Charge conserving Capacitive Spring.

[Armagdn03]

To take full advantage of this effect, we must know and understand how flux density, permeability and the non linearity of cores work.

This was mastered a long time ago, and the devices which took advantage were known as Satruable Reactors, and Magnetic amplifiers.

Below is a graph of the B-H curve of a permeable core.



We can see as the ampere turns per inch increases, the lines of flux per inch changes in a NON LINEAR way.

Thus there is a portion of the curve (the more vertical section of the curve) which allows a small change in the ampere turns per inch causes a large change in lines per inch. Said another way, a small change in the magnetic field applied to the core, causes a LARGE change in the permeability of the core (and hence a large change in the inductance of a coil associated with the core).

Thus we want to BIAS our core in the unsaturated region (A) so that a small change in the input causes a large change in permeability. See section A of graph below:



This amplification effect can be seen in the graph below, where an input is shown to affect an output.




In the case above, a control winding is used to alter the permeability of the core. An inductor with an AC current through it sees a change in the impedance of its winding. Thus a small change in input current causes a large change in current through the secondary winding, this is how a mag amp works!

Lets take a look at the Alexanderson # 1,328,797 figure 1.



Notice the two windings, and two flux paths.

If the internal winding (labled in figure as: #2) is wound in such a was as to have NO mutual inductance with the other winding (#4) then there will be no transformer action. This can be accomplished with two mirror windings. Winding 2 will create a flux circuit which circles the two smaller legs around which it is wound. The winding 4 will see flux paths in two opposite directions within its center, and so will have NO EMF induced.

Therefore, winding 2 can create a change in inductance for the core, without affecting winding 4. This can operate the other way around also. The important thing to take note of: The two windings have NO mutual inductance. They do not operate as a transformer. One changes the impedance of the core only.


The issue is this....

If you have a control winding used to change the inductance of a secondary winding, no EMF will be induced on the secondary with only a change in inductance to cause it.

But, if the secondary sees a change in a coherent magnetic field, then an EMF will be induced.

Therefore magnets will serve two purposes when added to a magnetic amplifier.

1) The magnets will bias the core, bringing it into the region where a small change in input creates a large change in permeability.

2) The magnets will create a coherent magnetic field within the core. As the permeability changes, the flux density within the core changes as well. This creates a change in B field with respect to time for the inductor winding to see! Now instead of doing nothing as the permeability is changed, an EMF is induced.

I hope I have explained this well, and the ramifications are understood.

[Armagdn03]

Once this has been thoroughly digested, I suggest the reader look over once again this post from Eric Dollard.

I have a device, built for the Army Air Corps during World War 2, A/N number PP-18/AR Power Converter, which self-sustains the electrical system in my car. It uses the same theory of operation as Chris’s device but involves a different mechanical implementation utilizing a vibrator, several capacitors and 12V and 24V batteries that are connected in parallel through the device, rendering them as one.

I had a young student from Korea visit me a few years back. He had no problem understanding the basic concept of producing an energy synthesizing apparatus, because his mind was uncontaminated by all of the Bedini/Bearden falsehoods. The term Scalar Wave is an oxymoron, as scalar is part of the propagation constant that is NOT A WAVE! (Idiots!)

Most are clueless about the importance of the Variation of Inductance and Capacitance with respect to time – and synchronous parameter variations. Read chapter 21 (XXI) titled REACTION MACHINES in Charles Proteus Steinmetz’s book titled “Alternating Current Phenomena”. There is also a Russian paper (brought to me by the Korean student as a gift) titled: “UBER DIE ERREGUNG VON ELETRISCHEN SCHWINGUNGEN DURCH PARAMETERAENDERUNG” von L. Mandelstam und N. Papalexi, published in 1934 in: J. ZEITSCHRIFT FUR (umlaut on the U - as should also be on the first U in the title of the paper) TECHNISCHE PHYSIK Band IV, Heft 1, that continues with what Steinmetz teaches in his books, and takes it all the way (Title translation: Concerning the Excitation of Electrical Waves Through Parameter Changes). In one picture in the paper, there appears to be a brightly glowing incandescent lamp connected to a network, with no apparent connection to a power source. It appears to be an Alexanderson type Mag. Amp. operating in a self oscillation mode. (Alexanderson Patent # 1,328,797 Jan. 20, 1920): Even though my copy of the paper is in Russian, the equations speak for themselves and echo the work of Steinmetz and Alexanderson. Ernst Alexanderson emigrated to America because of Steinmetz’s book, - he was determined to work with Steinmetz after studying it. Steinmetz was forced to reverse many of his equations in later books and was severely criticized by physicist Michael Pupin of Columbia University for not using Maxwell’s ideas and instead developing a methodology that was actually useful and practical for engineers. (Read, “Steinmetz, Engineer and Socialist” written by Ronald R. Kline.) Here it was said that General Electric gave Steinmetz permission to create Electricity form the square root of minus one…

[SuperCaviTationIstic]

I believe we should be using core material that resonates well magnetically in the audible frequency range. That way, we can build a transformer in the shape of a tuning fork or something similar and match the magnetic resonance to the mechanical resonance! And by having the core actually move or bend, we can vary the inductance of the windings resonantly! I have a metal tuning fork somewhere in the house......I'm thinking like poles on the fork ends, so repulsion occurs, and you could have the mechanical resonance twice the frequency of the circuit's.

[Armagdn03]

I believe we should be using core material that resonates well magnetically in the audible frequency range. That way, we can build a transformer in the shape of a tuning fork or something similar and match the magnetic resonance to the mechanical resonance! And by having the core actually move or bend, we can vary the inductance of the windings resonantly! I have a metal tuning fork somewhere in the house......I'm thinking like poles on the fork ends, so repulsion occurs, and you could have the mechanical resonance twice the frequency of the circuit's.

this sounds like a very very dificult build, mainly because cores that are specially designed for mag amps (they definitly exist) are pressed cores. Its a good idea, and I really like the concept of using physical resonance, however in practice it would be very hard.

Here is a good challenge. Create a mag amp, to which you put DC in, and get an oscillating AC wave out. How can you create a mag amp which has input and output inductively free of each other, yet still creates an EMF in the output, and uses feedback to induce oscillation? Pull this off, and you are well on your way.

[Web000x]

If the internal winding (labled in figure as: #2) is wound in such a was as to have NO mutual inductance with the other winding (#4) then there will be no transformer action. This can be accomplished with two mirror windings. Winding 2 will create a flux circuit which circles the two smaller legs around which it is wound. The winding 4 will see flux paths in two opposite directions within its center, and so will have NO EMF induced.

Can you elaborate on the above statement?

Are you talking about using true mirror imaging or are you speaking of what the Alexanderson diagram that you posted shows?

I hope to get some more experimenting in soon using parameter changes. This traveling job kind of limits me..

Sorry for the confusion,

Dave

[Armagdn03]

Can you elaborate on the above statement?

Are you talking about using true mirror imaging or are you speaking of what the Alexanderson diagram that you posted shows?

I hope to get some more experimenting in soon using parameter changes. This traveling job kind of limits me..

Sorry for the confusion,

Dave

Sure I can elaborate!

The inner winding which goes up one leg, down the other should create a north and south at one end, and north and south at the other. Because each leg of that inner winding creates a magnetic field in two opposed directions, there is no way to induce an EMF in the second winding which covers the inner two. No reason to wonder however, just think the problem through, how do you wind a winding on those inner legs which cannot inductively couple to the outer legs? In fact, this has been done in another way with toroids by JLN labs.

Naudin S2Gen



Here he is CLEARLY biasing a toroid to be at the transition point of a core specifically designed for Mag amps.

The neodymium magnet is important. Without the magnet, the lamp does not light up, there is no current in the pickup coil because there is no EM coupling with the toroďdal coil

The output coil must be fully EM decoupled from the input coil (no mutual inductance), so this why the toroďdal coil is used as the input coil and a cylindrical or a flat coil set at 90° as the output coil.

He also stated this on the site, which is wrong in my opinion.

The magnet is used only to set the operating point in the MH curve of the toroďdal core. The magnet is not the source of the excess of energy. The ferromagnetic core is used on the highly non linear portion of the MH curve (where the core permeability drops quickly)

One can change inductance ONLY without charging the inductor with a coherent magnetic field, and there will be NO EMF generated in the pick up coil. That is why on post 89 of this thread I gave the following example (it is quite easy to see why the magnet DOES in fact play more of a role than simply setting the point on the B-H curve):

If we take a capacitor and separate its plates, we decrease its capacitance, but we DO NOT INCREASE its energy state.

If we take a CHARGED capacitor and separate its plates, we decrease its capacitance and INCREASE its energy state.

Similarly we can change the permeability of an inductors core, and we DO NOT change its energy state

But we can change the permeability of a CHARGED inductor core, and affect a change!

[Armagdn03]

Consequently I also posted an interesting way to accomplish this many moons ago.

The idea is to use two transmitters, eccentric to each other, to create a beat frequency between the two.



The beat frequency will be orders of frequency different from the transmitters, and so will be far outside their bandwidth. (Meaning that the BEMF of the pick up will not really be seen by the transmitters, thus not affecting their Q and resonant rise.)

The pick up should be tuned to the beat frequency.

When the pick up is also biased with a magnet to give it a coherent magnetic field AND set its bias point, the beat frequency will scatter the domains during one portion of the cycle, and let them re-cohere the next portion of the cycle, hence the term Degaussing Generator

Have fun!

[SilverToGold]

Going back to Capacitive Parametric Change, I built and tried out your florescent tube capacitor and got about the same variation. 4 pF OFF and 590 pF when ON.

The problem is that with this test setup is that it has a very poor ability to hold any type of high voltage due to leaks. Also, this setup has poor dielectric strength since only a thin layer of glass is your dielectric.

Another interesting observation, when the tube turns ON and the cap charges up, it will induce extra currents within the tube that light up the tube even more. When you turn off the tube, the charges will also move to neutralize and light up the tube once again. This means that the current will also run through your tube and this limits your power output since it would burn out the tube faster.

The positive news is that this device is a true variable capacitance device that will actually move charges like a mechanical variable cap.

Getting variable capacitance measurements is easy with lots of different solid state designs but they do not move/pump charges! This one does.

For example, take 4 plates and stack them between some dielectric, 2 outer plates are one cap, the 2 inner plates are another. You can very easily change the capacitance reading on one pair of caps by just touching the other caps leads together or using a switching transistor to connect them. It seems like a true variable cap BUT it does not pump charges.

If you do some calculations, you can derive how much possible energy this system could theoretically create and this is the basic equation you should get:

Power Out = f*(Co+Cmin)*Vi^2*( 1 - (Co+Cmin)/(Co+Cmax) )

This is for the same system in the first post where a variable cap is in parallel to a static cap connected through a load.

Co = static cap value
Cmin = minimum value of variable cap
Cmax = maximum value of variable cap
f = frequency of tube lighting
Vi = Initial voltage on caps

These tubes are generally operated above 20 kHz at around 10 Watts for the ones created in the video, above the hearing frequency to make them more quiet.

Looking at the equations, we can make these assumptions for design of this system:

1) Use the highest initial voltage possible. This is more important than using a large variable cap. Power out is the square of the voltage.

2) Using a small variable cap is better so you can operate it at a higher frequency.

3) The greater the difference between Cmin and Cmax is ideal.

4) Use as large a possible value for your static cap (Co) as reasonable.

5) Higher frequency equals higher power output but the input power is pretty much the same regardless.

You could in theory easily operate the system to give out kilowatts of power from one tube but remember that the power will also transfer through the tube so it would quickly burn out these tubes if not properly handled. This limits the synthesis of energy in this system more so than any other factor since the design is so simple.

One simple way to compensate for this is to use a larger Load Resistance. This will lower the current and transfer more power to your Load since in a closed current loop, percentage of power transfer is proportional to percentage of the resistance of that element in the current loop (think voltage divider). This is another reason for a smaller variable capacitance, it allows larger Load Resistance, which allows larger power transfer to the load and the ability to run the system at higher overall power levels (due to less stress on the tube from smaller currents per cycle).

From my calculations, Loads in the 100's of KILO-OHMS are possible when using pF variable capacitance at 30 kHz.

Just for fun, here are some calculations based on reasonable values.

A) f = 30 kHz, Co = 2 uF, Vi = 10 kV, Cmax = 44 pF, Cmin = 4 pF

Power Out = 120 Watts

B) f = 30 kHz, Co = 2 uF, Vi = 30 kV, Cmax = 44 pF, Cmin = 4 pF

Power Out = 1050 Watts

For both cases, the input power is the same. It takes the same power to light the tube in both cases, less than 20 watts.

Again, this is mostly theoretical but based on very reasonable assumptions.

[Armagdn03]

A) f = 30 kHz, Co = 2 uF, Vi = 10 kV, Cmax = 44 pF, Cmin = 4 pF

Power Out = 120 Watts

B) f = 30 kHz, Co = 2 uF, Vi = 30 kV, Cmax = 44 pF, Cmin = 4 pF

Power Out = 1050 Watts

For both cases, the input power is the same. It takes the same power to light the tube in both cases, less than 20 watts.

Again, this is mostly theoretical but based on very reasonable assumptions.

WOAH! same input!!! cool uh?

good rough calculations to show how the system acts, you definitely get it.

one more thing to consider, is the time constant created by the circuit, if your load does not let most of the charge through it in one period of the frequency, then you shuttle less coulombs per second, meaning less power output, so as I said before, the less impedance the load has, the more average power out to a point.

[SilverToGold]

Well, I haven't actually build any circuits but I do believe the power in is the same. The power through the variable cap circuit goes through the tube but should be somewhat separated from the tube's power supply (in theory).

There are several approaches to this system, I believe the low variable cap is the best method given the inherent small capacitance of the tube's setup. So we go for higher frequencies and higher voltages to make up for the small capacitance.

In this system with the small capacitances, I believe the larger resistance in the load is better. The RC time constant is so small with this system that it is dwarfed by the comparatively larger tube's 30 kHz. It's on the order of 7 significant figures smaller for the system described above for a 1 ohm load!

As I pointed out, the tube will run a closed loop current through the variable capacitance circuit as the tube charges up and discharges. This means it will have some resistance within the tube. That means the tube will eat up power and burn out. So you want a much larger resistance in the loop to absorb most of the power, that would be the load resistance. You want as little power in the tube and most of it in the Load, so this is the main reason I would go for a large load resistance in the order of kilo ohms probably.

In your initial design with the large caps, a small load resistance would be fine but that system obviously was theoretical and there was no plasma involved.

My thoughts on this system may be wrong also, I am confident it would work with low frequencies and low voltages but when you start working with high frequencies and high voltages, weird things start to happen. Things happen that don't otherwise occur with "normal" low voltage systems and plasmas actions are not linear or even common sense at times.

I also have not taken into account any resonance effects which may hamper or help this system.

I'm sure this system "synthesis" electrical energy. Have you gotten any verification that it does outside or Mr Dollard? It's so simple, hard to believe someone hasn't already done this in the private sector.

WOAH! same input!!! cool uh?

good rough calculations to show how the system acts, you definitely get it.

one more thing to consider, is the time constant created by the circuit, if your load does not let most of the charge through it in one period of the frequency, then you shuttle less coulombs per second, meaning less power output, so as I said before, the less impedance the load has, the more average power out to a point.

[SilverToGold]

One other point I would like to make.

With the small variable cap and large static cap, even though this system operates at high voltages...... during each cycle, the voltage change across the caps are less than 1 volt (0.20 V) for a 10 kV initial voltage! So technically, i would say it's a low voltage high frequency system I think this is a GREAT benefit to the system since that means reduced EM output and reduces the chance of interference with the system's operation and of close by equipment.

Most charges move off of the tiny variable cap but since the large cap is so large, the voltage does not change much. Also, since the large cap is so much larger than the smaller variable cap, the time constant's C value is roughly equal to the smaller cap's max value.

I think this is best because I want as little power going through the tube as possible per cycle with most of it going to the load.

One major problem with this system is voltage leakage from the high voltages. It is of utmost importance to create smooth surfaces and very good insulation on these devices or it will not work very long or very efficiently.

[Armagdn03]

One other point I would like to make.

With the small variable cap and large static cap, even though this system operates at high voltages...... during each cycle, the voltage change across the caps are less than 1 volt (0.20 V) for a 10 kV initial voltage! So technically, i would say it's a low voltage high frequency system I think this is a GREAT benefit to the system since that means reduced EM output and reduces the chance of interference with the system's operation and of close by equipment.

Most charges move off of the tiny variable cap but since the large cap is so large, the voltage does not change much. Also, since the large cap is so much larger than the smaller variable cap, the time constant's C value is roughly equal to the smaller cap's max value.

I think this is best because I want as little power going through the tube as possible per cycle with most of it going to the load.

One major problem with this system is voltage leakage from the high voltages. It is of utmost importance to create smooth surfaces and very good insulation on these devices or it will not work very long or very efficiently.

I assume the .2v change is across the caps in parallel once the system has re-equalized?

You must have a much larger second capacitor.

You can adjust the potential difference between the capacitors by changing the ratio of the capacitances between them. For example, if you make them equal capacitance, the voltage will asymptotically move towards 2x the original voltage, which would be 20kv in your case. The change in voltage (along with capacitance) will give you your change in coulombs and you have your time already, you can then figure out the work you have done. High voltage insulation is definitely a must as you said.

Something to think about, how can you make a "Flux Conserving Inductive Spring"? Might be worth considering due to the fact that a magnet can charge an inductor.

[SuperCaviTationIstic]

My first utube video
YouTube - ‪Gas discharge tube with relative capacitor pulsing‬‏

I did an experiment with a u.v. purifier bulb wrapped in foil and the flyback hv from a television.... The tube is connected to hv positive, and then through a spark gap back to the t.v. ground. The relative capacitor goes to a spark gap, which goes to a water pipe. Each time the closed circuit gap fires, the relative capacitor gap fires also.

I'm getting nice EMP type bursts I guess too, since my garage door opened twice during my experimenting.

Next thing to do is make some real chokes, and a capacitor so I'm not flashing the tv tube on and off every time the gap fire and so I don't blow the circuit board. Then I guess a load needs to go between that water pipe and the relative capacitor plate....

Any suggestions?

[SilverToGold]

Yes, the static cap is intentionally very large. If the equation is correct, then this is the ideal case. If I used equal capacitances as you suggest, the energy output goes down by half according to the equation.

The large cap can provide more energy than a smaller static cap. The reason is that the coulombs are moved out but at a basically constant high voltage. With a smaller static cap, the voltage drops significantly as the charges move out so less power is delivered.

The amount of charge is proportional to the change in the variable cap. So in my case going from 4pF to 44pF, the amount of Coulombs moved is (1-4pF/44pF), about 90%. So most of the charges are moved even though the voltage only changes by 0.2V.

To be honest, I really do NOT understand how all this works and I really don't think anyone here really knows either. If they do, then please explain it to me.

To get the equation, I simply calculated the energy in the capacitances before the tube was turned on. Then I calculated the energy after the tube is turned on. Then I ASSUME the lost energy is then consumed in the impedances of the circuit. Makes sense but may be incorrect.

BUT when the tube is turned OFF, the system springs back to it's initial state. WHERE DOES THE EXTRA ENERGY come from???? I turn off the tube (remove energy from the system) and the energy the system goes up?

I also question my original assumption of the energy going to the impedances. If energy can be created in one cycle seemingly out of nowhere, why can't it be destroyed in another?

All I am saying is that using traditional energy analysis with this system is not straightforward.

The inductive spring is interesting but I don't even fully understand the cap one enough to move on and the cap system is relatively "easy" to understand. This stuff does not make sense when you look at the numbers. So I must go back to the board till I can account for it all.

I assume the .2v change is across the caps in parallel once the system has re-equalized?

You must have a much larger second capacitor.

You can adjust the potential difference between the capacitors by changing the ratio of the capacitances between them. For example, if you make them equal capacitance, the voltage will asymptotically move towards 2x the original voltage, which would be 20kv in your case. The change in voltage (along with capacitance) will give you your change in coulombs and you have your time already, you can then figure out the work you have done. High voltage insulation is definitely a must as you said.

Something to think about, how can you make a "Flux Conserving Inductive Spring"? Might be worth considering due to the fact that a magnet can charge an inductor.

[Armagdn03]

the best possible connection to load is an impedance matched transformer to your load. This allows for an almost impedance free movement of coulombs through the load, which is optimal.

You are definitely correct that there is a trade off between cap size ratio, to voltage and coulombs moved. To get higher voltage differential, you sacrifice coulombs moved and vice versa.

There are lots of cool applications and im just exploring some of the ideas, helps to understand the situation from several angles.