OverUnity Research

QuoteSo the race is on.

3 AA energizer batteries or 4.5 V DC Input

100 Watt Bulb as load.

Frequency of operation about 1MHz

Must be a resonant system to increase the power factor and drastically cut the input reactive resistance of an ordinary transformer.

We could call it a TPU  ;D

Anyone fancy designing a circuit to try.

OR

is it as simple as placing a cap across a mains transformer primary and secondary and connecting the bulb across the secondary and cap

QuoteA pure reactance will not dissipate any power.

but how can you use a standard transformer if the resonant frequency is 1MHz surely we need an air transformer like Don used

My aim is to build a pocket version of Don's table top device, but driving the primary using a blocking oscillator.

Construction and Test Data coming Soon

First Job was to build the primary resonant coil for 1Mhz Now Done

Then the next job was to buid a fet based Blocking oscillator to drive the above coil. Now Done.

Next up is the secondary and capture circuit. Not yet Done

A picture of the blocking oscillator and 1.04Mhz Primary

Scope shot from GND to Fet Drain.

Note amplitude is 248Volts and the Frequency is 1.04Mhz, the resonant frequency of the blocking oscillator is set by the tuned LC

I am using my psu set to 125 Volts Current drawn is 6mA, i didn't want to go with the 5volts as above, i wanted to get a little closer to a tesla system but using an oscillator to drive the coil as large as i could.

Also note total power consumed is about 0.75 Watts

Here's a FFT Shot of my scope probe laying near the primary coil.

Here's the circuit

OOPS Note i missed out a 330K resistor in the original Circuit and C4 is 47nF not 470nF so i have uploaded the correct circuit now.

Sir ,
I've been inviting guests .
I hope you don't mind?

Chet

Hi Chet

No problem, thats why i document my builds whether successful or not, hopefully something can be learned along the way for others to build upon.

My intention here with this build is to study Don Smiths setup and see what i get.

With Don's device there are 2 secondaries, one is parallel tuned, but the other has no tuning but goes straight to a diode we have always suspected that there maybe in part a component missing to deceive us from the correct function of this second coil.

Don refers to this coil as the current supply to the system.

It has occurred to me that maybe this second secondary coil is infact series tuned.

QuoteIn a series RLC circuit at resonance, the current is limited only by the resistance of the circuit

QuoteAn equal magnitude voltage will also be seen across the capacitor but in antiphase to the inductor. If R can be made sufficiently small, these voltages can be several times the input voltage. The voltage ratio is, in fact, the Q of the circuit

So it would be relatively easy to hide a capacitor in series with the secondary coil, and when tuned this indeed could provide the current aspect to the final output, also note he uses thick wire for his secondaries, now we know why.

Anyway just thinking out loud.
You heard it here first  ;D , so 2 secondaries 1 parallel tuned one series tuned  ;)

Primary Coil details.

Wire Length = 430.8 cm
Coil Length = 29mm
Diameter = 16.5mm
Resistance @ 120Hz = 0.928R
Inductance = 49.95 uH
28 SWG Wire

For the secondary i want to use the magic 4* wire length which gives me 17.232 Meters
I will use a heavier wire maybe 22SWG

In which case i will aim for about 170uH which would require 137pf cap which will be good for using a low value trimmer for adjustment also.

If anyone is going to try building the Blocking Oscillator use a pot first for the bias resistor chain, set this to turn the fet gate fully off by adjusting for low volts and then adjust up to find the point oscillation starts, if you dont do this then there is danger that the fet is biased on to much and when you apply 125 volts the fet will burn.

Hi Peterae

Another trick I learned in building these oscillators is to add some resistance in series with the 10nf back to the gate so that when you hit the TVS diode conduction point you don't hog too much current out of the circuit.

This is important for blocking oscillators where high efficiency is required, if you don't care how much energy you burn in the "tickler" circuit then no problem.

As you may have seen in some of the circuits I have posted, I sometimes use a potentiometer divider to precisely adjust the feedback into the gate, rather than slamming it hard.

For a Don Smith circuit, this may not be an issue.

I have always liked playing with LC series circuits, and yes, the current can go to infinity in the ideal world.

Good luck

Hi ION

I did try a 5k pot in series with the feedback cap but having already built a 1 turn pickup into the coil, i could not get any adjustment, as soon as i turned the 5k pot even slightly i could not get oscillation back, maybe in hind sight i should have used a few more turns and tried this.

Also interestingly the gate is held way lower than it's hard switch on point to stop excessive turn on current.

Thanks for the advice  ;)

PS just ordered some 70pf trimmers for each secondary coil.


Peter

A Scope shot of the gate notice the DC offset

Here's an interesting thought.

Don's second coil fed a rectifier diode, which then connected to a capacitor bank, so would a series connected Inductor-Diode-Capacitor Bank constitute a series tuned resonant circuit. :o what effect would the diode have?

Quote from: Peterae on 2010.10.29, 21:05:07
Here's an interesting thought.

Don's second coil fed a rectifier diode, which then connected to a capacitor bank, so would a series connected Inductor-Diode-Capacitor Bank constitute a series tuned resonant circuit. :o what effect would the diode have?

The circuit as you described would probably not oscillate, as there is little to no current flowing in the reverse direction.

It would seem that in this case the diode capacitance would predominate in the reverse direction forcing a much higher frequency.

Still, it is an interesting idea. I once tried splitting a resonant tank into two separate inductors with  diodes. If i remember correctly it did resonate, but there were losses in the diodes.

Thanks for the reply ION , i may carry out some tests at some point and try a few different permutations.
It was a bit of a long shot anyway as the cap bank is either 5 x 2uF or 5 x 8uF in parallel.

This puts the fet in linear operation mode. Of course I am preaching to the choir.
Nice sine output!
Saw that with the audio circuit I built from Bolt's suggestion for my Trinity hammer / 3 speaker canon.
Side note: I play Pied piper with a community of voles behind my backyard fence. I am watching for birth defects.

Quote from: Peterae on 2010.10.29, 13:06:45
A Scope shot of the gate notice the DC offset

Thanks gk

I have built one of the secondaries it came out @ 123.43uH
It uses exactly 4 times the wire length of the primary for the turns.

I am waiting on 2 x 70pF trimmers to arrive, using a 150pf cap across the secondary and the trimmer across this should give me an adjustment range of 970kHz to 1.17Mhz.


Coil uses 22SWG wire

Having wound the last secondary coil i decided it was way too long, so i halved the wire length and then wound a new primary coil using a quarter of the primary wire length.

Primary Inductance is now 17.62uH
Secondary Inductance is 56.52uH

So new primary cap value will be 1.47nF
and secondary will be 440pf

I also made a new oscillator and cleaned this up using board, i will try a zener diode in place of the TVS Diode and reverse protection 1n4007 diode.

As you can see i still need to connect the board and check it's operation.

Peterae,

The best advice I can give you on simple RF tank circuits is the same info EM shot down as being incorrect. He cited the book but his book must have been several generations from the physical world.

For the best storage, use and exchange of energy between the capacitor and coil..... Think of the same simple LC circuit but put a transmission line between the L & C, for concept only.

The best operating LC tank circuit has matching 'at-frequency impedance' between the capacitor and inductor.

The book thumpers are quite correct in saying 'it doesn't matter'. It doesn't matter as long as you are not concerned with overly short oscillations due to energy loss in the tank circuit.

I communicate via PSK to Antarctica and South Africa on a regular basis with less than 1 watt using this understanding and the same long-wire antenna and tank circuit that charges my station batteries.


And no, simply adding a capacitor to a transformer has little to do with gain of the system. Otherwise, my main bench isolation transformer would be OU. Here is the schematic and photo of the strange winding arrangements.

QuoteThe best operating LC tank circuit has matching 'at-frequency impedance' between the capacitor and inductor.

So you are saying i should be matching the impedance of the capacitor and inductor at resonance.
I need to look at this then.

Thanks WW
PS that transformer is way over my head, i cannot understand why some windings are Open ::)

So for the Primary @ 1 Mhz I Have

L = 17.62uH
C = 1.47nF

Xc = 1/(2 x pi x f x C)
Xc = 108.2687 Ohms

and

Xl = 2 x pi x f x L
Xl = 110.709 Ohms.

Wow don't ask me how, but it a fairly close match.

Secondary

L = 56.52uH
C = 440pf

Xc = 361.715 Ohms
Xl = 355.126 Ohms

thats pretty close as well, probably even closer if i allow for the trimmer to tune for the 1mhz resonance.

So how come they are close, is it becausewhen i used the resonant formulas to calculate my cap value for a given inductor that the impedance automatically gets figured into the equation.

The normal math will make it happen anyway. L & C will end up equal (but opposite).

Resistance will take up the differences you see, or make it worse  :D

Never mind. You were doing it right. I should have assumed you were doing it right  :)

Hi WW
No in fact thanks for the heads up.
I hadn't even thought about the reactance values and certainly a good idea to calculate them.

Just to be clever i thought i would use Don's suggested charts as i have the book at hand.

I have attached a scan, with Don's desktop device against my device, it's interesting his device has much lower reactive Ohms.

Peter
PS i wonder how the ratio of primary reactance against the secondary reactance alter the performance of the energy transfer.

If it would be possible to transfer the energy in the primary to a lower reactacne value in the secondary, i wonder if there would be a net power gain.
It's also interesting that this should be an efficient method of transferring energy from one impedance circuit(Primary setup)  to another impedance value system(secondary) without the normal mismatch losses.

OK tested and tweaked Coil and Oscillator Build 2

I am using a 18v Zener in place of the TVS and 1n4007 diodes this time.
The coil on this one is a quarter of the wire length from coil 1

In operation it draws more current @25mA and i need to get a heat sink for the fet because as the temperature starts to rise slowly climbing the until an untimely death would occur.

The coil get's hot after a while as well, and the feedback pickup has 3/4 of a turn.
A few resistor value changes and an increase in the gate capacitor.
The resonant frequency of this one is just above 1Mhz, i decided to use a 1n5 which brought it up to 1.14Mhz

A Technical question, my second builds primary is a quarter the wire length of my first build, is this why my circuit is drawing a lot more current or is it because i am not using the correct component values to bias and drive the gate properly.
I notice my gate has a slightly larger sine on it than build one, i did try a pot and to get consistent oscillation it did require the resistor values i choose, maybe the pickup coil is still too much at 3/4 of a turn.

Quote from: Peterae on 2010.11.03, 20:40:41
A Technical question, my second builds primary is a quarter the wire length of my first build, is this why my circuit is drawing a lot more current or is it because i am not using the correct component values to bias and drive the gate properly.
I notice my gate has a slightly larger sine on it than build one, i did try a pot and to get consistent oscillation it did require the resistor values i choose, maybe the pickup coil is still too much at 3/4 of a turn.

Hi Peterae

Your circuit has a resonant tank as the main coil. This will not perform as a blocking oscillator. It will perform as a nice class C oscillator.

If you want this circuit to perform as a blocking oscillator, you will need more gate drive, not less. Also you would need to operate it without the tank capacitors.

From the earlier pics, it looked like you barely had enough drive to force the gate high and low enough. looks like your FET is operating closer to Class A, hence the high power dissipation.

If your output is a sine wave, you are not hitting the gate of the FET hard enough. The more you reduce the gate drive coil turns, the less drive you will have until it stops oscillating.

Better to have a little extra drive, but adding resistance in series with the gate coil slows the circuit down. With the existing circuit arrangement, and at 1 MhZ there may be a limit to how fast you can charge and discharge the gate of the FET.

The 47nF is sucking up all your drive current.  

I prefer a slightly different circuit arrangement with the drive coil not grounded. The coil instead is inserted between the gate and the bias network. This eliminates the 10nF cap. The placement of the 47nF is slowing the gates ability to swing. By re-arranging the circuit the gate will not "see" the 47nF except as part of a stiff bias supply. Keep the protection diodes close to the gate.

If this is not clear, I'll try to supply a schematic. Sorry I didn't catch that 47nF problem before.

Edit: OK posted the schematic. This is what I was getting at. Now the 47nF is in the proper place and not sucking base drive. Note: designators in my circuit are arbitrary and do not conform to yours.

Hello ION

Thank you for the time you have given to this, i am sure there are many that will find this circuit invaluable for future experimentation.
I will built it and post the details.

Quote from: Peterae on 2010.11.04, 10:46:28
Hello ION

Thank you for the time you have given to this, i am sure there are many that will find this circuit invaluable for future experimentation.
I will built it and post the details.

Hi Peterae

Let me explain the circuit a little further. The driving end of the coil, that is, the end connected to the gate now will "see" only the gate capacitance as a limiting factor. The end which you had grounded is now the biasing end and sits on a "virtual ground" as far as AC frequencies are concerned. The bias voltage (DC) is applied to this end via the voltage divider or pot. Keeping this end "stiff" with the 47nF to ground insures that all current generated by the drive coil is sent to the gate.

What you need to decide is: Do you want the circuit to be a blocking oscillator (rectangular pulses out) or a resonant tank oscillator (sine wave) or some combination thereof?

Typically, blocking oscillators are operated such that the driving transistor is fully on and fully off for each cycle. Also typically there is no resonant tank (LC) load, just the L. The circuit I posted shows no capacitors across the drain connected L, hence is a typical blocking oscillator. It will produce rectangular pulses at the drain of the FET.

If a capacitor is added across the L. it becomes a resonant tank oscillator, and will produce a nearly pure sine wave at the drain of the FET.

A resonant oscillator will drive the transistor less hard and the transistor will typically be biased inside the Class A region of an amplifier. If you try to drive the transistor harder, you will begin to distort the sine wave, and the circuit becomes a hybrid resonant tank / blocking oscillator.

In any case, be sure to maintain proper polarity of the gate drive coil or the circuit may operate in a parasitic mode.

Best of Luck

Ok ION i see, my understanding of the Blocking oscillator was wrong in that case, i am after the tuned pure sine wave output across the primary coil, as you have posted an incredibly simple FET based pure sine oscillator i would like to explore and build your circuit and i am sure others will also follow. ;D

I have had some real good fun playing with the fet oscillator so far ;)

There is another form of this resonant tank oscillator called the "tuned plate-tuned grid" oscillator from vacuum tube days.

In this case the resonant tank is calculated for the desired frequency and a small variable tuning capacitor is placed across the gate drive coil in order to tune it and allow it to "resonate' at the same frequency.

I have now connected my secondary caps and trimmers to each secondary coil.
Here's a snap of the setup
(http://www.overunityresearch.com/index.php?action=dlattach;topic=397.0;attach=2479)

And here's a video of me pulling the primary in and out of the 2 secondary coils, each secondary is being scoped on chan 1 and 2
The scope is set to 100V/Div

@ ION

Would you know if it's possible to have a fet oscillator that uses the LC to set the resonant frequency of the oscillator but does not require the pickup coil for feedback.
Maybe a capacitor coupled to the drain as feedback?

Thanks
Peter

The criteria for an oscillator is that you need a gain > 1 in the amplifier and >180 degrees of phase shift, which the pickup coil provides.

Your pickup coil can be located some distance from the resonant tank, since the tank will be transmitting to the pickup coil so this should not be a problem. You might need more turns on the pickup coil if it is located a distance from the tank as the drive signal will fall off with distance and you need to replace the lost voltage gain.  Also tuning the pickup coil with a small variable cap will  up the gain.

Note also that physical orientation will change the phase of the pickup coil so be sure to keep the vertical polarization of the coils in alignment.

If you absolutely must eliminate the pickup coil, you might consider a colpitts oscillator, a phase shift oscillator if you have enough gain in the circuit, or a Wein bridge.

Quote@ ION

Would you know if it's possible to have a fet oscillator that uses the LC to set the resonant frequency of the oscillator but does not require the pickup coil for feedback.
Maybe a capacitor coupled to the drain as feedback?

Thanks
Peter

A capacitor from the drain probably will not work, as it does not provide the required phase shift, it is negative feedback instead of positive feedback

The criteria for an oscillator is that you need a gain > 1 in the amplifier and >180 degrees of phase shift, which the pickup coil provides.

Your pickup coil can be located some distance from the resonant tank, since the tank will be transmitting to the pickup coil, so this should not be a problem. You might need more turns on the pickup coil if it is located a distance from the tank as the drive signal will fall off with distance and you need to replace the lost voltage gain.  Also tuning the pickup coil with a small variable cap will  up the gain.

Note also that physical orientation will change the phase of the pickup coil so be sure to keep the vertical polarization of the coils in alignment.

If you absolutely must eliminate the pickup coil, you might consider a colpitts oscillator, a phase shift oscillator if you have enough gain in the circuit, or a Wein bridge.

Just an idea ION, i was not sure if it was possible Thanks for your reply anyway.

I have now added a diode of each end of each secondary as in Don's table top and these feed a 2uF cap, i then use a 90V spark gap to drive a 12V 5 Watt bulb, the gap fires like mad the bulb does light but not very brightly.

Psu was set to 125V @35mA
4.375 Watts

SG is on left and bulb filament on right

Phase shift oscillator seems to require RC for timing, but the Colpitts uses the capacitor but center tapped, now this is interesting.
(http://www.play-hookey.com/oscillators/images/colpitts_oscillator_cd.gif)

but then the article goes on to say biasing resistors are needed

(http://www.play-hookey.com/oscillators/images/colpitts_oscillator_cb.gif)

Those circuits get messy, if you've got a lot of time and FET's in you parts bin you might get something working.

I think the circuit you now have or the ones I supplied are the cleanest approach.

OK ION thanks for the advice, i will soon build your circuit up as my build2 is proving temperature dependent.Doesn't always start from cold.

It is a good idea to make sure you've got the voltage divider for the FET bias worked out. I prefer to temporarily replace the resonant tank with a resistor, say 10 K, then make sure the divider has been worked out to supply the required 2 to 8 volts to the FET gate. This is why a potentiometer is better as you can precisely adjust the turn on threshold of the gate.

The divider string will change as supply voltage changes, so if you plan on a lot of different voltages, best to zener regulate the divider. I'll put together a good circuit I have tested.

I fully understand what you are saying, the regulated reference supply for the pot shouldn't be needed for me as i will always use 125V, i will use a pot, i could even leave the pot in circuit this time.

Here's a preferred circuit. Actually R2 can be much lower than 100K, low enough to bias up the Zener to 10 volts.

That is a pretty handy circuit ION thanks for sharing with us  ;)

Quote from: Peterae on 2010.11.05, 13:16:10
Just an idea ION, i was not sure if it was possible Thanks for your reply anyway.

I have now added a diode of each end of each secondary as in Don's table top and these feed a 2uF cap, i then use a 90V spark gap to drive a 12V 5 Watt bulb, the gap fires like mad the bulb does light but not very brightly.

Psu was set to 125V @35mA
4.375 Watts

SG is on left and bulb filament on right

Interestingly, the youtube link does not show up in your post #37, but shows in the quoted text, only before I save the quote, then it disappears. I copied it out of the text quote, here it is. Seems to be a glitch in the program.

http://www.youtube.com/watch?v=YHmtAEU1rgQ

As a control maybe take some light meter measurements of your bulb at various power levels directly off your power supply. I like to use a distance of one meter through a darkened cardboard tube. You now have a calibrated light source. Then hook up to your circuit and using the light meter / tube again and you can reference the power level output versus input.

That is strange shows for me ok in my post, anyone else having trouble.

I've had a few hours playing with my build so far today.

I ended up driving a 2.4V bulb in series with a 90V Spark gap, it's quiet bright, i added another 10uF cap to the cap bank bringing in a total of 13uF now, the bulb then flashes at a number of hertz(Difficult to judge), but this gives me an indication of how much energy the tuned system is provided to the bulb, all i need do is change tuning and config to look for faster flashes which inturn would indicate more captured energy from the system.

I have not done any calcs, but it is definitively way under unity.

I have tried all sorts of coil cap configs and can conclude as follows.

2 tuned secondary's do not provide as much power as 1 tuned secondary no matter how much you play with the primary position and tuning of both secondary's, i believe the reason for this is that when 2 identically tuned coils are placed next to each other they will inherently pull each resonant frequency away from the tuned primary frequency.
This actually is quiet interesting because when you scope the 2 unloaded secondaries there's equal potential in each secondary, so if you can draw 1 watt from one secondary what do we need to do to draw 1 watt from the other at the same time and add this together to get 2 watts.What ever the answer i could not do it.

I tried one secondary with series capacitive inductive resonance and it made no addition to the net energy whats so ever.

I tried one secondary with no tuning at all as in Don's bench top coil with no cap and again it appeared to make no more gain to the net energy whats so ever.

Not sure where to go with this, maybe tune one secondary to a harmonic on the other secondary

Quote2 tuned secondary's do not provide as much power as 1 tuned secondary no matter how much you play with the primary position and tuning of both secondary's, i believe the reason for this is that when 2 identically tuned coils are placed next to each other they will inherently pull each resonant frequency away from the tuned primary frequency.
This actually is quiet interesting because when you scope the 2 unloaded secondaries there's equal potential in each secondary, so if you can draw 1 watt from one secondary what do we need to do to draw 1 watt from the other at the same time and add this together to get 2 watts.What ever the answer i could not do it.

maybe 2 tuned secodaries just introduce more losses to the system hence less output

If i build a coil tuned to the 3rd harmonic of my 1Mhz driver coil, will it resonate at the 3rd harmonic, if not what do i need to do to get it to.

I tried just the trimmer across my secondary coil, using math i worked out it should resonate at the twice the frequency, but it didnt, all i had was the 1mhz imposed from my primary?

Do i need to drive the primary using a square wave or something.

Peterae,

You are trying to build a passive frequency multiplier and I suspect your signal source may be fairly clean of harmonics.

In any case, the energy found in the harmonics will be quite low compared to the fundamental. If you wish harmonics then you need a dirty signal (distorted, clipped, square wave with switching transients, etc.) A spark gap is a good switch for generating harmonics.

The tank just passes or attenuates harmonics.

I have some notes on a functioning passive frequency multiplier to share. It may be a bit. I'm into 17 hours at work this day.

Thanks WW
I did suspect this, I thought about using my white noise generator, but the frequencys are too high that i require.
Not a problem, now you have confirmed what i thought i can research my problem.

Cheers for your time, i know you are a very busy man  ;)

Your system is not driving anything that has a mechanism for gain.

You might try changing one of the secondaries to a tertiary magnifier coil and locate it so it is not inductively coupled to the other two coils.  Now you will have a two-coil Tesla transformer driving a magnifier coil.  Ground on leg of the secondary and the other leg of the secondary goes to the tertiary coil and the other and of the tertiary coil has the large surface area top load.  You can bring the ground and top terminal together for plasma between them.

Tesla also stated in the CS Notes that you can drive two tertiary coils with a standard transformer, with a ferrous core, and achieve any potential desired.   Tesla states that the rise is pressure is due to the factor pL/R.  p is the spark rate per second, L is inductance, R is resistance of magnifier coil.  He also explains that high self capacity absorbs the energy and ruins the magnification.

Quote from: Grumpy on 2010.11.14, 15:45:09
Your system is not driving anything that has a mechanism for gain.

Just curious.... 'Gain' as in antenna-like? (meaning no increase in energy just tighter focus). Or, gain in energy?

Quote
p is the spark rate per second, L is inductance, R is resistance of magnifier coil.  He also explains that high self capacity absorbs the energy and ruins the magnification.

So he was specifically limiting the tertiary coil to be a planar, single layer, with series opposing bi-filar? (lowest self-capacitance and typically lowest SRF while minimizing inductance and resistance).

Since the idea of the infinite negative inductance can only work with an open-ended inductor, your suggestion makes some sense.

That is interesting advice G Thank you, i am now trying to put these things into my next 3 coil design.  ;)

The problem i have then , is that i will wind for bifilar which will no doubt cause high self capacity(So this looks bad)

Following what WW said maybe i should then wind a quad coil and connect in pairs to form a bifilar version of Tesla's special coil but solenoid in type.

I also need to get as lower self resonant frequency for a small amount of wire as possible but i need to use bifilar to have 1 coil driven by a fet and the other to be mutually coupled to the driven coil.

Maybe it would be better for me to build this device flat using pancake coils, this way i can place the 2 identical coils on top of each other, while still maintaining matched self resonance for each pair, the problem i have with this arrangement is that i dont know of any device being quoted as successfully OU in this config.

Because i am aiming for low frequency 35khz,70khz and 105 khz maybe instead of pancake i could wind 1cm think coils but large diameter, i could still use Tesla's 2 wire coil arrangement, i could do some tests to see which gives me low SR while using less wire.

I was going to use 12 inch pipe to wind on.

Quote from: WaveWatcher on 2010.11.14, 16:12:06
Just curious.... 'Gain' as in antenna-like? (meaning no increase in energy just tighter focus). Or, gain in energy?

So he was specifically limiting the tertiary coil to be a planar, single layer, with series opposing bi-filar? (lowest self-capacitance and typically lowest SRF while minimizing inductance and resistance).

Since the idea of the infinite negative inductance can only work with an open-ended inductor, your suggestion makes some sense.

Gain in energy.

Yes, the third coil was planar, windings slighty separated.  Tesla used various configurations but Dollard shows the math in one of his charts that when the coil length is the same as the coil diameter, then the self capacitance is at its minimum and the self-inductance is at its maximum.  A coil with very high self-capacitance can never gain energy.  "The Buzz" specifically mentioned this in reference to Thane's early e-core transformer thing, which Buzz replicated.

The value we call inductance is also related to the space around the coil.  The fields are in the space, yet coupled to the coil, so they are accessible.  For example, place a ferrite core in the coil and increase the permeability of the space within the coil and increase the inductance of the coil.  People like to see the physical core, but overlook that fact that it alters space.

No bifilar.

Resonance was just a means to build up the potential from a small potential to excite the magnifying coil.  Thus, as Tesla stated, you can often do better with a standard transformer.

Quote from: Grumpy on 2010.11.14, 17:02:30
windings slighty separated.

I should have included that. The conductors need to be spaced 2-2.5 X conductor cross-section apart center to center. Otherwise, turn to turn capacitance is high (not negative inductance).
Yes, I do suggest pancakes and Yes, I do suggest bi-filar but wound to minimize capacitance.

Peterae,

Design a pancake coil to meet your needs without worrying about resonance frequency, not bi-filar. Maximize conductor cross-section and minimize coil turn count. Add the turn spacing I mentioned, above. Tell me what the actual coil SRF is (without a capacitor added) and what conventional resonance frequency you wish. I'll show you how to lower the conventional RF resonance point to your preferred frequency, or close to it.

BEP


Edit>> Hopefully we can get this done before I start a new week at work  ;)

ok so it looks like i am going flat coiled  ;D and would you belive it the Mrs has just asked me to make pancakes, wow she was lucky i nearly sneaked away to my lab  ;D

WW so if i was to make a 6 inch flat coil, are you saying you can tell me how to make it 35kHz no matter how much wire i used :o

OOPS no just reread your post, ok first off this device i am building needs to rely on the SRF of the coil to work, no adding caps.

A rise in voltage via a transformer effect can not explain the rise in voltage by the magnifying coil.  It is not even coupled inductively to the other coils.

There may be a minimum voltage threshold.

yes i see G, everyone is trying HV spark gaps ect, this is OK if you want a knock on your door and have all your equipment taken away.

I am convinced the same principles can be used to make a LF solid state version, (TPU) and when we work it out, i bet they all work on the same principle.

I have a feeling a LF device is capable of capturing more power than a HF device as well(Don used LF but never divulged the true design), theres only one way to find out, the thing with LF is the parts are cheap. OK so i will need more wire and the coils will be harder to make, but this is only a matter of mastering.

I am trying to settle on a coil design to get me started, trouble with WW pancake is i cannot think of a working device that uses this method of coil setup.

Maybe i should just use solenoid coils but large maybe 12 inch diameter, i need bifilar, because one coil is driven by the fets and the other is used to mix the 3 harmonics together, i have isolated them this way because i am driving each coil, where as the HV guys are using a dirty Spark Gap to excite all 3, my system should be much more energy efficient because i am not spewing noise across the whole spectrum.

The annoying thing is that i need a degree of tuning the SRF and the only way i can do this is to place taps every now and then :(

Quote from: Peterae on 2010.11.14, 19:51:25
OOPS no just reread your post, ok first off this device i am building needs to rely on the SRF of the coil to work, no adding caps.

There are limits on how low you can go.

35.705 kHz or anything below around 300 kHz is very difficult.

Better go ahead and make those pancakes for the wife but before you go... tell me what size and material solid wire you have. Hopefully, greater than 1mm diameter.

Quote from: Peterae on 2010.11.14, 20:53:37
I am trying to settle on a coil design to get me started, trouble with WW pancake is i cannot think of a working device that uses this method of coil setup.

I'm pretty sure the TPU uses the method I will help you with.

Quote
The annoying thing is that i need a degree of tuning the SRF and the only way i can do this is to place taps every now and then :(

You will need to trim one end until things are best. The idea is to design a coil that audibly resonates around 6 kHz but has an SRF in the extreme high ultrasonics and a magnetic resonance near the VLF band.

Kinky, no?

And you won't need any caps  :)

Edit....

Working on an example in case the World demands our attention.....

Well the pancakes were lovely   :P

Sounds good WW you talked me round.
I will buy what ever wire i need,i understand the trimming of the wire to tune, but at such low frequency's i was worried about not being able to tune each coil until in it's final resting place, which would be when all 3 coils are made and placed together then the tuning would be done i guess.

Am i still able to have 2 coils for each frequency, one being driven by the fet and the other matching coil being excited by it.

Frequency wise, i guess it doesn't matter too much, i was hoping for 35kHz fundamental because of all the talk in the past with the TPU but i guess we could go 400, 800, 1.2MHz if it makes it easier to build.

QuoteThe idea is to design a coil that audibly resonates around 6 kHz but has an SRF in the extreme high ultrasonics and a magnetic resonance near the VLF band.

You really seem to know where this should go, i guess you dont have enough time to build yourself.

QuoteAnd you won't need any caps  Smiley

;D

maybe we should switch to the proper thread for the new device.
http://www.overunityresearch.com/index.php?topic=417.0

I'll make the next posts to that thread.

I'll take you through getting one coil 'resonating' at oddly low frequencies without caps. You'll have to expand from there.

World calling at the moment.

Later

Quote from: Peterae on 2010.11.14, 20:53:37
yes i see G, everyone is trying HV spark gaps ect, this is OK if you want a knock on your door and have all your equipment taken away.

I am convinced the same principles can be used to make a LF solid state version, (TPU) and when we work it out, i bet they all work on the same principle.

I have a feeling a LF device is capable of capturing more power than a HF device as well(Don used LF but never divulged the true design), theres only one way to find out, the thing with LF is the parts are cheap. OK so i will need more wire and the coils will be harder to make, but this is only a matter of mastering.

I am trying to settle on a coil design to get me started, trouble with WW pancake is i cannot think of a working device that uses this method of coil setup.

Maybe i should just use solenoid coils but large maybe 12 inch diameter, i need bifilar, because one coil is driven by the fets and the other is used to mix the 3 harmonics together, i have isolated them this way because i am driving each coil, where as the HV guys are using a dirty Spark Gap to excite all 3, my system should be much more energy efficient because i am not spewing noise across the whole spectrum.

The annoying thing is that i need a degree of tuning the SRF and the only way i can do this is to place taps every now and then :(

Yes spark gaps are noisy, but I have experienced NO interference with just avalanche transistors and no gaps.

Your own experiments with the bifilar coil showed that it produces a compressed pulse of higher magnitude.  HV from LV.

OK. Put a large trashcan outside the window of your lab.  Open window.  Throw all resonant stuff in trashcan.  Resonance is just means to build up something small into something larger.  There is not OU there.  It is just a child in a a swing.

Yes for LF.

Eric Dollard used pancake coils in one of his videos, but he drove them with 20kv from a Tesla Transformer (pancake therapeutic type).

In the end, you are just hitting a CCU with HV  pulses...  ;)

Test question:  What "change" to space is required to "magnify voltage"?

Hint: 

What is inductance (L)?

How does permeability affect inductance?  Increase the permeability and you increase the self-inductance.

It magnifies itself, so to speak.