by Fred B. Epps ( 04/27/97 ) - JLN Labs

In this analysis of how the Hendershot machine might have worked, I will make certain simplifying assumptions. First, I assume you are familiar with the circuit diagram and construction features of the Hendershot device. I also assume the device can be analyzed as two parts, the coil/capacitor assembly and the "clapper" resonant circuit. The coil/capacitor (CC) is my major focus here, but I will point out later how the resonant energy-storing principle is essential to the operation of the device.


In examining the CC assembly we find a very interesting component that is quite likely to create overunity performance if used properly.

(The basket-weave coil consists of braided windings over a modified electrolytic capacitor).

It is common sense that the capacitor and coil are intended to interact. The question is, what is the nature of that interaction?

Logically, there are three possibilities:

1) The coil affects the capacitor, but not the reverse.

2) The capacitor affects the coil, but not the reverse.

3) The coil affects the capacitor, and the reverse.

I reject option 2 because there is no element in the coil that can be affected by a changing electric field in the capacitor.

I reject option 3 because no true overunity device can be reciprocal-- it must be nonreciprocal. A reciprocal or 'two-way" device must always load the input and cause power loss at the input equivalent to the power gain. Let me explain what I mean by "reciprocal". Most systems that are encountered in everyday life and engineering practice are reciprocal in nature. What this means is that the energy relations are reversible. A good example is EM induction, where if the output of a motor becomes the input, the motor becomes a generator, and the energy relationship is reversed without being changed. This is an extension of Newton's law of action and reaction.

It has been proven (1) that certain systems are nonreciprocal, that is, the outputs cannot be made inputs. It has also been proven (2) that any nonreciprocal device with electrical inputs and outputs must contain a magnetic field. The magnetic field has the property of changing the direction of applied forces without doing work-- this is essential to these types of systems. In nonreciprocal systems the output does not load the input. Imperfect, lossy nonreciprocal devices can be constructed in many forms, among them gyroscopes, gyrators, ferromagnetic amplifiers, microwave phase shifters, and Hall effect devices.

Only the first option is capable of creating overunity performance: there must be a one-way interaction between the magnetic field of the coil and the capacitor. Otherwise the magnetic field must be loaded in some way, either inductively or parametrically, and the coil will lose energy. For there to be no losses in the core, it must 'see' the capacitor as basically an air core with an unvarying u of 1.


There are three ways to look at the braided windings of the coil:

1) They are inductive.

In this case the braiding is not intended to reduce flux and there is considerable flux inside the capacitor. The operating principle involves standard magnetic fields.

2) They are noninductive.

The flux is essentially prevented from entering the capacitor. The operation involves the magnetic vector potential.

3) They are both inductive and noninductive.

I am not equipped to analyze the noninductive (scalar) aspects of the coil (perhaps Bob Shannon would like to look at that). I don't believe that it is necessary to invoke scalar fields to explain its operation. Thus I make the further, possibly incorrect, assumption that the coil uses only standard magnetic fields.


How might the magnetic field of the coil interact with the capacitor? At the risk of boring "old hands" I think it necessary to explain the action of an electroytic capacitor to show how this might happen. An electrolytic capacitor has an extra layer of liquid or solid electrolyte between one of the plates and the dielectric. This layer has two main functions-- it provides better contact between the metal and the dielectric, and seals tiny holes that form in the dielectric by electrical action.

A liquid dielectric such as used in the Hendershot device has interesting electrical properties. It has a high dielectric coefficient k , as well as ionic conduction (3). Since obviously the ionic content has something to do with the value of k, and since moving ions are subject to Lorentz forces over their free path, it seems at least possible that a magnetic field through the coil would change the value of k and thus the C of the capacitor. I do not know enough chemistry to describe all the details but I have posted material on the magnetovoltiac effect and the experiments of Weiss who showed that magnets will affect the chemical and thus electrical conditions in wet cells. Hall effects exist in electrolytes (have to find the ref.)

The particular mechanism that is involved need to be determined by experiment. I think it will prove to be an ionic Hall effect, but it might be scalar or something else. I am not as focussed on the specific mechanism of interaction as that there IS a one-way interaction where L changes C without C changing L. This is a nonreciprocal system.

You may be familiar with the experiments that Jean-Louis Naudin and I are doing with varactors. These APPEAR nonreciprocal at the voltage

levels we are using because the output voltage is not high enough to appreciably affect the control voltage. At higher output voltages the device goes into nonlinear operation (frequency doubling mode). This mode does affect the input, sometimes cancelling the control voltage, soit is definitely not a nonreciprocal sytem. Also, keep in mind the existing proof that all electrical nonreciprocal systems must contain a magnetic field.

Hendershot's coil/capacitor appears to be a true nonreciprocal system. The theorems for nonreciprocality were proven for a low-frequency system consisting of an electromagnetic transducer mechanically coupled to an electrostatic or piezoelectric transducer.

The general form of these devices is "magnetic--mechanical or material coupling--electric". Hendershot's device fits this pattern well, being "magnetic induction--mechanical properties of ions--capacitance".


Nonreciprocal systems are usually lossy, although some microwave ferrite systems can operate with very low insertion loss. How does sucha system go overunity?

Through resonance. Remember that the nonreciprocal system by definition does not load the input. It cannot change the input in any way, so whatever state the input was in at the beginning of the process is the same state its in at the end of the process. If the input is a resonant circuit, only the electrical losses in the input need be considered because we can ignore the output by definition.

The fact that power is being transferred to a load through the nonreciprocal element cannot influence the operation of the circuit. At a certain Q in the input circuit and at a certain level of loss in the noreciprocal element, the circuit will go overunity. It should be pointed out that the output in Hendershot's device was due to variation of the capacitance in the CC assembly and thus drove the output circuits parametrically.

This is why his circuit required constant tuning. The Mathieu equations for parametric oscillations have many areas of instability for different ranges of frequency and power.

These considerations allow for the design of many types of overunity system in many media. Everything from purely mechanical devices like gyroscopes to solid-state ceramic resonators could be called into play. They also allow us to understand, predict, and duplicate the operation of devices like the Testatika and the Hendershot device. I would be happy to discuss building projects with anyone who is interested.

Fred Epps


1) "Violation Of The Reciprocity Theorem In Linear Passive Electromechanical Systems" by Edwin McMillan, J. Acous. Soc. Am. (18), 344 (1946)

"Coordinates And The Reciprocity Theorem In Electromechanical Systems" by John W. Miles, J. Acous. Soc. Am. (19), 910 (1947)

2) "Reciprocal Relations In Irreversible Processes I, II" by Lars Onsager, Phys. Rev. (37) , pp. 405-426 (1931)

Some Aspects Of Onsager's Theory Of Reciprocal Relations In Irreversible Process" by H.B.G. Casimir, Nuovo Cimento Suppl. (6), pp. 227-231 (1949)

3) Electrolytic Condensers, by Philip Coursey, Chapman and Hall, 1937

Return to Parametric Home page or Return to Home page