We try a Melnichenko setup.

We have a 20 turn coil L1 (forward biased), a 10 turn coil L2 (reversed biased) and a 10 turn coil L3 (forward biased). When we switch on Q2 and Q3, the fluxes of L2 and L3 cancel. When we let go the coil setup, the MOSFET switch consisting of Q1 and Q4 shorts L2. The idea is that we charge the core (coil setup) with low inductivity and discharge with high inductivity (possible overunity gain).

There shall be an air gap between L1 (top half) and L2/L3 (bottom half) as can be seen below.

We use a delay line of three 74LS04 inverters (3 x 10ns) to make sure L2 is shorted before Q2 is switched off. We omit L4 and C2 in this experiment.

• (yellow) FG signal
• (blue) drain of Q2 with respect to ground
• (red) voltage of C1 (100nF)
• (green) voltage over 125R load (4 x 500R parallel) 84.7V RMS

Pout = 57.3W for Uin = 60V (63.3W) COP = 90%

We try something else:

No avail: 50%