   have you tried different methods for controlling the motor? I think that most of these issues are caused by not having a direct motor rotation sensor (or I least I don't have it), but there is not much you can (at least easily) do.

Happy new year!

J.

Dne pátek 22. prosince 2023 v 0:58:18 UTC+1 uživatel joseph...@gmail.com napsal:

[Joe Dinius]

Thanks for responding Jaroslav.

I think you are correct: the issues I was seeing are due to the lack of a direct rotation sensor on the motor. I was able to fix my cogging/commutation problem by lowering the step response rate from 10000 rpm/s (the VESC default) to 4000 rpm/s.  I was able to tune the speed controller gains sufficiently at this point, so I moved on.  I hope to set up continual adjustment to the gains online at some point, but for now I am ok with the lower step response rate.