I missed the question, I could have told you about -S
I was using it when experimenting with the floating point math when there is no FPU.
You can see the __aaebi_f functions there
Hmmm⌠you will still get phase matches or close matches across all motors at some periods anyways, since the duty cycles are changing so fast statistically you will get hundreds if not thousands per second rising front or falling back overlaps. No? I canât quite put a quantitative estimate but that should happen.
Good point. Yes, staggering them would be statistically irrelevant. And the benefit from avoiding simultaneous switching wouldnât be worth attempting any more complicated approach.
Would it even be possible to keep the motors PWM âout of syncâ when they spin with individual RPM?
Ideally the hardware should be tough enough to deal with worst case PWM scenarios.
No, if the timers are in sync, and running at the same frequency, you could set them up to be out of phase, just like the current code would set them up to be in-phase.
While it is true that there would be overlap when the duty cycles are >50% on both motors, the centres of the PWM would still be entirely out of phase at all times.
Are you sure? I think it could make a difference, especially at higher power levels, but I admit I have no data to back up this claim.
The PWM frequency is constant and not dependent on RPM.