A little Brushless Motor Tech 101:
According to Richard Fechter's schematics and accompanying description of the operation of a Crystalite controller, the FET's that drive the coils rapidly alternate between + and - voltage during operation:
Is this due to the phenomemon of attracting the permanent magnets as they approach the coils, and repulsing them after they pass?
Generally that is how the motor works but not the way the FETs work. The switching of the hi and lo FETs happens very quickly all the time and is continuously controlling the current to each of the 3 coils. The frequency of the switching could be from hundreds of times a second to tens of thousands. I don't know what they use for larger EV controllers but I use 30 khz on the smaller brushless DC motors I work with at my job.
The controller has a sensor to tell it where the rotors or the magnets are in their rotation relative to the coils. The sensors are often Hall effect magnetic sensors and referred to as "hall sensors" or "hall switches". By controlling the current to the coils you push or pull relative to the poles of the magnets as they rotate around. If you graph the current on each coil you would see a sine wave on each coil, each 120 degrees apart from each other.
If you haven't already figured out this is all done with computers, usually digital signal processors (DSP) that can use math algorithms to figure out how to control the FETs. The math takes the hall information, the throttle input and information about the current speed and current used in each coil (measured with the shunt resistors) and decides how much current should be applied to each. Then it sets the duty cycle of the PWM (pulse width modulation) circuit.
The PWM circuit is what switches the FETs on or off rapidly as needed and also makes sure you not only don't have both the hi and lo on at once (which would short across the battery!) but also adds extra time when neither one is on. This is called the "dead band" or "band gap" and is because FETs sometimes are a little sticky turning off and you could still end up with both the hi and lo energized at the same time.
This hi-lo FET circuit and PWMs are also used on the brushed motor controllers too but instead of 6 FETs (3 hi-lo pairs: one for each coil) you have 4 FETs or a hi-lo pair on each side of the brushed motor. Here the controller just controls the current either forward or reverse and the brushes in the motor provide the commutation to the internal coils which are on the rotor now while the magnets are fixed on the outside.
So again the controller and PWM will switch very rapidly to control the current to the motor it happens fast enough that the motor just sees a constant current.
And that is as well how the brushless hub motor works? Just curious.
Also, while I'm asking...
Someone said somewhere on here recently that the hub motor is totally quiet. I don't find that true at least with my XB600 with 600 watt brushless hub motor. When I first got it I thought "totally quiet" but I was comparing it I guess to a gas motor. You can hear a "shhhhhhhhhing" sound that comes from it. And I also hear sometimes a "high pitch" sound if I'm just letting the scoot roll without applying any power with the throttle...just a faint high pitch that gets higher the faster I roll. Course everything around me has to be really quiet. I also "think" (if it's not my imagination) that my hub motor makes a bit more sound than when I first got the scoot. I'm assuming like any motor that as it wears in (I've had it about 6 months now) the sound of the motor may increase just a bit. Anyway, not saying something is wrong. It's still virtually whisper quiet and running great and strong. I hope it continues cause I really love this ride!
Thanks for the post. More graduate work than 101, but what's a heaven for.