Jensen's UDT is covered by Thane Heins' patent. Nothing new here apart from the feedback winding.
Yes, I'm looking at the differences between the UDT and the previous circuit and they are clearly different.
Here's my take on the above circuit--theory of operation:
L1 and L2 are bifilar wound coils. K1 and K2 are initially closed at T
0, K1 effectively shorting out L1. C
0 energizes L2 and and L3, creating a magnetic field within both halves of the dual E-core separated by an air gap much like a flyback transformer. At T
1, K1 and K2 are opened and the magic begins. With L1 and L2 being bifilar wound, any magnetic field present in the lower half of the core is effectively cancelled out electrically. L1 and L2 become nothing more than a wire-wound resistor. But the magic is they dump the magnetic field built up in the lower half of the core to the upper half of the core. Now the voltage built up in L2 and L3 plus the source voltage at C
0 is directed to the output at RH via the diode. So far, so good, but Lenz Law should bite us as it always does. So the current flowing to RH should push back in the reverse direction, opposing our input, just like it would do in a typical transformer or generator. Not so fast though, L1 and L2 are bifilar, meaning the magnetic field gets reversed, but it's still cancelled out. Lenz Law cannot get back to the source to push against us anymore.
Granted I'm still trying to wrap my head around it, but the switching and bifilar wound primary with an air gap between primary and secondary cores seems absolutely ingenious to me if I'm deciphering this all correctly. It also looks like something that can be scaled to whatever size you need. In addition, there is talk on the OverUnity forum that Tariel Kapanadze buried this simple technique under a cloak of spark gaps and funky looking coils to deceive anyone trying to replicate his work.
What I am convinced of is this seems simple enough to prototype, connect to a scope and see whether it flies.