Self running coil?

[NextGen67]

<...>
ADDED

I added another scope shot where I added capacitance to the toroid coil to lower resonating frequency so now the pickup coil now has .35vdc on the 1K Ohm load (twice the power) without raising the 150mv RMS across the 100 Ohm current shunt and still sending back -.000006

What do you get if you take out the pickup coil in the above test (still -.000006 or less or more), and how does the scope shot looks like then?

--
NextGen67

[gyulasun]

....
Also, take note of how low of voltage I can A this mosfet!... don't think it's not driving the gate as I checked and it is. Since the pickup coil has 0.22vdc on the 1K Ohm load.

This is the lowest RMS voltage (156mv) across the 100 Ohm resistor I got so far with it sending back energy.

What do you make of this? 

Hi Luc,

In your 2SK2806 MOSFET the socalled V_th gate-source threshold voltage is about 1/3 of that of the IRF640  i.e. between 1 and 2V only.  This means an advantage in input power wrt the IRF640 (the need of lower gate-source voltage means lower input power).
Here is the data sheet: http://skory.gylcomp.hu/alkatresz/2sk2806.pdf

ADDED

I added another scope shot where I added capacitance to the toroid coil to lower resonating frequency so now the pickup coil now has .35vdc on the 1K Ohm load (twice the power) without raising the 150mv RMS across the 100 Ohm current shunt and still sending back -.000006

Did you add the capacitance in parallel with the toroidal coil?

Did you have any capacitance in the pickup coil circuit to make it resonant at the toroidal coil's lower resonant frequency?  The two freqs have to be the same to get the most output.  Hint: at the best setup as you wrote above (150mV across 100Ohm etc), try to place capacitors in series with the 1KOhm load to reach a resonant pickup coil situation (or as Mark suggested, connect it in parallel and use an audio transformer to step down the output power from the pickup coil), to approximate best a matched condition between the pickup circuit as a whole and the 1k load.

Thanks,  Gyula

[gyulasun]

double post

[mscoffman]

Ok since I was asked: On the test #11 and 12# Scope pictures

Waveform Picture #1 â€" The weird one I don’t it looks like dual frequency,
the driving one and the resonant one are fighting or combining with each
other. Not very important.

Waveform Picture #2 â€" clipping of the top of the pickup coil voltage signal
due to the rectifier diode turning on. If you connected the bridge rectifier
as a bridge rather than using one diode from the bridge then both the top
and the bottom of the signal would be clipped. The way to look at it is the
coil signal is pushing up on the cap voltage and the cap under load is
pushing back on the signal squashing it back down. This shows that the
load resistor and the coil impedance is (badly) mismatched. If the bridge
was connected then only 1.4Vdc of the coil signal (squarewave)  would
be  showing that is .7 volts pedestal +| -.7| pedestal whipped around
by silicon diodes in the bridge. With the bridge, both sides of the signals
polarity try to push the cap voltage up rather then just one. The bridge
would be more efficient than a single diode…So connecting the full bridge
would improve efficiency of rectifying the pickup energy.  Signal total =
0.42Vdc pp peak-to-peak.

Waveform #4 â€" unloaded, there is no load resistor pulling the cap voltage
down. So the cap charges up to 11.2Vdc revealing the full pickup coil
waveform. Only a very small part of the signal pushes up on the cap
voltage. *No work is being done* by the pickup coil â€" so there is no
impedance mismatch….But now one sees a *voltage phase shift*
occurring between the driving (clock)signal and the pickup voltage.
What do we know about AC signals; When no work is being done the
voltage waveform and the current waveform are 180 degrees out
of phase. It be hard to see the current waveform when thing are
disconnected but we could see it with a one megohm load or something.
When an impedance mismatch favors more load the voltage and current
waveform will be 0 degrees out from each other or in phase and closely
aligned to the coil drive phase because work is being done. Note that
the signal phase *leads* the drive phase. I hope that means inductive
reactance but it may mean capacitive reactance. This tells “How” our
two coils are coupled to one another.


Waveform #x-
A little bit further on in the thread. We have waveforms from the
2SK2806-01 mosfet. Notice that the waveforms there are less flat
ie. less digital…Most likely this is due to the higher final drive
impedance of that mosfet so this means the IRF604 is a preferable
transistor. Generally that ST transistor is very newly designed so
will take advantage of the latest in mosfet design tech.

I want to correct the above as pointed out in gyulason. The gate
voltage is much better on this transistor...which may explain the
difference in output waveforms. So the 2SK2806-01 has preferable
gate characteristic...the Rd's is better but Cissy is still too high.

---

It's important to note. When you couple an generator to a load
*The maximum power is transferred between them when the
impedance (resistance) of the source and load are equal to
each other.*  Note that power = continuous  or instantaneous
energy.

So this is why I tend to emphasise impedance matching in overunity
projects. Mechanical impedance is exactly the same way. If we simply
use ballast resistors the energy in the resistor is converted to heat
and lost. As long as we are not converting it to heat it is either staying
in the circuit or coming back to the circuit as flyback energy...(a
beneficial part of the conservation of energy law).

:S:MarkSCoffman

[mscoffman]

Reason being the IRF640 needs quite some energy to actually switch it ON, so *if* some of this energy triples through the circuit, it will be amplified and in such a way *could* be responsible for the cap 'charge up'... However a replacement mosfet (STS2DNF30L) for example which requires *way* less energy to switch on, hence would *also* cause *less* energy triple trough the circuit... Now *when* the STS2DNF30L results in a *better* cap 'charge up' then the IRF640 it can be argued leakage is not the cause.

On the other hand, if the STS2DNF30L shows a cap 'charge up' that is far lower then when he used the IRF640, it can be argued that indeed the hidden leakage could be responsible for the cap 'charge up'.

Sounds plausible ?

--
NextGen67

Yes, this is correct. I am hoping to get a mosfet transistor that requires
*less drive voltage* and shows spiffy ie not sluggish Coil resonant effects.
Our mosfet gate drive voltage at ~11Vpp is too high in my opinion. We
are feeding our transistor too heavily for what it is doing and it's
passing through that power. Ok, boost/buck converter using the
parasitic inductance of the long clip leads...I don't care how it
accomplishes this task of raising it's bulk input voltage... If we feed it
less it's going to have less available.  The combination of lower drive
and lower mosfet capacitance is much lower possible energy pass
through potential, Therefore if the capacitor still gains our gain must be
occurring from overunity. Unfortunately light rider's 4N35
opto doesn't suggest this. With the 4n35 we only feed the transistor
with energy we already have in the circuit. Things seem to get sluggish.

The reason I like the pick-up coil funding the gate energy is that; The main
toroid coil is going to take care of itself. Energy put out will reasonate
back in plus maybe some overunity gain. Only if the pickup coil is coupled
to main coil unit for unit will we never be able to accomplish anything.
As we will push down the gate energy required, less will be taken but
still no overunity gain might be present to raise the bulk capacitors.

The other thing is if the transistor wants to see more gate voltage
for it's operation for some reason (maybe I'm not as hot as I think
I am on designing with fet's) we could use that transformer to push
up the gate voltage while we decrease inputted power. This is kind of
a secondary cross check to decrease gate drive power if we have
really succeeded in increasing mosfet input impedance. If we can drive
the transistor at the voltage of the signal generator based in overunity
energy...ie not drawing anymore from bulk. Then the bulk capacitor is going
to charge up in exactly the same way as it did with the signal generator.
We don't really care, at this time, what circuit elements are causing that
bulk voltage increase.

Also, gotoluc could put the 100ohm resistor into the gate circuit
and show that it doesn't effect things there. Then the current
waveform going into the gate of the mosfet could be directly
seen and be measurable. Also reductions or increases with other
transistors would be measurable as a change in current.

:S:MarkSCoffman