Joule Motor

[synchro1]

I lit a 120 volt LED off the BEMF that would normally go to the charge battery in the Bedini SG by spinning a bearingless 1" neo sphere inside a Spiral bifilar, The circuitry inside the Lights of America 7 watt LED, converts the BEMF from the 12 volt 400 milliamp input @50k, into 70 Lux units of power. This is the first time I've been able to measure the BEMF accurately.
This equals about .o % of the input. Not enough to generate any meaningful EMF in the secondary power coil. I would look elsewhere for an explanation for the acceleration, like the effect connecting a charge battery to a Bedini rotor, or Lenz delay.

[Lynxsteam]

I have built a new motor, and narrowed in on the schematic.  The AC comes off the secondary flyback coil and goes to a bridge rectifier.  Lots of weird things occur and I can't even repeat some of the experiments because I don't have the equipment to control variables.  The most interesting things are:
If the battery voltage is higher than the zener diode breakdown voltage things get interesting.  Start up and it runs along, then connect the positive from the BR and the rpm shoots up and the charge on the cap shoots up above rating.
Also if I connect the negative of the bridge rectifier it slows the motor down noticeably and running gets rough.  Take the negative lead off and the rpm goes shooting up.
With the volt meter hooked up to the bridge rectifier, posiitve lead to battery, voltage on the battery will increase while running for awhile and then settles.
Overall, I don't see anything breaking laws of physics (but I didn't expect that). 
My preliminary thought is the flyback is getting amplified by the passing magnet and has more than a whisper of ma.
There are times where I think "wow".  And then a few minutes later I see battery voltage drop and I say, "yep".
Overall, I do see about 250 ma drawn from the battery, 25 ma going back if I measure with amp meter, and yet when i am done with 10 minutes of running battery voltage can be lower or higher.

Here's today's video
http://www.youtube.com/watch?v=EztG45B-qr0

[synchro1]

Apply torque to the drive shaft, and the battery will drain at the same rate it would without the recovery loop and speed up effect. The same thing holds true for Thane Heins's Lenz delay circuits. He can charge the source battery while the vehicle's coasting. Looping the a.c. BEMF from the secondary is actually increaseing the the hysterisis loop in the primary, causing Lenz delay. This causes a speed up that can actually send power back to the source battery with no additional load while ideling alone but nothing more.

[synchro1]

@LynxStream,
Two Zener diodes, rated zero to x hi-voltage, reversed from each other on each end of the secondary, looped to the primary, may produce the same desired phase shift acceleration effect from the capacitive nature of the inductor. 
   

[synchro1]

One Zener, an inductor choke with a ferrite core and a shottky diode at the return end to the power coil. Coupled with the Hall effect and power transister combination. I plan to try this Tuesday. I think this kind of simple tunable retarder may alter the way power is generated by the magneto effect of the spinning magnet rotor, and cause a Lenz reversal, propeling the spinner. I'll be be back with my results. The way you wired  the Zener was inspiring to me, thanks.