olaf-lampe wrote:IIRC the timing advance is depending on momentary RPM and load: the higher the RPM the more timing advance is necessary for the best output.
I believe the sensored infineons may have an auto-advance algorithm, because I felt some small vibrations comming from the motor under certain conditions and after a while they're gone. Seems it takes a bit time to find the right 'commutation angle'
The worst thing is too much advance, it's heating up the motor and controller.
Just finished reading the paper from Krumpi which give me wild ideas, given a non auto advancing controller like mine, I think we could have a very advanced fix hall sensor and separate controller to introduce timing delays, which then could dynamically changing the delays based on RPM