A101960,
Thanks for the expansion. Yes, it makes sense that bad segments could be passed over while motoring. Not that it means the dynamo is good, but, when running on the engine, those same bad segments are passed over while the good ones would be generating. All this because Faraday's Law is a two-way street. In the case of my dynamo, it motored and generated, which was a lucky coincidence. In the case of Franenstein's dynamo, it is rather clearly in trouble, so, despite my apparent mistatement of indicating a good unit, it would still be revealing to know if it motors. If so, and it is truly bad, it certainly would be an empirical proof of your point.
Richard