As opposed to having a physical sensor on the motor, this measures current flow into and out of the motor and can instantly direct coasting and regenerative braking back into the batteries to re-charge them. That's my understanding of it anyway. Probably more complex than that in real terms.
Does that mean I can use the new sensorless controller to drive my old 500watt brushless ebike sensored motor just by connecting the three phase wires.?
'Sensorless' addresses the idea that the controller can power the motor through throttle action with no additional sensors than that of the hall sensors at the motor. 'Dual' addresses the point that additional sensing, like rpm (of the wheel), to determine a more accurate position when hall sensors and decoupling capacitors variegate under heat. I think the narrator has just cut over the spokesperson, who is explaining the difficulties maintaining a sensored motor, and how a sensored motor works. I'm just a hobbyist. I hope 'back emf' is possible, it seems to just describe sensored application, as hall sensors read and return a data current dependent on the magnetic current that the rotor magnets produce when circulating around the stator coils. The controller will produce a powerful 3-phase square wav constant emf current to the coils, causing propulsion in the magnetic rotor. The halls positioning in the motor is dependent on motor pole count, that is, in what pattern the coils are wrapped over the armatures. I've shaved the stacked sheet metal armatures from square to cylindrical and recoated them (will epoxy and high temp enamel work as thermoplastic diffuser of residual current in the armatures…??) I hope this will give sensors more space from hot coils and not disrupt the sense accuracy (unfortunately i have had calamitous doubt over the integrity of my coil wire. I wish I'd stored it between bits of foam in a suitcase). Maybe heat fins welded to the outer hub casing will dissipate heat. I've put some silicone over the halls to reduce vibration and heat, and I'm looking forward to soldering the connections for the cycle analyst (more sensors!) on to the controller pcb.
Nah, in the vid, dude stated that when the motor is not spinning there is normally nothing to use to determine rotor position. No motion, no BEMF. Instead, I think what the controller is doing is dumping current into each of the coils and measuring their "ringdown" oscillations in order to determine rotor/magnet position. Each coil will have it's inductance affected by the rotor magnets and once all the positions have been thusly "fingerprinted" it should be trivial to determine postion.