Negative discharge effect

[ayeaye]

TheComet, what matters is research, like possible overunity in the coil.

Maybe the mosfet leaks through resistance, and not through internal capacitances, as it is often said. Look at the horizontal region of the gate current on the currents diagram. What is this? This may as well be a continuous leakage from the input capaciance to the circuit, through a 42 k resistance, well, calculate. It does not show any correlation whatsoever with the circuit though. Is it a power necessary to attract the electrons in the mosfet? But whatever it is, it cannot be seen where it really goes, and that a part of it doesn't go to the circuit. These are the problems, we cannot see inside everything, thus also not possible to see for certain, what the real leakage is.

We can know what the maximum leakage is, this though is almost certainly a lot more than the real leakage. The only real leakage i think there certainly is, is leakage through the input capacitance when the mosfet switces on, which is very small though and lasts for a very short time. But it provides the intitial voltage to the coil, and starts that oscillation. The maximum leakage though can also be useful to know. One may try the simulation with coils with higher inductance, for example, and different frequencies and duty cycles. I don't exclude the possibility that with some of these there can be overunity even considering the maximum leakage.

I don't assume the overunity, i just don't exclude the possibility. This is what the research is all about, we assume nothing, just find out what happens. Just as Faraday did, he did it to find out how god made the world, because he was very religious. Well and, without him we didn't have the electricity today. And overunity, it looks like, is only some small effect, at least when using conventional ways of doing things. Later the small effect can be increased, like it has happened a lot in technology.

Like Faraday first saw only a slight declination of a compass needle, for a very short time. Something which some may not even notice, or consider some interference. To exclude that possibility he, btw, invented his Faraday cage, to protect his galvanometer from interference. But this slight effect is what all the electric motors and generators are based on today.

[TheComet]

Look at the horizontal region of the gate current on the currents diagram. What is this?

I'm not sure what you mean, can you please be more specific? Where is this diagram? What current doesn't make sense? Please point it out by downloading the image, drawing on it with an image editing tool and re-uploading it and explaining what doesn't make sense.

I can help you understand.

It does not show any correlation whatsoever with the circuit though.

Why? What is your reasoning behind this statement?

But whatever it is, it cannot be seen where it really goes, and that a part of it doesn't go to the circuit.

Baseless statement. Download the simulation and show me with your own measurements what you mean.

[ayeaye]

I'm not sure what you mean, can you please be more specific? Where is this diagram? What current doesn't make sense?

The diagram below, from the simulation results, showing the currents at the gate and in the circuit. You see the current drawn with red there, this is the current at the gate from the beginning of the input pulse, up to the half of the cycle. The pulse ends when the gate current goes to zero.

Look at the current drawn with red there, this is the gate current. It first goes down (it is drawn reversed), then up again, and then there is a horizontal region, it stays at the same value for some time, and then goes up again to zero. This horizontal region means continuous consumption of the gate current, not charging the input capacitance, which happens by an exponential curve. This means a continuous power which goes to the mosfet. We cannot see inside the mosfet, so we cannot see where it goes, whether it is consumed by the mosfet, or it goes to the circuit.

Correlation, the two things that are somehow connected, should show some dependence on each other. Like the gate current, when it leaks to the circuit, should somehow reflect the current oscillations in the circuit. It does not though, there is a perfectly straight horizontal line. I cannot help you to understand, what concerns things like that. This needs an analytical general thinking, the people who lack that ability, are never able to understand some things. And there are many such people.

[ayeaye]

I now also did a simulation of the same circuit, but with the inductance of the coil 10H. The energy going to the capacitor in a single cycle was only 0.7pJ, which is much less than the mosfet's leaking energy with the circuit with an equivalent impedance. Which by the thread here on how much the mosfet leaks http://overunity.com/16062/how-mosfet-leaks/#.VhBEDG943iY , is 0.7nJ.

This doesn't necessarily mean though that the output energy cannot be greater in a real circuit. As simulators don't always simulate everything so as it happens in the real world.

I also added the circuit diagram as it should be, below. One thing to notice there, the resistance of R3 is written there as 100000, not 10M. This is very important, because ngspice recognizes all suffixes such as k, m, u, n and p, but doesn't seem to differentiate between M and m, or doesn't recognize M. So the values with the suffix M will have a wrong value, and thus instead of using M, the value should be written either with a full number, or then with a suffix that is recognized, such as k.