Good videos! Nice and clear explanation. This will give the starter a good general idea what this is all about.

One thing I must address though, I don't know how you controlled your attraction motor 20? years ago, but this one is in a sense a 'pulsed attraction motor' Your switching type does not care much for mechanical loads the pulse is always short, in a ("more true") attraction motor the coils will be powered aslong as it takes to get the rotor in alignment. So that mean that there will be a higher current draw per on-time at 3rpm then compared to 3000rpm since at 3rpm the coil will be on for like seconds. If we want to also have some considerable torque of these designs we must (in my opinion) make the system so that the coil will be on as long as is needed to get the rotor in alignment and not just a short blast. Maybe some reed pulse with variable pulse width. Or some opto controller sensing a wheel with holes that are as long as a rotor alignment stroke. I know it is not really necessary at this point but I noticed from my old flux-gate gen/motor tests that you also need control to advance or retard the on-time of the coil. Since at certain rpms the slowness of the steel start to become an issue with a fixed pulse on-time.

Regards,
Steven