18-12-2012, 10:01 PM
Yes very interesting !!
I like the idea of using the fets instead of the flywheel diode. makes sence with the on resistance of the fet being much lower that of a diode. The drama I can see there is accurite switching. Did you notice on the Cro trace there were quite large spikes on the rising edges of the switching wave form?
I'll bet that doesnt help either. Good electronics design would have that waveform nice and clean
The thing is, with Pylon it's all flat out, well mostly (90% of the time flat out running of the ESC). When they are run like that they don't have the PWM signal superimposed on the switching signal so that is not nearly as much of an issue. It's just the fets switching between the three phases
so the energy is dumped into the next set of windings via the two fets as it switched.
I know inductors don't like to be driven in one direction then another it takes power to change the field around and even more so if you drive the inductor far from resonance. It's like phically trying to make a pengelum swing faster than it's natural frequency by manually draging it back and forth rather than just keeping it going by adding a little energy at the end of each motion, like you do when pushing a child on a swing. I wonder if you could set up a motor so that when running flat out it would be on or close to resonance that would save a massive amount of power as you wouldn't have to dump anything much across a fet or a diode
Or even if you could make it close to resonant at the superimposed PWM frequency that would help a lot at lower than Max power settings and that frequency doesn't change with RPM so it would be resonant right through the range.
I like the idea of using the fets instead of the flywheel diode. makes sence with the on resistance of the fet being much lower that of a diode. The drama I can see there is accurite switching. Did you notice on the Cro trace there were quite large spikes on the rising edges of the switching wave form?
I'll bet that doesnt help either. Good electronics design would have that waveform nice and clean
The thing is, with Pylon it's all flat out, well mostly (90% of the time flat out running of the ESC). When they are run like that they don't have the PWM signal superimposed on the switching signal so that is not nearly as much of an issue. It's just the fets switching between the three phases
so the energy is dumped into the next set of windings via the two fets as it switched.
I know inductors don't like to be driven in one direction then another it takes power to change the field around and even more so if you drive the inductor far from resonance. It's like phically trying to make a pengelum swing faster than it's natural frequency by manually draging it back and forth rather than just keeping it going by adding a little energy at the end of each motion, like you do when pushing a child on a swing. I wonder if you could set up a motor so that when running flat out it would be on or close to resonance that would save a massive amount of power as you wouldn't have to dump anything much across a fet or a diode
Or even if you could make it close to resonant at the superimposed PWM frequency that would help a lot at lower than Max power settings and that frequency doesn't change with RPM so it would be resonant right through the range.
“The knack of flying is learning how to throw your machine at the ground and miss.”
"When once you have tasted flight, you will forever walk the earth with your thoughts turned skyward, for there you have been, and there you will always long to return." ~Leonardo Da Vinci