Hi folks! I succeeded to repeat the experiment. The effect was less than at previous time though, and appeared on another frequency. What it depends on, i don't exactly know. The video about this repeating the experiment is the last on that page, number 4. The circuit is also now simpler:

    -uuuu-
    |        |-
    \        =
    |        |+
    -->1---

https://archive.org/details/ndischarge

The circuit diagram is in the video, but i will try to upload the original diagram also to the archive.org page.

Basically yes, there is a capacitor, a coil and a diode, and there is only switching, no external power source, yet voltage appears on the capacitor.

I'm well aware of the mosfet leaking, or maybe what i call leaking, you call capacitances passing power, but this is what i mean by mosfet leaking.

I did everything to measure the mosfet leaking, you see if you see the videos. I even tried to measure the input current of the mosfet, but it is infinitesimal, thus difficult to even detect. I used a circuit which had only one capacitor, no coil, and by measuring the voltage on the capacitor found that any mosfet leaking which there can be, can only cause more than 100 times less power than the power generated as an output. This was also done with the same frequency and duty cycle as the experiment.

But whatever mosfet leaking there is, all will end up in the capacitor. The mosfet's output capacitance empties itself to the circuit in each switching cycle. So the more frequent is the switching, the more the mosfet leaks, there is nothing more special to it in my simple mind. Thus with the ten times lower frequency the mosfet cannot leak more than with the ten times higher frequency.

To be more certain, one certainly may measure the leakage on different frequencies, but i don't think that it gives much more. In the last video i made a leakage test only by measuring the voltage on the capacitor with the frequency on which the effect clearly didn't work. Why i made only that test, is because i don't want to make people to do a test with only a mosfet and capacitor. Because this is dangerous and creates a condition similar to short circuit, which is a quick way to burn the mosfet and maybe even the signal generator.

Mosfet is very good for such applications. Some say mosfet switches faster than a transistor and its input power and leakage is much less than that of a transistor. When meaning by leakage both the capacitance leakage (or passing power) and current leakage.

Just my opinion, but I havent watched all your vids.

I think the pos bemf is passing through the transistor's diode and charging the backside of the cap.

irf 630 is avalanche rated, meaning its diode will pass pos current
http://www.vishay.com/docs/91031/sihf630p.pdf
Just my opinion but work it out for yourself
Check everything,

Hello,

you are pumping gate-charges via the Miller-Capacitance into the Drain-terminal , pumping it via the inductance to capacitor ( LC-tank)
if pulse is shut off, negative puls is created ( collapse-spike ). After a while you will damage your electrolythic capacitor.


 http://electronics.stackexchange.com/questions/83712/gate-capacitance-and-miller-capacitance-on-the-mosfet

also see Vladimir Utkin-paper how to gather negative charge. Here starting on page 58, but be sure only to use
metal-foil-capacitors

http://www.free-energy-info.co.uk/VladimirUtkin.pdf

Regards

Kator01

Here starting on page 58

Yes the buck converter generates neg energy.
A pos pulsed coil has a neg bemf.