Selfrunning Free Energy devices up to 5 KW from Tariel Kapanadze

[verpies]

The variable corresponds to the secondary winding of the parametric transformer, the inductance of which is varied periodically in time by the alternating current in the primary winding.

The saturating AC in the primary winding would vary the self-inductance of the primary winding as well.

None of the flux produced be either winding links with the other winding, and there is no mutual coupling whatever the relative directions of the two fluxes.

That is quite a feat.  None of the conventional transformers work that way.  In fact conventional transformers try to maximize mutual flux coupling between the primary and secondary.

This parametric transformer should not be able to transfer any power from the primary to secondary (or induce any emf in the secondary) when used as a conventional transformer. 
In fact such an EMF induction attempt would be a good test to verify that the design is well executed and there is indeed no mutual coupling between the primary and secondary when operating in linear mode (while no saturation effects are present).

Wave diagram below: Analysis of the loud humming noise in the TK 2004 Video superimposed with a 50Hz sine wave. The mark »0.100« denotes 1/10 second. This means the humming noise correlates with 50Hz and 100Hz.

ElementSix already noticed that complex sound here, and I also heard it on the soundtrack of the video that he cited.  This sound definitely contains frequencies well above 50Hz.

[Zeitmaschine]

This parametric transformer should not be able to transfer any power from the primary to secondary (or induce any emf in the secondary) when used as a conventional transformer.

And maybe that's the trick. Just another coincidence:

No matter what input wave type is, the resonant output is SINUS WAVE

Is that so?

The parametric transformer consists obviously of a two-part core like the two-part core of the yoke device. Now my problem is, where can I see a two-part transformer core on the Kapanadze device? Or could there be another way to do this with an ordinary off-the-shelf E-core transformer? (Will keep thinking ...)

If my understanding of this functional principle is correct then the intent of the divided core is to have a loose magnetic coupling between the primary coil and the secondary coil. But it does not really matter how this loose magnetic coupling is achieved (by having a gap between the two half cores or by rotating one half core 90 degrees with respect to the other). So the current supplied to the primary coil is just to excite the secondary coil, which then produces its current (or voltage) by itself from thin air (due to parametric excitation). Now someone try this and report a working OU device.

And furthermore, why is it that the toroidal shape of the yoke device reminds me on the toroidal shape of the Marks device? Don't say that's just another coincidence.

[verpies]

No matter what input wave type is, the resonant output is SINUS WAVE
Is that so?

Well, yes except for the temporary/transient states when the device is powering up or pri/sec phase difference is changed rapidly.

This is because in this type of transformer, the current in the primary does not induce any EMF (voltage) directly and consequently no current in the secondary.  Because of this, it is a very bad conventional transformer.  It should not transfer any power in the linear range of the BH curve.

The oscillations in the secondary LC circuit are due to parametric excitation (periodically varying the inductance of the secondary).  Also a parametric damping is possible in this secondary LC circuit, depending on phase difference between the primary and secondary current. 
Curiously, in this type of transformer the nature tries to keep this phase difference at -90º or +270º where maximum excitation occurs. The only deviation from this is temporary, thus parametric damping is also temporary.

There is a delay of many cycles in response between the primary current and secondary current, thus the output is largely immune to input transients and off-resonance harmonics.

If my understanding of this functional principle is correct then the intent of the divided core is to have a loose magnetic coupling between the primary coil and the secondary coil.

Not quite. The reluctance of the flux paths must be low to reach the non-linear region of the core's BH characteristic (see Hopkinson's Law for magnetic circuits) in order to cause the resulting permeability/reluctance/inductance modulation.
Achieving loose coupling merely by high reluctance of air gaps defeats this purpose.

IMO the mutual inductance between pri/sec of the parametric transformer should be zero while keeping the reluctance of the two flux paths as low as possible.
This might sound like a contradiction at first, but upon deeper consideration it becomes obvious that this possible by orthogonal positioning of the two flux paths or establishing only a small common volume of the core where the pri/sec flux meet and crossmodulate, while disallowing the primary flux from ever reaching the secondary winding (and vice versa). The orthogonal double C geometry cleverly incorporates both of these techniques.

But it does not really matter how this loose magnetic coupling is achieved (by having a gap between the two half cores or by rotating one half core 90 degrees with respect to the other).

In light of what I wrote above it matters very much.
Loose coupling itself is not the ultimate goal. Min mutual inductance and max reluctance modulation, is.

So the current supplied to the primary coil is just to excite the secondary coil, which then produces its current (or voltage)

Be careful with the wording here, because the above sounds like a description of a conventional transformer, which I'm sure you have not intended.
CONVENTIONAL TRANSFORMER:
varying primary current -> varying pri & sec flux -> EMF in secondary -> secondary current -> tapping all of the secondary current by the load.
PARAMETRIC TRANSFORMER:
varying primary current  -> varying primary flux -> crossmodulation in common core volume -> varying secondary permeability/reluctance/inductance -> parametric excitation or damping in secondary LC tank -> diversion of some of the secondary LC tank's current into a load.

...by itself from thin air (due to parametric excitation).

The author of this paper does not admit that varying the permeability/reluctance/inductance of the secondary circuit does not require any power delivery from the primary or that it requires less power than power tapped from the secondary LC tank.

In fact the author rightly notices that the parametric transformer also works in the reverse.

As the symmetry of the core implies, the dependence of the primary reluctance on the secondary flux is assumed to have the same form

The initial oscillations in the secondary circuit create variations in the primary reluctance, the phase of these ensuring that maximum energy is drawn from the primary flux, and thus from the primary voltage supply, the energy being absorbed by the primary reluctance variations.

**********************************************

And furthermore, why is it that the toroidal shape of the yoke device reminds me on the toroidal shape of the Marks device? Don't say that's just another coincidence.

Toroidal shape is not a coincidence. It is the most efficient shape for magnetic flux.  It "likes" to form closed circles...

[Zeitmaschine]

The author of this paper does not admit that varying the permeability/reluctance/inductance of the secondary circuit does not require any power delivery from the primary or that it requires less power than power tapped from the secondary LC tank.

In fact the author rightly notices that the parametric transformer also works in the reverse.

But »unfortunately« this has been proven to be wrong by the working yoke device, which is -according to the diagrams- essentially based on the same parametric principle.

Or has anyone any idea why the web page of the yoke device has been removed (404) if not because this device actually worked?

BTW: It would be a bad idea to write about free energy in a doctoral thesis, wouldn't it?

Toroidal shape is not a coincidence. It is the most efficient shape for magnetic flux.  It "likes" to form closed circles...

Maybe it likes even more to form unclosed circles. This means it is also an efficient shape to cut the magnetic flux by a gap.

[verpies]

BTW: It would be a bad idea to write about free energy in a doctoral thesis, wouldn't it?

Oh, it would be an academic suicide!

However, he is correct that the Parametric Transformer with a symmetrical core, works in reverse equally well as forward.
However, there is no unbreakable rule, that it must work symmetrically like that, ...besides energy conservation.

If it is possible to prevent the secondary flux from affecting the reluctance of the primary flux path, but not vice versa, then it would be a hit.