Sensorless, halls, encoders, motor efficiency?

[Lebowski]

Only the backemf itself can tell you what it's phase is and therefore give you the correct phase for supplying current (into the motor)...

I doubt I understand what this means. Are you saying that it's possible to always, with perfect accuracy, to read rotor angle with zero sensors and that sensors are simply used to compensate for some sort of inadequacy?

No, what i'm saying that from an electrical energy point of view the only thing that matters as far as signal timing goes is the backemf of the motor.

 

[Solcar]

Thinking of a rider on a swing, the push should be timed at right after the rider reaches rear peak height. With a motor, right after peak magnetic field falls on the coil would be the time when adding energy contributes most efficiently to the rotational energy stored in the motor.

 

[justin_le]

This whole topic of wave forms, pulse widths / frequencies and timing has some interesting possibilities /results. Lots to explore beyond the simple/gross efficiency question. Harmonics and how a controller best deals with them is another whole area to explore. I need to learn much much more :?

I'm not sure if bowlofsalad ever got a clear answer to his original inquiry, but in spite of the many variations here I can say a few things in regards to ebike controllers and motors encountered in practice.

1) If you have a trapezoidal drive motor controller with a relatively low eRPM motor, and the hall sensors are properly located on the stator core for neutral hall timing, then the efficiency and motor performance will be pretty much identical whether you run the controller sensored or sensorless.

2) If you have the same system in 1, except that the halls are poorly aligned in the motor so that their timing is off, then you'll get noticeable current ripple and "buzz" from the motor when you run it sensored, but sensorless it will run nice and smooth. In this case running sensorless should have marginally better performance. Probably undetectable riding the bike, but likely measurable with an accurate dyno system.

3) If you have a motor with a high eRPM's, then many models of sensorles trapezoidal controller will have some commutation lag time as a result of firmware limitations in their back-emf sensing code, and this will result in delayed commutation timing and excessive ripple current and inefficiencies. How much really depends on the particular controller and eRPM at which you run it, but I've seen some very ugly oscilloscope waveforms of ebike motors running at high speeds on sensorless controllers not really up to the task. In this situation, sensored is generally better.

HOWEVER, some digital sensored controllers also have a firmware lag time from when a hall transition happens to when they commutate the phase outputs, and some really bad ones only synchronize the change in commutation with the start of the next PWM duty cycle on the mosfets, leading to an unnaturally discretized and jittery commutation frequency even if the hall sensors themselves are perfectly timed. Meanwhile some sensorless controllers will still be spot-on in having neutral commutation timing even at high eRPMs. So with high eRPM motors the answer is very much one of "it depends" on the implementation details of the particular controller.

4) If you have a field oriented controller, then best behavior for sure is sensorless for the reasons that Lebowski said. The only time it makes sense for a FOC to base it's output on the hall signals is when the motor is at a very low speed and the controller can't reliably figure out the rotor position.

 

[bowlofsalad]

This whole topic of wave forms, pulse widths / frequencies and timing has some interesting possibilities /results. Lots to explore beyond the simple/gross efficiency question. Harmonics and how a controller best deals with them is another whole area to explore. I need to learn much much more :?

I'm not sure if bowlofsalad ever got a clear answer to his original inquiry, but in spite of the many variations here I can say a few things in regards to ebike controllers and motors encountered in practice.

Thanks for the response to this thread, words cannot say how much I appreciate you taking the time to give your thoughts and experience on this question. I've read many things and came to basically identical conclusions as to what you've post. Really your testing with the BAC500+ is what shed a lot of light on this subject and forced me to wonder a bit deeper on the subject. What I had really wondered is if others had created dynamometers similar to yours and done various plotted and overlayed comparisons of all different kinds of configurations. This isn't to say I don't appreciate the commentary and discussion, usually when I ask a question I look for less anecdotal data.

I've watched some of your videos demonstrating your direct link dynamometer and I hope to emulate yours in order to do some tests of my own. I don't mean to derail this thread too heavily, but one question I've wondered is where the break away RPM comes in between efficiency of BEMF versus hall sensors, among other things. This isn't a request, just idle chatter pointlessly sharing whatever floats across my mind.