Setting up 5i25+7i85+8i20

I made up the NPN optical interrupt board and gave it a try on the spindle. I can fine tune in a P-I loop that maintains speed reasonably unloaded but the values are so low that as soon as it's loaded up it's no good. It takes way too long for the current to rise to make it usable. So it's just not going to work with the 4ppr spindle disk but I'm wondering just how much a suitable encoder is going to help.

Now that I look at my HAL file again I realize I'm probably not doing this as well as it could be but I'm wondering how much it's affecting the results. For example, with a 4ppr feedback. I need to run the velocity output from the encoder interface into lowpass, then I scale it to rpm, then send it into the PID loop. I also have the encoder "filter" on, which slows it down a bit.

I'm thinking that since motion has an RPS pin, I can just send that into PID to avoid scale. I assume that with a real encoder I can turn the hm2 encoder filter off and remove the lowpass component. I'm wondering if after all this I'll wind up with something I can use.

You could try running the PID with the unfiltered signal and let the integrator do the filtering.
(You will probably not be able to use much P in this case)

I remembered about a cheap 600pr AB encoder I had lying around and whipped up a quick coupling for it. HUGE difference. Once I got rid of the filters and scales between the hm2 velocity signal and the pid input I had something much more useable. It still took some tuning and there was still some quill banging in the bldc=h mode. Once I got it in qh mode it got much quieter.