Slayer's third video in the series:

https://www.youtube.com/watch?v=OrHCh0N5UQU

Have a look at "Lidmotor's" mechanical Joule thief:

https://www.youtube.com/watch?v=1U5lXqkoMzE

Here's a picture of an inductance measure of an air core coil. The same coil was wrapped and measured for inductance around the two AA batteries in the picture: One, a Radioshack alkaline and the other a Duracell Quantum.

The inductance of the copper coil around the Duracell was 170 mH and around the RadioShack merely 70 mH. This invisible difference in inductance is enormous. It comes down to lighting over twice the amount of LED's for the same input with the Duracell core.

Here's an interesting version of Lidmotor's mechanical oscillator; Winding a AA  "Duracell Quantum" copper top battery with a copper wire coil would increase the backspike output tremendously: Ludic states the frequency of this model is 500 Hertz.

https://www.youtube.com/watch?v=iHmTc0PwiyY


The challenge is to oscillate the contact between two positive poles of multipal swapper batteries.

One can easily see how the AA can be turned to a neutralization pulse coil with the addition of magnets on both ends of the battery. This arrangement would require a commutator to separate the flyback output from the regauging output generated by the reappearing permanent magnet field.

My tests have shown that the operating frequency needs to slow down to 1 or 2 Hertz to harvest the regauging output; Therefore the magnets and pulse coil need to grow over sized to compare with the flyback gain.

2 tiny magnets on a AA battery would not match the flyback output alone up in the 500 Hertz range.

The pay dirt appears to be in splitting the positive on the AA higher frequency scale as George Cianiotakis demonstrated in his plasma reed OU video.

The best approach looks like the high perm AA core boost converter powered by split positive power source.

A 9 volt and 2 AA in series doubling as flyback inductor cores, would yield a 6 volt differential.