youtube channel "skycollection" circuit, using a pnp and npn combination, bringing the circuit into auto oscilation
(not further related, posted because of interesting design)

(not related to above skycollections circuit)
its like a ab amplifier, but with a standing wave. so up and down at the same time.

connecting a pnp and npn so it switches negative and positive together as replacement for the sparkgap of a hairpin circuit

A joule thief circuit based on a 1:1 toroid, with a PNP and a NPN , so both pulses (from the toroid) can be used.

Would this work? I need a little help with this design.

The  npn and pnp should switch simultaniously. Short duration hairpin pulses.

V+ and V- would also connect to 2 capacitors , with the bifilar coil in between.
But ... this circuit is based on dc. does the hairpin circuit work with dc?

designing circuits isnt my strong point...

I've got a tip 122(npn) and tip127(pnp)  darlington(100V 5A) to experiment with. 

If it works, the V+ an V- should be made a lot higher(but with low amperage) This would require a higher voltage rating pnp npn pair.
they should be able to shortcut the power supply.  They also should be very fast. fast= better. in the end the resonant frequency should be matched to the pulses. (radio frequency transistors?)

the pnp at the above picture is wrongly connected.
the arrow (emitter) and collector should be reversed i believe

This is correct I think.

But I still dont get it. Its like my brain is on a holiday.

played withe circuit. it basicly works,but it doesn,t supply the pulses im searching for. it makes blocked clipped signals like a square wave.

connecting the pulses from a joulethief circuit (npn only) doesn't either. one side is stronger that the other side. (thats why i added a pnp)

Im goimg to try something different.

my igbt creates the right pulses, to make the bifilar coil ring at its resonant frequency.

if I have the resonant standing wave voltage from the coil, i could redirect the wave into another coil (with the same resonantfrequency)

setup; center coil pulsed by igbt. top and bottom coil signals taken so, they are 180 degrees out of phase.
both signals get 1 diode to rectify to positive and negative.

these 2  diodes then connect one to the center tap and one to the outerrim tap, of the middle coil of a
second stack of 3 coils (all tuned to the same resonant frequency of the first set of coils. (by adding parallel capacitors).

the resonant standing wave of the first set of coils now gets redistributed.

they are opposite polarity, and now also counter rotating from inside to outside and outside to inside.

top and bottom coil of the second set resonate but now without a polarised magneticfield. they can be closed looped, and experimented with.

never knew there where radio frequency relais. good to know.can be used to switch a capacitor with both leads between HV source and bifilar coil load.

i believe Nelson uses this in his nano pulser. joulethief can be used to power the relais

I've got the pnp npn opposite pulses working :D

I doubled the ring toroid windings, and mirrored the connection. 
It looked like a bug with its red tape body, and its 8 leads.

I didn't connect a load yet. but I should now be able to use the darlington tip 122 npn and darlington tip 127 positive negative pulses from the collector.

Now find some really fast transistors that can handle some current

I'm feeling a bit smarter again now it works.
But the frequency is  still very low relative to the resonant frequency of the bifilar coil.

The idea of the radio frequency relais, switching a capacitor, has another advantage. The capacitor that discharges into the coil from both ends, is at the same time tuning down the resonant frequency of the coil, because it is added in parallel. When the right size is choosen, the coil would be pulsed at its resonant frequency

Next step.
How to feed both pulses into the bifilar coil.
Connect it at the npn pnp collectors. Nope, still a magnetic field.
Is it because the coil is acting as a shortcut between the collectors of the npn and pnp?

adding diodes between the collector and the bifilar coil, would still create a shortcut

those tip darlingtons are slow. better replace them with BD240C and BD241C or something fast like that (not darlington)

Common emitter, would that work?
based on the double bifilar wound toroid seen above.

common emitter works!
the voltage can also be higher 1A 5V

i could also place the bifilar coil between the 2 emitters (without capacitors)

By adding 2 ultra fast diodes between the toroid coil and the collectors, I can provide the capacitors with High voltage.
The pulses(from double joule thief torroid), then create a discharge/shortcut of the capacitors, that energizes the bifilar pancake coil in between, like a tesla hairpin circuit.

The caps need to be matched to the bifilar coil.

the npn/ pnp not only short the caps, but also short the HV power supply. It needs to be a low amperage supply. Or I could use current limiting resistors between the HV + and - and both capacitors.

so far so good. I tried to connect the HV to the V but that didnt work.
Did add some uf4007 diodes, over the 2 resistors, to get the spikes better.
Also added the capacitors and the bifilar and it seemed to work. Tried several different capacitor sizes, and indeed it changed the resonant frequency of the coil. but it is added in series, not in parallel.
Could not get the coil to resonate properly. used a 3 coil stack, with center coil connected to the two capacitors. measured the resonating coil.

 Ah. now i get it. i didnt ad capacitors in parallel with the resonanting top and bottom coil. DUH... I saw a spike, and ringing...
I need to keep the 2 capacitors of the circuit, in a low value, pF range. and high voltage rating, I played with pulses around 75 to 100V.

I also replaced the darlington tip 122 and tip127, with BD240C and BD241C. These are high frequency transistors. darlingtons are to slow.
I could get higher frequencies. easily got 20kHz. maybe also add some diodes in reverse with the transistors over the collector emitter.

Never saw the needle move of the compass, but the voltage rise was to low to be sure.

I need a cooling body on the transistors. Higher voltage/amperage means higher frequency. Till now I worked with 5V 0,2A to keep them cool. this gave 20kHz. The higher i can get, the less capacitance I will need to tune the resonant coils down.

Those coils of Neslon his setup, would have a resonant frequency in the MHz range, but he also showed in his videos the process of tuning with differnt capacitors. The solid state pulses, with this toroid technique wont go high enough. But It is working!

sorry if the diodes are wrong. bit tired. diodes over R should be blocking (reversed).
diodes at collector should be open to the pulses.

the voltage spikes in the resonant top and bottom coil, seemed higher voltage then the pulses going in, but I cant remember clearly.
If it is , then its a good sign.

heatsinks, plus tuning with parallel caps is the nest step

heatsinks are in place. removed all the diodes again. signal on the collectors looks good.
Resistor on the base of the PNP, gets hot. signal between the coil and the resisor is good (positive pulse), but the signal between the resistor and the base is suddenly distorted, almost inverted (negative pulse). Still the PNP gives the correct pulse at the collector.
Maybe the emitter to the base current of the PNP (from the emitter of the NPN) leaks through toward the resistor???
to much power coming in from the npn into the pnp,  creating to much current from the emitter to the base?

ehhh.... i dont know. everything is in reverse with these pnp's I'm not used  to that. everything is upside down.

diode in series between the resistor and the base of the pnp. to stop the reversed current? It needs to go to the collector. not to the resistor at the base

need brains, eh no FOOD
:)

I have placed a diode in between the resistor and the base of the pnp. now the resistor doesnt get hot any more. I still wonder how it is possible that the pulse before the resistor is positive and after the resistor is negative.

It might have to do with the common emitters. its floating. maybe a earth ground would stabilise it. or 2 capacitors, one to plus and one to minus, with large resistors accros it, equal size, to create a stable virtual earth.

this is done because the NPN pulse seems to shoot up through the emitter of the pnp, and shoots out of the pnp base.

The diode might be needed due to the mismatch between the pnp and the npn. the npn is a bd241C rated at 40W and the pnp is a bd240C rated at 30W.

the diode used is a uf4007 (ultra fast).

The voltage sets the frequency. more voltage=higher frequency pulses/

With a 7.07V dc at 0,25A I get a 60V positive pulse and a 60 v negative pulse. one at the pnp collector and one at the npn collector.
The frequnecy, calculated from the scope: 3,5x10uS=35uS  1/35uS=28,6 KHz.

That is  a high enough frequency for the skin effect, but very low compared to the resonant frequency of a bifilar coil.

So far so good

R is now 650 Ohm,

For now, this goal is reached. I'm happy with the results. I now have opposite polarity simultanious pulses.
But now I will have to further investigate the longitudinal effects of the bifilar coil. And find out if it only happens at a fixed (high) frequency, or not.

to let it work properly the capacitors would need a extra voltage supply to charge, when the pulse isnt closesing the circuit to discharge the capacitors.

the extra voltage supply for the caps should not interfere with th toroid pulses

this uses 2 power supplies. one for the toroid (low voltage low power), to create the oscillations around 20-30khz
The HV (doesn't need to be KV, 100V?)  power supply is to charge the capacitors.
 when the pnp npn is not conducting (largest time) the caps get charged by the HV.
The diodes prevent the HV to flow into the toroid when the transistors are not conducting.

diodes are ultra fast (uf4007)

when the toroid gets the voltage(low), the base resistor creates a current, and opens the transistor. a big current starts to flow, through the collector side of the toroid, giving rise to a magnetic field in the toroid, that shuts down, the base current, because of its opposite polarity.

At the same time, the HV+ an HV- gets shorted out. Its only a short pulse, but still, the HV supply must be able to handle this short cut. Also the transistors need cooling, because of this shortcut.

Most importantly, the capacitors, (charged, at the relative long duration of no pulse). get discharged due to the shortcut(created by the pulse)
Due to the fact these are 2 capacitors in series, the bifilar coil, in between the 2 capacitor gets a pulse from both ends of the 2 capacitors. A pure dielectric discharge. fast and energetic. (hairpin circuit)

The whole setup would need to be tuned properly.  the transistors, would need to handle high voltage/high current, matched to the HV power supply, and the capacitors, need to be tuned to the time the transistors are off. to charge to a high enough voltage (pF range i guess for the 20-30 khz range).

then the most important... the bifilar coil. it needs to be tuned to that relative low frequency. A cap in parallel(what I normally tune with), would mean another cap in series with the system... eh...??? :shocked: most of my coils resonate at 600+ KHz...

And Why do I want all this? because these pulses bring  the coil in to its resonant voltage rise. At the same time, there is no magnetic field in the coil, only a dielectric (due to counter rotation of the pulses).

This would give a gain in energy, due to the resonance. that's what I want to measure.
The loss in this system, is the heat produced, from the shortcut of the power supply of the capacitors. but, even that could be engineered out. (use another pulse/transistor to disconnect the HV power supply, when the caps discharge)