Found this toroidal current transducer here;
http://sentrancorp.com/If the outer coil was wound afterwards onto this toroid (look at the white one), then using the bucking wind method would make real sense. Also, if you look at the outer coating being a plastic, this would explain why the outer wind started to embed itself. If this is what SM used then for sure he would not want us to see under the toroid, which we never did. The dimensions of the shown white coil is fairly close to SMs. You will never find a standard ferrite core to come close to SMs toroid dimensions. We had spent months on that off and on during these few years. If SM wanted to use a choke, he could have picked from thousands of standard ones. His toroid was not a standard format. Actually 90% of the OU ability could be right in that toroid if you consider the FTPU did not have anything more then that and a few turns of wire.
The point here is that if the core is a current transducer that usually picks up the energy of a wire going through the center hole, then this indicates the transducer has a good enough sensitivity and should therefore be able to pick up energy from a very low level to a higher level. Low level to enable system start up and higher level to withstand an increase and leveling off of gain. The outer wind would then compound the energy imparted to the current transducer. Can't really know for sure unless it is tried so I will look for a few of these coils and do some tests. I know @GK and I have tried with horizontal windings inside and outer bucking coil but our wire awg was way too low.
The other factor I would like to look at is how to start the whole process. Now to start such a process from a near zero position, this goes above my acumen to identify the right component model but I think all you would need is a transistor and 1, 2 or 3 zener diodes of different values. The transistor would have to have three main attributes;
1) The transistor base would only require the most minimal amount of voltage to start switching, possibly in the mV level.
2) The transistor base would also have to withstand a voltage increase to the lowest value zener diode.
3) The transistor will not require heat sinking. (This may be the 20 minute limit.)
If the TPUs all share the same center toroid, then the only other variable here is the coils in the outer ring that will add more inductive (as well as some capacitive) value to the complete process. If the process ran via feedback coils going to the gate of 1 or 2 alternating transistors working like a virtual Slinky Toy, the outer circumference of the ring would in fact be responsible for the overall working frequency, just as SM had indicated the frequency is a function of the circumference of the ring and this only define length and has nothing to do with a circular length.
So, we again should not be fooled by the actual physical shape of the SM TPU, by thinking that this circular form has any advantage on the overall function. Ether energy should not care about shape. Give it a good reason to come in and give it a good place to land and it will. It always does. We just don't yet know how to play the game to our advantage.
Then I am thinking that If SM made the conscious effort to show the video with his ammeter showing highest readings over the toroids. Now why would he show this and for him, why was this so important to see amperage (current) in them toroids. Well if the toroid was a current transducer, knowing about the current at the toroids would be an important thing to know.
Then consider the center of the toroid is epoxy filled. Well you can make a small circular circuit with one or more transistors and embed them in the center then have those four center wires and bingo. If you knew those two wire pairs had different uses, then this would give you a good reason to vary those wire gauges. Then for the LTPU you parallel two toroids to increase the overall output amperage.
But of course, these are all theories that remain to be tested.
wattsup