Selfrunning cold electricity circuit from Dr.Stiffler

[armagdn03]

Nice post, one has to respect a post backed up by numbers.
I did all the math and you are definitely correct in your assessment,
Of the circuit, but I think that something has been missed. I could definitely be wrong on this one, and im sure ill be corrected. But, you cannot just simply stick a 50 ohm resistor across the inputs of the circuit and expect that to solve your impedance matching issues. According to the equations for adding resistances in parallel (counting one resistance as the 50 ohm resistor, and the other as the whole circuit behind it by Theviens Theorem) if you have one low resistance and one extremely high resistance the total resistance will be close to the low resistance, which should be intuitively correct. But I doubt that the resistance of the circuit is very high even with that 10k ohm resistor in place. The resistor is in parallel with the capacitor, and we all know that capacitive reactance is inversely proportional to frequency, and we are driving this set up at high frequency so the electric signal would bypass the 10k resistor. Now there are a lot of inductive elements in the circuit for which the opposite is true, but we would need to know their inductive values to know how many ohms of reactance they have at each frequency.

Basically, what im saying is that the resistive nature of the circuit at each frequency changes, and im not totally convinced that throwing a 50ohm resistor across the inputs will give you accurate enough results to base all of your math on, which by the way was correct, and I wish more people would take the time to work out the numbers! GOOD POST!

[armagdn03]

Also this brings up something interesting that I have not thought of before. Many cores have a resonant frequency based off of their physical dimensions, I never really thought about looking for resonant points related to the molecular makeup of the core. What is the resonant frequency of the iron in the core? What if one were to have high power frequencies driven over the same core one at irons frequency and one at bariums frequency, would we see a disassociation of the material much like what Hutchison saw? I have no idea, Im definitely going to have to do some research on this one.

[armagdn03]

Well, the video was pretty cool, but now that we have the schematic and parts list, it's easy to see that the circuit is powered by the signal generator.  It's capacitively coupled through C2, bypassing the 10K resistor, and then through the input capacitance of the 2N7000 MOSFET, which is quite high.

R2 keeps the gate-to-source voltage around zero, and the data sheet for that MOSFET (http://www.ortodoxism.ro/datasheets/vishay/70226.pdf, for example) shows that the added 1V p-p signal from the signal generator is not enough to make it conduct at all, so capacitive coupling to B4 and L1 is its only function.

The inductances, C1, and other capacitances form a resonant circuit through which the signal generator drives the step-up transformer formed by L2 and L3.

L3 transmits the power to the LED at high voltage through one wire, which is pretty cool, and the D1+D2+D3 circuit picks it up.  I'm not able to analyze this part very well, but I'd guess that it uses the stray capacitiances beween the parts of the breadboard that Dr. Stiffler has already mentioned.

All in all, this circuit is pretty wierd.  I wonder why anyone would build such a thing?

Cheers,

Mr. Entropy

I guess alot of what I said was already said.
If this is negative energy, entropy, then we should be able to remove the ?Plug? part of the circuit from the bard, and it will still light the LED distance not really being an issue.

[fritz]

mmmH,

Lots of things to bring together....
The impedance of the 10pf cap is 318Ohms at 50Meg
and 160Ohms ato 100Meg (thought it would be higher).
So this means a 5mW load which reduces the amount
of resonance voltage needed to suck even more power.

Another thing came to my mind:
The blue LED is operated pulsed with a frequency of
10MHz and a very short pulse duration.
Could be that the visual effect of such operation is
higher than the equivalent rms value DC operation
would be. Even the video camera can make the LED
shine brighter (on the recording).
My instinctive voice tells me that the losses increase
dramatically in this 1N4148/914/LED if operated in the
100 MHz range - a normal 4148 is slow as hell and stores
more charge than it switches (;-)))).
Maybe the state of the art blue LEDs have different behaviour.
So my contribution to that will be that I make an investigation:
Blue LED pulsed vs. DC, measuring the light intensity and
envelope with my highspeed photodiode.
Could be that the secondary voltage is rectified in the diode
itself, the transformed charge stored in the diode, 4148
works just as protection "resistor" hihi, hoho, hihi

rgds,

Wolfgang

[edork]

armagdn03 HI!

I had it all wrong, I do 50ohms transmitter with 50 ohm line and put 50 ohm load on it. Now I see, it is really 25 ohm, no wonder standing wave there.