I've appended new versions of Fig. 3 and Fig. 4 to: http://www.mtaonline.net/~hheffner/HullMotor.pdf
I think these figures are the main problem with my communicating the hysteresis explanation for the Marinov Ball Bearing Motor.
I also included in the above two photos of an experiment under construction. Slightly kludgey, but hopefully effective eventually. The idea is to run two motors in series, one with magnetic bearings and the other with non-magnetic bearings. This scheme is not looking good at present due to the only sporadic conduction of each of the bearings, and therefore almost never simultaneous conduction. I've tried orienting the bearings vertically to increase tension and contact, but without effect. Looks like I'll have to run each motor separately.
I'm putting the bearings through a series of washes to attempt to increase surface conduction. I am concerned that clean surfaces will increase the possibility of welding. A capacitor across the battery would clearly improve the torque that can be obtained with a given sized battery, but also increase the likelihood of welding. Though I had the experiment fully assembled and wired, I haven't applied power yet because I want to be fairly sure of getting a meaningful run without having to replace any welded bearings. I already spent more on this than I would have liked.
I expect to try a powdered graphite lubricant to encourage better contact and hopefully to avoid welding.
Some of the youtube experiment videos show motors that seem to run well with low current. I would not have expected the contact to be so poor. Maybe it *is* indeed poor and that is key to getting enough torque. The fact flywheels are required on some experiments is a sign that contact is poor. Torque is proportional to i^2 in a good operating range, so putting all the current through one bearing at a time will maximize torque for the total current draw.
Any suggestions for improving steel-steel electrical contact in the bearings would be appreciated.
Best regards, Horace Heffner http://www.mtaonline.net/~hheffner/
