We don't have the plans of either the TPU or the Kapanadze machine. The best method is therefore, in my opinion, to try to understand what the basic principle of operation might be.
Even if I don't think these two devices are overunity, it is possible that unusual effects may occur that surprise their builders.
All these devices show many coils that are clearly capacitively coupled, in addition to magnetic couplings, and simultaneously used with various frequencies. Could displacement currents in intercoil dielectrics cause unusual effects? The subject seems to me to have very little coverage in the academic literature.
It would join this thread. We notice that it is easy to pass a current through a dielectric, we obtain it between two plates, simply by applying a potential difference, it is the principle of the capacitor.
But have you tried to induce a current in a dielectric from the EMF of a variable magnetic field, and not from a potential difference?
Until today (maybe) I had tried and failed. For example, the immersion of the coil of an LC oscillator in water, a dielectric of high permeability (ε
r=80), shows no change when one would expect that the current induced in the water-dielectric would disturb either the amplitude or the frequency of the oscillator.
I only recently understood why. The EMF acts on both positive and negative charges, so on a loop, the symmetrical effects compensate each other perfectly, unlike the case of the capacitor where the displacement of charges in the dielectric is stopped at the plates of opposite polarities, creating asymmetry.
Could we then simulate this asymmetry by opposing two EMFs in two dielectric half-circuits, separated by two plates across which a voltage, such as that of a capacitor, would be recovered? See the attached schematic diagram (inductionInDielectric-Principle). Two U-shaped dielectrics are separated by two plates, and EMF of opposite direction are induced in each one, from two coils powered by currents of opposite direction. Each EMF is looped back through the diameter between plates (not shown in the diagram).
At the interface between the two dielectrics, where the displacement currents tend to oppose each other, the plates should recover a potential difference due to the opposition of the two EMFs.
I tried to highlight it experimentally, with a first rough experience. Two coils of about 100 turns of wire are laid over each half of a double U dielectric circuit (ferrite). Copper plates have been plated at the interface of the two dielectrics and are connected to the scope.
I inserted an aluminium foil connected to the ground (removed for the photo), between the coils and the ferrite, to prevent the probable electrical coupling between the coils and the plates.
The result seems positive. I'm clearly observing the signal on the monitor. If I change the direction of the current in a coil, the signal becomes 7 times weaker, but not zero because despite the aluminium screen, there is still a direct capacitive coupling between coils and plates (without the aluminium it is preponderant, we see almost no change).
There may still be experimental biases. A more convincing experiment remains to be done, I'm thinking about it.