OverUnity Research

"Nickolay E. Zaev worked on creation of the prototypes of converter energy, which do not require
any fuel. The direct conversion of environmental heat to electric power is possible in the processes
of charge-discharge in non-linear capacitor or by means of magnetization-demagnetization of
ferrite. Such converters of energy create cold and electric power without any fuel."

Some efforts were made to reproduce device based on non-linear inductance. It seems to be very difficult
obtain any significant (measurable) amount of energy using standard and widely available core materials.

It was not possible to experiment with devices based on non-linear capacitor because there is no such capacitors.
However, recently I became aware of chip (SMD) capacitors with X5R and X7R dielectrics. These dielectrics contains BaTiO and have interesting properties I would like to demonstrate.

But for beginning - N.E. Zaev "Research on the Capacitance Converter of Environmental Heat to Electric Power"

http://jnaudin.free.fr/doc/Capacitance%20converter.pdf

More Zaev's papers

Capacitors:
https://www.ebay.com/itm/291870890748
https://www.ebay.com/itm/291870890748

picture from datasheet (attached)

In first experiment I charge capacitors from constant current source and discharge through small resistor.
On pic. 1 (attached) shown capacitors I tested
and on pic. 2 scope traces (yellow - control signal, blue - voltage on capacitor)
There are only 3 set of traces, bank of non-electrolytic capacitor behave same way as electrolytic i.e. is is linear

In second experiment I replaced discharge resistor with constant current sink.

And last experiment I tried discharging capacitor with very short pulses (5us) through small resistor (4.7Ohm)

Attaching traces for linear and nonlinear capacitor and graph how capacitance (C) depends from voltage (U) cal from this traces.

One piece of lost knowledge from 50s, time of electronic tubes.

In tube amplifiers, capacitors between amplifier stages were often under significant bias
voltage (because one side connected to anode and other side to control grid)

Bad capacitors often caused distortion and even oscillations.

Possible practical application of non-linear capacitors - Weber circuit, supposingly self running.

Here is the page in German with one of Webers circuits:
https://www.psiram.com/de/index.php/Gustav_Hans_Weber

If anyone has more information about this, please share.

Dear Vlasik

Thank you for all the info generously offered on non-linear capacitors as a possible source of excess energy.

I recall the work of Zaev first seen on JLN site many years ago, but I never pursued work in this area.

Have you considered double layer caps or supercaps for these experiments as they have extreme non-linear V vs Cap curves, but the slope may be in the wrong direction.

Regards

Quote from: ion on 2018.12.22, 17:47:21
Have you considered double layer caps or supercaps for these experiments as they have extreme non-linear V vs Cap curves, but the slope may be in the wrong direction.

Thank you ION,
I tried using super capacitors for energy storage, e.g. run joule theft, bit not for these experiments.
Supercaps have very big capacity and charge/discharge cycle will be long or very high currents needed.

Regards
-V.

Quote from: Vasik041 on 2018.12.22, 18:36:07
Thank you ION,
I tried using super capacitors for energy storage, e.g. run joule theft, bit not for these experiments.
Supercaps have very big capacity and charge/discharge cycle will be long or very high currents needed.

Regards
-V.

Yes large capacity, lower switching frequency, high currents involved, however, more energy delivered per cycle.
Also note that the currents used in the charge / discharge cycles need not be lost energy, as these heating effects are still recoverable as part of overall input energy.

e.g if the entire apparatus were run in a styrofoam box and temperature rise noted, we could compare against a equivalent resistor fed from a power source to determine comparative heating / cooling effects over a long period of time.

If there are exothermic or endothermic factors, these can be separated out with a careful test setup.

I have noted extreme non-linear behavior of double layer caps, but agree the time constant is long. I did charge / discharge using a current source and paper strip chart recorder, then held a ruler against the produced graph. Of course this could also be done with a data acquisition device and PC.

I need to refresh on Zaev's work.
Regards

Quote from: ion on 2018.12.22, 19:17:59
Question: In a non-linear capacitor based energy gain system, do the opposing plates of a capacitor demonstrate a temperature difference? Been a long time since I read Zaev, but seem to recall this being mentioned.

I can't remember such detail. As far as I understand energy comes from dielectric's heat and I would guess that both plates cool simultaneously. This is similar to magneto caloric effect (https://en.wikipedia.org/wiki/Magnetic_refrigeration).

from Zaev's paper:
"Energy is taken from free (heat) energy of ferroelectric. B.B. Golizin (https://en.wikipedia.org/wiki/Boris_Borisovich_Golitsyn)
showed the possibility of such mutual conversion in dielectrics in 1893 [5]. It is a pity, that there are no mentioning of this basic
article by B.B. Golizin in any works on thermodynamics of dielectrics."

Regards,
-V.

Got to thinking about Hans Weber's self oscillating circuit and thought I would play with it on paper before going to the bench. In the process I think I have boiled down a few simplified versions that may eliminate the need for a transformer, and in the end can use an inductor (or tapped inductor if the original FW rectified version is preferred).

In one version, I have preferred to make three items tunable such that a sweet spot can be found.

Note that quite possibly a lower value of storage capacitor and a half wave rectifier may provide more "pumping action", but this is TBD.

Also a non-linear load may facilitate operation.

The initial startup can include a signal generator and loose coupler to induce operation while tuning for the sweet spot.

Unless the original circuit relies upon leakage inductance in the 1:1 transformer, I see no reason why the simplified versions using an inductor would not be worthy as primary to secondary isolation is not required.

Regards

Quote from: ion on 2018.12.24, 17:32:37
Got to thinking about Hans Weber's self oscillating circuit and thought I would play with it on paper before going to the bench.

As far as I understand simplest setup is just L, non-linear C and voltage source for offset.
I am trying figure out what is condition for non-damped osculations, but so far did not get it.

Regards
-V.

Quote from: Vasik041 on 2018.12.25, 20:23:34
As far as I understand simplest setup is just L, non-linear C and voltage source for offset.
I am trying figure out what is condition for non-damped osculations, but so far did not get it.

Regards
-V.

I think the bias voltage on the capacitor is a requisite for oscillation, which an LC alone does not provide. The proper point on the capacitor's non-linear curve must be found, I presume.

Reversing charge on the capacitor probably doesn't get you to the sweet spot on the non-linear curve.

As someone said: "things must be made as simple as possible, but no simpler". (don't remember the exact quote by Einstein?)

Regards

edit: Attached is a proposed test setup using a signal generator with LF offset output to induce a variable bias on C2.
       Look for small HF burst as variable V traverses sweet spot.

Hi all,

I was "Exnihiloest" in the past and I am now back with my ham radio callsign. I left for reasons that are no longer valid today, so I am very happy to be back among you, especially since some of you asked me to on ou.com, and I greet the former contributors I have known here, as well as the new ones.

This thread caught my attention because I'm experimenting on dielectrics. For me parametric systems like non-linear capacitors cannot a priori give overunity because it takes energy to change a parameter, like reducing a capacitance to increase its energy by virtue of W=Q/C. The energy required to reduce C is exactly the electrical energy that will be recovered in W, this is easily verified for a parallel plate capacitor.

But what if the capacitor is "open" and the energy for parametric change can be drawn from the environment? Then non-linear capacitors could be the gateway to the hidden source, which could be ambient heat (Maxwell's demon), LENR, ZPE or other unknown sources...
I am now investigating the first case, according to Zaev's idea, and so I am looking for a dielectric whose permittivity decreases when the temperature increases. Did you find one? Because in general, what we find is the opposite. If anyone has an idea, I'm a taker.

On my side I started to test ferrites as dielectrics. It is known that they have a high permeability but less known that they also have a high permittivity, depending on frequency (and perhaps temperature?). NiZn ferrites are the first ones I test. There are other types such as MnZn with higher permittivity but with too many losses because they are not insulating enough.
I think this is a little explored field, including interactions when the two magnetic and electric fields superpose in ferrites and could have cross parametric effects. Just an idea.

François

Francois

A large welcome back, you were deeply missed when you left as I (and others, I'm sure!) greatly valued your input and always learned something. Looks like you have a good grasp of what is needed for parametric operation of capacitors.

Hopefully we can try a few ideas.

Kind Regards

Hi François,

Welcome back!
These capacitors I've been testing decrease capacitance with temperature.
I found this video https://www.youtube.com/watch?v=j0MLo_BYLHQ but haven't tried it myself.
Capacitance decrease with temperature rise and this cause voltage rise also.
My goal build device which does not require heat source (except ambient environment heat).
It would be nice to learn more about your experiments.

Regards,
-V.

PS there is related thread about ferrites  here http://www.overunityresearch.com/index.php?topic=3453.0
Perhaps you find it interesting.

Quote from: F6FLT on 2019.01.02, 15:45:19
Hi all,

[snip]

On my side I started to test ferrites as dielectrics. It is known that they have a high permeability but less known that they also have a high permittivity, depending on frequency (and perhaps temperature?). NiZn ferrites are the first ones I test. There are other types such as MnZn with higher permittivity but with too many losses because they are not insulating enough.
I think this is a little explored field, including interactions when the two magnetic and electric fields superpose in ferrites and could have cross parametric effects. Just an idea.

François

Hi Francois,

Welcome.  Perhaps you may find the attached papers interesting regarding an anomaly I found in a few MnZn ferrite cores some time back.

Regards,
Pm

Thank you everyone for your welcome

@Vasik
Do you know what this "nanocrystal" is, which is referred to at https://www.youtube.com/watch?v=j0MLo_BYLHQ_BYLHQ that you have given?
That is what we need. Someone says in the comments that it would be a tantalum capacitor.

It would also be interesting to know if the effect is reciprocal, i.e. if reducing the voltage reduces the temperature (if there is an effect, it must be small, and I am not equipped for this kind of measurement).

@Partzman
We saw on ou.com how difficult it is to measure high frequencies. I just screwed up right now while playing with my NiZn toroids  :(!
When the COP is large, for example >3, it should be possible to replace the FG with a low-consumption 1 transistor oscillator, which could be powered by the rectified output of the setup. It's easy to build. I think that's the only way to be sure of the COP.

Quote from: F6FLT on 2019.01.03, 17:35:59
Do you know what this "nanocrystal" is, which is referred to at https://www.youtube.com/watch?v=j0MLo_BYLHQ_BYLHQ that you have given?
That is what we need. Someone says in the comments that it would be a tantalum capacitor.

I guess it is SMD capacitor. I tested capacitors with X5R and X7R dielectrics. They are really non-linear (see beginning of this thread).
As far as I understand, we can have OU heater or cooler, similar to magnetic refrigerators.
With regular ferrites temperature difference is very small. I haven't tried yet with capacitors.

If you charge(discharge) capacitor fast enough, you create rapid entropy change in dielectric and it will cause interaction with ambient environment (temperature difference).

Regards,
-V.

I tested several capacitors and found only one whose capacity decreases with temperature, it is a component from the 60s, see the scan (it's the first time I scan a component instead of making a photo, it works well  :) ).
The effect is only a few % on the resonance frequency from 20 to 70°. We would need much more non-linear components. I think that temperature stability is one of the criteria for choosing dielectrics in industry, so I don't think we can find very effective ones, i.e. with a large and tight variation as a function of T. We'll have to find the right dielectric.

Quote from: partzman on 2019.01.02, 17:23:59
Hi Francois,

Welcome.  Perhaps you may find the attached papers interesting regarding an anomaly I found in a few MnZn ferrite cores some time back.

Regards,
Pm

Hi partzman ,

Very interesting anomaly that I confirm.

I tried with 3 toroidal coils from the electronics industry (I didn't wind them myself). The ferrite is likely NiZn for all but not sure.

One has its two windings wound like a bifilar coil. The other two have two diametrically opposed windings.
At the middle point, when the connections are made in such a way that the windings oppose their magnetic fields, I have less than a tenth of the input voltage, whereas we expect half. Same result for all three.

In fact as both windings are opposing, with ideal coils 100% coupled we would have a short cut and the FG coudn't generate the signal.
So we have to deal with non ideal coils. I made a LTspice simulation. While the coefficient of mutual induction is less than 0.9 (which is a good coupling), a generator with a 50 ohm impedance provides enough current to maintain a voltage near to that one in open circuit. Nevertheless the simulation shows half the voltage at the coils mid point, regardless of the coupling coefficient and of the loss in the internal FG resistance.

Very strange, I don't have an explanation yet.

Quote from: F6FLT on 2019.01.04, 15:57:56
Hi partzman ,

Very interesting anomaly that I confirm.

I tried with 3 toroidal coils from the electronics industry (I didn't wind them myself). The ferrite is likely NiZn for all but not sure.

One has its two windings wound like a bifilar coil. The other two have two diametrically opposed windings.
At the middle point, when the connections are made in such a way that the windings oppose their magnetic fields, I have less than a tenth of the input voltage, whereas we expect half. Same result for all three.

In fact as both windings are opposing, with ideal coils 100% coupled we would have a short cut and the FG coudn't generate the signal.
So we have to deal with non ideal coils. I made a LTspice simulation. While the coefficient of mutual induction is less than 0.9 (which is a good coupling), a generator with a 50 ohm impedance provides enough current to maintain a voltage near to that one in open circuit. Nevertheless the simulation shows half the voltage at the coils mid point, regardless of the coupling coefficient and of the loss in the internal FG resistance.

Very strange, I don't have an explanation yet.

Hi F6FLT,

Yes it is puzzling!  I had purchased a number of these cores on Ebay and I used this particular one on a number of bench tests over time so I may have "conditioned" it somehow along the way.  I initially suspected magnetostriction so I placed felt on the inner and outer edges of a virgin core to float the windings off the core and it did exhibit some voltage increase at the tap.  I had planed to 3d print special covers that would allow as little support as possible for the core and windings but haven't done this to date.

A similar condition can be forced with a capacitor connected to the tap but it is not asymmetrical like the anomalous core.  This led me to believe that perhaps it might be related to the permittivity of the material.  I've attached a paper relative to the dielectric measurement of ferrite materials that I had used for study.

Regards,
Pm     

@partzman

I just tried with a very large 9x14 cm rectangular ferrite core with a 3.3x3.3 cm square section, which was used in a KW power supply.
I have no anomalies.
The difference compared to my other tests is that here the coils are strongly coupled, distributed on both long sides of the rectangar core, and wide to the point of touching in the middle, so that it seems impossible to me that the opposite flux loop between the two coils.

I also tried with another toroidal ferrite, smaller than the other three already tested. I have the anomaly, with always Vout about equal to Vin/10.

Could it be that the anomaly only occurs when the 2 flux can loop back between the two coils?
Have you some ferrite toroidal core with no anomaly, and if so of which geometry?

Quote from: F6FLT on 2019.01.04, 18:50:23
@partzman

I just tried with a very large 9x14 cm rectangular ferrite core with a 3.3x3.3 cm square section, which was used in a KW power supply.
I have no anomalies.
The difference compared to my other tests is that here the coils are strongly coupled, distributed on both long sides of the rectangar core, and wide to the point of touching in the middle, so that it seems impossible to me that the opposite flux loop between the two coils.

I also tried with another toroidal ferrite, smaller than the other three already tested. I have the anomaly, with always Vout about equal to Vin/10.

What is your winding geometry on this core with the anomaly?

Quote
Could it be that the anomaly only occurs when the 2 flux can loop back between the two coils?

Actually, this is all I've tested that is, two identical counter-wound coils operating in a bucking mode.  My original intention was to experiment with the H field in the center of the toroid that is the result of the opposite poles at the start and finish gaps between the windings.

Quote
Have you some ferrite toroidal core with no anomaly, and if so of which geometry?

Out of twelve identical (supposedly) ferrite toroid cores as described in the paper, only one exhibits this anomalous behavior!  That's why I suspect I somehow changed the molecular structure in some previous tests.  Otherwise, I have not seen this type of anomaly in any other cores I've tested or used in the past.

Pm   

A magnetic simulation such as FEMM can show that flux will escape through the full length coils when in bucking mode.  It is interesting that a transformer where the two coils are wound on opposite sides of a ring core will exhibit significant leakage flux similar to that of the coils in bucking mode.  This is of course because the secondary mmf from the load current bucks the primary load mmf from the input current (they are in antiphase and don't contribute to the magnetization, only the small 90 degrees magnetizing current does that).  It is easy to see that leakage flux across the centre of the ring core using a sampling coil there, and it is proportional to load current.

I like to think of the two bucking coils as two bent solenoids, C shaped electromagnets that have been brought together.  Bringing each pole close to the opposite one of the same polarity just  makes the ring core into an electromagnet with N and S poles diametrically opposite to each other.
Smudge   

Hi Smudge,

I agree with you on the physical effect. The question now remains as to why we do not measure half of the voltage at the mid-point.
I performed realistic simulations with LTspice, including coil capacitance, coil resistance, capacitance in between, oscilloscope probe capacitance. Anyway, LTspice shows that we must have half the voltage at the mid-point, and we don't have it.
I made new measurements this morning with an HP3468A instead of the scope, and same observation in agreement with the scope.

Quote from: partzman on 2019.01.04, 19:16:01
...
Out of twelve identical (supposedly) ferrite toroid cores as described in the paper, only one exhibits this anomalous behavior!
...

How is it possible?
On my side, all the four toroids that I tried show the anomaly. Only the big one doesn't (see photo).

Note that the anomaly does not occur at frequencies too low. The impedance must be significantly higher than the internal resistance of 50 ohms of the FG, i.e. the working frequency must be high enough that the voltage drop due to the impedance of the coils does not exceed 10 to 20% of the no-load voltage of the generator (with the cores I have used, frequency always higher than 800 KHz)

Quote from: F6FLT on 2019.01.05, 13:21:36
How is it possible?
On my side, all the four toroids that I tried show the anomaly. Only the big one doesn't (see photo).

Note that the anomaly does not occur at frequencies too low. The impedance must be significantly higher than the internal resistance of 50 ohms of the FG, i.e. the working frequency must be high enough that the voltage drop due to the impedance of the coils does not exceed 10 to 20% of the no-load voltage of the generator (with the cores I have used, frequency always higher than 800 KHz)

I think at 800KHz or higher you just have some resonance effect with parasitic capacitance.
Have you tried attaching some load resistors (to Vout on your schematic) ?

Quote from: F6FLT on 2019.01.05, 13:21:36
How is it possible?
On my side, all the four toroids that I tried show the anomaly. Only the big one doesn't (see photo).

Note that the anomaly does not occur at frequencies too low. The impedance must be significantly higher than the internal resistance of 50 ohms of the FG, i.e. the working frequency must be high enough that the voltage drop due to the impedance of the coils does not exceed 10 to 20% of the no-load voltage of the generator (with the cores I have used, frequency always higher than 800 KHz)

The major difference I see is in the operating frequency and generating source.  In my paper, I use a solid state power amp with a very high damping factor thus a very low source impedance and my test frequencies are in the range of 6-7kHz.

With these test conditions, none of the other identical toroids produced the anomaly.

However, this same anomalous toroid driven with a 50 ohm SG still exhibits a tap to input gain of 1:07 in one direction at 1kHz!  Reverse the input and output winding connections and the gain is 2.01.  The inductance of the primary with the secondary shorted is typically 760uH.  The individual winding inductance is 34mH.

Edit:  I might add that the measured inductance seen by the generator at the input of the series connected bucking coils is 648uH.

Regards,
Pm

Quote from: Vasik041 on 2019.01.05, 15:16:22
I think at 800KHz or higher you just have some resonance effect with parasitic capacitance.
Have you tried attaching some load resistors (to Vout on your schematic) ?

No resonance. Same level from 800 KHz up to 3 MHz or 4 Mhz (resonances are above).

With a load resistance of 1 Kohm, Vout is divided by about three.

I have just improved the configuration by putting an isolation transformer (which is also a toroid with 2 coils) between the FG and the toroid to be tested. Then I can connect the oscilloscope ground on either side of the bi-coils.
Same measure! I have about Vin/10, which I measure from one side or the other. The configuration being now symmetrical, this means that it is the choice of the ground point that determines the coil on which only Vin/10 will be measured.
We face a measurement problem.

Quote from: partzman on 2019.01.05, 15:27:19
The major difference I see is in the operating frequency and generating source.  In my paper, I use a solid state power amp with a very high damping factor thus a very low source impedance and my test frequencies are in the range of 6-7kHz.

With these test conditions, none of the other identical toroids produced the anomaly.

However, this same anomalous toroid driven with a 50 ohm SG still exhibits a tap to input gain of 1:07 in one direction at 1kHz!  Reverse the input and output winding connections and the gain is 2.01.  The inductance of the primary with the secondary shorted is typically 760uH.  The individual winding inductance is 34mH.

Edit:  I might add that the measured inductance seen by the generator at the input of the series connected bucking coils is 648uH.

Regards,
Pm

For my configuration, I guess that it is a question of measurement method, probably that the impedance seen from the midpoint is high, and therefore the capacity of the scope probe collapses the voltage (nevertheless rather strange because not depending on the frequency). Not yet verified this idea.

On your side, I don't know if it's the same anomaly or another. What do you mean by "input gain"? Is it the voltage gain? How to connect the FG by inverting the input and output, do you keep the same ground point and connect the FG to the midpoint of the coils? (because if you double the voltage, everything happens as if your coils were in series but not in opposition).

This thread being dedicated to non-linear capacitors, I propose to open a new thread on your anomaly. Do you agree?

Quote from: ion on 2018.12.26, 18:03:14
I think the bias voltage on the capacitor is a requisite for oscillation, which an LC alone does not provide. The proper point on the capacitor's non-linear curve must be found, I presume.
Reversing charge on the capacitor probably doesn't get you to the sweet spot on the non-linear curve.
As someone said: "things must be made as simple as possible, but no simpler". (don't remember the exact quote by Einstein?)
Regards
edit: Attached is a proposed test setup using a signal generator with LF offset output to induce a variable bias on C2.
       Look for small HF burst as variable V traverses sweet spot.

Hi ION,

Today I tried this test setup (see attached picture).
I use signal generator (with isolation transformer) and power supply with variable output voltage to create sine with DC offset.
I think it is close to setup you suggested.
With load resistors 1k and 100 ohm I don't see any non-linearity, but with 10 ohm I see sawtooth like distorted signal.
It is also interesting that amplitude depends on offset voltage, the higher offset the lower amplitude gets. I still not sure about FE, but at least it should be possible build a volume regulator :)

I tried simulate this setup and simulation looks close to reality. However I am observing some extra small peaks on top of sine (see attached photos).

This could be just isolation trafo non-linearity, or it also could be something...

Regards,
-V.

Quote from: F6FLT on 2019.01.05, 16:14:12

[snip]

This thread being dedicated to non-linear capacitors, I propose to open a new thread on your anomaly. Do you agree?

Yes, I agree.

Edit:  Actually IIRC, there was a thread where this was discussed so let me take a look to see if it would work.

Edit2:  Here it is, see if you can access the thread-  http://www.overunityresearch.com/index.php?topic=3278.msg55390#msg55390

Pm

I can't access. Message "An Error Has Occurred! The topic or board you are looking for appears to be either missing or off limits to you."?

Quote from: F6FLT on 2019.01.05, 17:35:24
I can't access. Message "An Error Has Occurred! The topic or board you are looking for appears to be either missing or off limits to you."?

Try the link again.

Pm

Quote from: partzman on 2019.01.05, 18:46:19
Try the link again.

Pm

It works! I will continue there.

I'm testing unknown recovery components of ma collection.
I get a huge effect with one of them: 1.9nF at room temperature down to less than 1.3nF when above the soldering iron. The increase of capacity is faster than the decrease.

Does anyone know what it is?

Quote from: F6FLT on 2019.01.06, 11:16:44
I'm testing unknown recovery components of ma collection.
I get a huge effect with one of them: 1.9nF at room temperature down to less than 1.3nF when above the soldering iron. The increase of capacity is faster than the decrease.

Does anyone know what it is?

May be this ? https://www.allaboutcircuits.com/technical-articles/safety-capacitor-class-x-and-class-y-capacitors/

Edit: see dielectric properties here https://www.vishay.com/docs/28536/geninfo-cd-rfi-safcaps.pdf

Thanks for the information, Vasik.

I hadn't classified it with the capacitors because I couldn't read a capacity value. Could it be the mention "2.22" (nF), and that the component has drifted (I measure 1.9nF)?

Quote from: F6FLT on 2019.01.06, 12:06:02
Thanks for the information, Vasik.
I hadn't classified it with the capacitors because I couldn't read a capacity value. Could it be the mention "2.22" (nF), and that the component has drifted (I measure 1.9nF)?

Yes, I think it is capacitor marked with 222 i.e. 2.2nF

Edit: it 10% or even 20% tolerance, so 1.9nF quite ok