this is just an idea,
use a stick as the bar, and lightweight iron balls or magnets on the ends of the wooden bar
have magnets covering 50% of the circle, with their fluxs even as possible.
like those cheap magnets for testing,
not drawing is Rough, but its just a picture.
so if you have 2 ends weight the same, from same length from the middle then its even yea?
if you provoke that with your finger it will move a bit and stop , yea?
if you used magnets on 50% of the wheel arch, when it turns 50%, then the other ball takes its turn,then i can see that if it did work, it would be extremely hard to prove false and wipe away the , '' it was running on Batterys '' comment.
it could,mind you , thats why im saying use very cheap parts, and play with fiddle sticks so that it , if it did run, then you would know.
what do you think?
that grey line and outer gray circle line was to help me make it a bit neater, the flux lines would cover 50% as much as possible, useing just mades and predicting how strong they are .
and the magnet layout would be neater then that, that picture is to help people, if they don't get what im saying, although the magnet layout would have to be neater then that,so im saying its just a rough picture.