Extra energy in a capacitor

[F6FLT]

In this paper: https://vixra.org/abs/2207.0084 Eue Jin Jeong explains that when a capacitor is charged, the charges of same sign on each plate repel each other, and that consequently a potential energy of repulsion is stored in the capacitor in each plate, in addition to the usual attractive energy between the plates.

The author calculates this energy in the case of the spherical capacitor, and shows that it is much higher than the classical energy, according to the ratio of the sum to the difference of the radii of the 2 spheres of the spherical capacitor, multiplied by the relative permittivity of the dielectric.
For example with a dielectric of permittivity 1000, the stored repulsive electrostatic potential energy would be 100,000 times the classical attractive energy 1/2*C*V².

From a theoretical point of view, even if I have doubts about equation 4 and the intensity of the repulsive energy, it remains that there is one, and I don't see how to deal with it since it is not included in the calculation of the work of the capacitor charge. Any ideas?

From a practical point of view of FE assuming the theory is right, how would it be possible to use the repulsion energy without reducing the attraction energy by the same value?

[Hakasays]

Very interesting concept F6, not least of which because it makes intuitive sense as well as mathematical.    Both electrostatic attraction and repulsion are already well-known but they are always treated as separate entities, 'never the two shall meet'.

I've been simmering this idea wondering how it might be related to both transients as well as parametrics.

For example, what happens to a charged capacitor when the capacitance is decreased?  Does the repulsive field take over the attractive one, or does some asymmetry arise?

Same with transients, which might explain why high voltage lightning bolts can be both straight as well as branching/diverging.

[Smudge]

Big mistake in that paper!!  His equation (4) is wrong.  He admits that the repulsive energy formula for each spherical shell is that used classically for determining the electron radius which assumes the electron to be a spherical surface carrying its charge.  That is the electron as an isolated sphere.  The self capacitance of an isolated sphere of radius a is given by C = 4*pi*epsilon₀*a.  Thus its electrical energy by his equation (1) is given by either of the two terms in his (4).  That is also by his admission the attractive potential energy and (since he has used it as such) the repulsive potential energy for such a single sphere.  I.e for an isolated sphere however you describe the energy you get the same result.  His mistake is in using an isolated sphere formula for each of his two spheres in his (4) and that throws the whole thing out of the window.  It is quite clear to me that the repulsive surface tension in a spherical shell of charge will be severely moderated by having another spherical shell of opposite charge close to that surface.  He has not taken that into account.
Smudge   

[NickZ]

    Guys:
    I found this video on YT, but don't know what to make of it. Looks interesting, using the big microwave transformer and capacitors to show a self running device. Let me know what you think.
    https://youtu.be/PxzKw2j4m_Y

   NickZ

[broli]

Big mistake in that paper!!  His equation (4) is wrong.  He admits that the repulsive energy formula for each spherical shell is that used classically for determining the electron radius which assumes the electron to be a spherical surface carrying its charge.  That is the electron as an isolated sphere.  The self capacitance of an isolated sphere of radius a is given by C = 4*pi*epsilon₀*a.  Thus its electrical energy by his equation (1) is given by either of the two terms in his (4).  That is also by his admission the attractive potential energy and (since he has used it as such) the repulsive potential energy for such a single sphere.  I.e for an isolated sphere however you describe the energy you get the same result.  His mistake is in using an isolated sphere formula for each of his two spheres in his (4) and that throws the whole thing out of the window.  It is quite clear to me that the repulsive surface tension in a spherical shell of charge will be severely moderated by having another spherical shell of opposite charge close to that surface.  He has not taken that into account.
Smudge   

I think his argument is that both of these potentials are perpendicular to each other and thus can be decoupled from each other.

This also reminds me of another paper (attached) where a generator idea is proposed by mechanically removing the dielectric of a capacitor. They have not build an experimental setup of this to validate the theory. But if the paper in this thread is right it would mean that you get more energy out of the capacitor than the mechanical energy needed to remove the dielectric.

I know there has been numerous discussions about this sort of topic but I'm not aware of any experiment that measures the work done to remove a dielectric compared to the energy of the capacitor after its removal.

[F6FLT]

I think his argument is that both of these potentials are perpendicular to each other and thus can be decoupled from each other.
...

Good point! I had prepared my answer to Smudge and Hakasays similarly, but you were the fastest. 


I agree with Smudge about the mistake of equation 4, but we shouldn't throw the baby out with the bathwater.

I don't see the author's text as telling us that charges of the same sign in the outer shell would tend to oppose the attractive force, but that there is stress between them along the surface, which represents energy.
In other words, the repulsive surface tension in a spherical shell of charge cannot be moderated by the presence of another spherical shell of opposite charge near this surface, because the repulsive forces are oriented along the surface, between charges of the same sign close to each other, while the attractive forces by the opposite charges of the inner shell, are oriented perpendicularly to the surface, towards the center.
The attractive and repulsive forces are orthogonal. So to answer Hakasays, I don't think the two forces are competing.


Then the author chose the spherical capacitor because the calculations are simple (even if he was wrong), but his idea is equally valid with a planar capacitor: the charges of same sign would repel each other along the surface, tending to concentrate at the periphery where perhaps they could be recovered.

I will try to clarify the idea with a capacitor that we will charge with only 2 charges on each electrode, see brief diagram attached.
At the beginning these charges represent two dipoles of positive and negative charges.
If they are far apart, each dipole sees the other as two charges reduced to a point, thus as a zero charge, the forces are almost zero.

Charging the capacitor is equivalent to spacing the positive charge from the negative charge of each dipole, i.e. increasing the length of the dipole. When the distance between charges of the same sign starts to be small compared to the length of the dipole, each charge of one sign will "see" the other charge of the same sign and feel the repulsive force since it is no longer neutralized by the charge of the opposite sign.
This repulsive force does not work. Only the energy against the attractive force is to be taken into account for the charge of the capacitor, and yet the repulsive force seems to be able to do some work now, hence the idea of extra-energy.

[Smudge]

This also reminds me of another paper (attached) where a generator idea is proposed by mechanically removing the dielectric of a capacitor. They have not build an experimental setup of this to validate the theory. But if the paper in this thread is right it would mean that you get more energy out of the capacitor than the mechanical energy needed to remove the dielectric.

No, the author admits that the mechanical energy is equivalent to the output energy.
Smudge

[Smudge]

@F6,
If the connection to a parallel plate electrode is an extension to the parallel plate what force drives the electrons along that extension?  I has to be the repulsive force in the plane of the electrode, so that force does do work.  In the spherical capacitor the two forces are normal to each other but I stand by my view that the other plate of opposite charge influences the magnitude of the repulsive tension.
Smudge

[broli]

@F6,
If the connection to a parallel plate electrode is an extension to the parallel plate what force drives the electrons along that extension?  I has to be the repulsive force in the plane of the electrode, so that force does do work.  In the spherical capacitor the two forces are normal to each other but I stand by my view that the other plate of opposite charge influences the magnitude of the repulsive tension.
Smudge

I also agree with this. It's sort of like blowing a balloon, the tension in the rubber causes the pressure to build up even though the pressure acts normal and the tension/stress tangential.

[Hakasays]

I also agree with this. It's sort of like blowing a balloon, the tension in the rubber causes the pressure to build up even though the pressure acts normal and the tension/stress tangential.

A balloon would still be a really interesting analogy because it would mean that energy is being stored in two different domains (air elasticity and rubber elasticity.)

Do we know if electrostatic attraction and repulsion both propagate in the same manner?  If they are different mechanisms, they may have different properties.

The last research I've read put Coulomb force propagation velocity as 'probably extraluminal' (ie: no propagation velocity)  https://link.springer.com/article/10.1140/epjc/s10052-015-3355-3
But if the repulsion force is kinematically derived, then it's conceivable that it propagates at or closer than C.
That would open up a lot of directly engineerable experiments.

[partzman]

With all the theory put forth on this thread, I'm hoping someone can explain the following simple parallel resonant circuit C1 displacement current to me.  This is a simulation but a bench circuit with equivalent values performs exactly the same.

The driving source is a low impedance sine voltage generator V1 running for 1 cycle of the resonance frequency at 22.508kHz.  Several things to note IMO.  First, the rise of IC1 at the very start of the cycle to a level of ~28ma.  I have added some esr to C1 which produces the rise time shown which is very close to bench circuitry.  Second, the peak current ratio of IL1 to IC1 is ~2 and the apparent load to V1 is a constant current of ~28ma.

Next is the same circuit but V1 is turned off at the 22.21us point in time thus clamping VL1 at ground level.  The current IL1 is clamped but note what happens to IC1.

Last is the same circuit driven by a constant current sine generator I1 that has a peak current of 100ma but is shut off at .3 of the cycle or ~13.3us.  At this point in time, the current form I1 is removed from node VL1 and L1 and C1 are still connected.  Again note what happens with IC1.  BTW, the current directions arrows for L1 and C1 are towards ground respectively.

Perhaps this is well known to you all but to me this is both interesting and confusing.  I have tried to tap this so called displacement current but to no avail.

Please delete if desired!

Regards,
Pm

[F6FLT]

@F6,
If the connection to a parallel plate electrode is an extension to the parallel plate what force drives the electrons along that extension?  I has to be the repulsive force in the plane of the electrode, so that force does do work.
...

I have constructed a diagram with a protocol that would allow the energy of the repelling charges to be recovered, see attached file.

It is a plane capacitor with each plate in 2 parts, a central disk and a flat ring around it, which can be connected to the central disk. We take by arbitrary choice that the rings added to the central discs double the capacity of the discs alone.

We have the impression that the cycle would allow a gain of energy, see attached file, except that...
Once the capacitor is charged with the rings disconnected, the energy is W=Q²/(2*C). But after connection of the rings allowing the charges to spread through the payloads, the capacity is doubled, so the recoverable energy of the capacitor will only be W=Q²/(2*2*C). The energy dissipated in the loads is well taken from the capacitor.

It would be necessary to use the repulsive force without changing the value of the capacity. Probably this is not possible.

[Smudge]

With all the theory put forth on this thread, I'm hoping someone can explain the following simple parallel resonant circuit C1 displacement current to me.  This is a simulation but a bench circuit with equivalent values performs exactly the same.

The driving source is a low impedance sine voltage generator V1 running for 1 cycle of the resonance frequency at 22.508kHz.  Several things to note IMO.  First, the rise of IC1 at the very start of the cycle to a level of ~28ma.  I have added some esr to C1 which produces the rise time shown which is very close to bench circuitry.  Second, the peak current ratio of IL1 to IC1 is ~2 and the apparent load to V1 is a constant current of ~28ma.

That 28ma is for that first cycle.  You don't say what resistance you give L but whatever it is the load current will not be constant, it will rise over many cycles, the time constant of that rise being L/R.

Next is the same circuit but V1 is turned off at the 22.21us point in time thus clamping VL1 at ground level.  The current IL1 is clamped but note what happens to IC1.

Yes, C1 now sees a short (the zero impedance source) so it discharges to zero volts and zero current over the same CR time constant as its initial charge.  L1 is clamped at that 56mA level but in practice that would decay over the L/R time constant.  With that short across the LC circuit it can no longer resonate.

Last is the same circuit driven by a constant current sine generator I1 that has a peak current of 100ma but is shut off at .3 of the cycle or ~13.3us.  At this point in time, the current form I1 is removed from node VL1 and L1 and C1 are still connected.  Again note what happens with IC1.  BTW, the current directions arrows for L1 and C1 are towards ground respectively.

Now we have a different regime, when the infinite impedance current source is switched off the LC circuit can continue to resonate as shown.  At the start the rising sine wave current gets shared between the L and the C with initially C taking all the current.  The IC1 and IL1 current waveforms there so not follow the sine function.  If you did not switch off the current at 13.3uS the L and C currents would gradually build up until they do each follow a sine function that is 90 degrees shifted from the drive current, they would reach peak values Q times the drive current.  That step function change in IC1 is the current in L suddenly not being sourced from the input drive, so it switches to being sourced from C instead.   

Perhaps this is well known to you all but to me this is both interesting and confusing.  I have tried to tap this so called displacement current but to no avail.

The displacement current through C1 is the IC1 current.  Tapping into it in the circuit is not going to lead to anything magic.  The only hope for anything unusual is getting inside the dielectric and doing something their.  Perhaps one type of experiment worth pursuing is to have a large ring core with a toroidal winding on it.  Place circular electrodes each side of the hole in the donut so that you have a working space in that hole.  Connect the electrodes together with a wire that passes round the outside of the core so the system becomes a transformer with a one turn secondary driving the capacity between the electrodes.  Now we can place a test device inside the donut gap that responds to the displacement current flowing through it.  That test device could be a smaller toroid that would show an induced voltage from the displacement current.  Not sure this would lead anywhere but it would be an interesting demonstration.

Smudge 

[F6FLT]

A balloon would still be a really interesting analogy because it would mean that energy is being stored in two different domains (air elasticity and rubber elasticity.)

Charges of the same sign tend to move away from each other. Unlike a balloon, a sphere of charges would be self-inflating.

Do we know if electrostatic attraction and repulsion both propagate in the same manner?

No experiment to my knowledge has been able to show a different propagation.

The last research I've read put Coulomb force propagation velocity as 'probably extraluminal' (ie: no propagation velocity)  https://link.springer.com/article/10.1140/epjc/s10052-015-3355-3
But if the repulsion force is kinematically derived, then it's conceivable that it propagates at or closer than C.
That would open up a lot of directly engineerable experiments.

Thanks a lot for the link. This is the first experiment that I find really convincing concerning the speed of propagation of the near field (without EM).
The coulombic field seems to be rigidly linked to the charge, which I always admitted. The idea that this field would be generated by the charge has always seemed to me unfounded. The charge and its field are one and the same object. We can see it as a rubber ball extending to infinity. Only when we accelerate it, it deforms, and the deformations propagate at the speed c. The charge does not emit virtual photons or other entities that would become independent, the charge is not a generator.

In other words the field of the charge does not propagate, it travels with it according to Newton's first law, it is itself a part of the charge.

[Smudge]

The coulombic field seems to be rigidly linked to the charge, which I always admitted. The idea that this field would be generated by the charge has always seemed to me unfounded. The charge and its field are one and the same object. We can see it as a rubber ball extending to infinity. Only when we accelerate it, it deforms, and the deformations propagate at the speed c. The charge does not emit virtual photons or other entities that would become independent, the charge is not a generator.

In other words the field of the charge does not propagate, it travels with it according to Newton's first law, it is itself a part of the charge.

So if I have a large sphere atop a high tower and I feed a fast rise time pulse of voltage to that sphere, and I have a measuring instrument atop another high tower some distance away, and I ensure that the instrument only responds to the radial E field from that transmitting sphere, I should measure zero time between the input pulse edge and the received pulse edge?  I am sure that this type of experiment has been done somewhere.  For my own part I had access to a Hewlett Packard TDR that had a specified rise time of 120pS max but which was actually nearer 50pS rise time.  So I could do measurements over quite small distances (that the equipment was designed to do for the transverse wave EM in cables) set up to measure the longitudinal field from an electrode.  And in my experience that longitudinal field propagated at velocity c in air.  You will of course argue that the electrons have to travel along the wire feeding the sphere where they then decelerate to become stationary, and the instrument merely recorded the distortion to the field lines that do travel at velocity c.  The only problem with that is the distortion accounts for the EM transverse wave and appears at the receiver as a transverse wave , not a longitudinal one.  I am quite sure that my measurements were not from a transverse wave so I do not go along with your perception of the Coulomb field.
Smudge 

[Hakasays]

And in my experience that longitudinal field propagated at velocity c in air.

I can understand transverse and longitudinal waves being present in every condition (surface waves on the ocean and compression waves inside the ocean, you can't have one without the other),
but it doesn't seem intuitive that transverse and longitudinal waves would/could propagate at the same velocity.

Charges of the same sign tend to move away from each other. Unlike a balloon, a sphere of charges would be self-inflating.

No experiment to my knowledge has been able to show a different propagation.

In other words the field of the charge does not propagate, it travels with it according to Newton's first law, it is itself a part of the charge.

There may be a distinction between the rate-of-propagation of the charged particle itself and the rate-of-propagation of the field.  Since capacitors work perfectly well in a vacuum.

A thought experiment would be charging a sphere in space to lets say 1 million volts, then discharging it in an infinitesimally small length of time.  How would a galvanometer 1 light-second and 10 light-seconds away experience this change?

(If electrostatic field equations are being used today without propagation values, it would seem we're already assuming an extraluminal field without realizing it)

[F6FLT]

So if I have a large sphere atop a high tower and I feed a fast rise time pulse of voltage to that sphere, and I have a measuring instrument atop another high tower some distance away, and I ensure that the instrument only responds to the radial E field from that transmitting sphere, I should measure zero time between the input pulse edge and the received pulse edge?
...
Smudge

Of course not.  At the beginning the field is zero because the positive and negative charges have their field mutually cancelled all around.
When you create the impulse, you move electrons so you shift their fields from the one of the positive charges. The ambient field is thus locally deformed, the deformation propagates at the speed c, and when it reaches the measuring instrument, this one detects the change and the evolution of the coulombic field until it is stabilized seen from its position.

In any case it is impossible to measure the velocity of the coulombic field in this way, because of the acceleration of the charges and therefore the electromagnetic radiation that accompanies it. You can't say that the wave you measured was longitudinal, it is impossible to eliminate the electromagnetic wave that pollutes this type of measurement (the same mistake made by Turtur, Monstein and others).

This is why Hackasay's paper is so clever: the field is measured when an electron at constant speed passes near the measuring device. No acceleration. This field is maximum on an axis perpendicular to the trajectory, at the moment when the electron crosses it. This leaves no doubt that at constant speed, the coulombic field accompanies the electron, it does not propagate from the electron. Only the field of an electromagnetic wave propagates.

An image of the electron could be to see its radial field lines as flexible metal rods, with inertia. Accelerating the electron bends the rods, and the deformation propagates from the center to the ends. At constant speed, the rods remain straight. In all cases they are centered on the electron and accompany it.

[F6FLT]

...
There may be a distinction between the rate-of-propagation of the charged particle itself and the rate-of-propagation of the field.  Since capacitors work perfectly well in a vacuum.

A thought experiment would be charging a sphere in space to lets say 1 million volts, then discharging it in an infinitesimally small length of time.  How would a galvanometer 1 light-second and 10 light-seconds away experience this change?
...

Same remark as to Smudge. Such an experiment proves nothing, because of the acceleration of the electrons and the radiation of an EM wave, and all the more so as the pulse has a steep front.

To make this measurement, it would be necessary to remain in the near field, in a quasi-stationary regime, thus to have a pulse with a slow front, whose wavelength of the components highest in frequency, is at least 10 times the distance between emitter and receiver.
But in this case, we lose all accuracy of measurement because we must determine a detection threshold, and as the front of the pulse rises or falls slowly, the background noise and the residual EM wave are incompatible with the required accuracy of measurements. The amplitude of the coulombic field weakens in 1/r² while the EM field weakens in 1/r, so it is very difficult if not impossible to measure the first in the presence of the second.

This is a matter of principle. The temporal accuracy is proportional to the frequency. If we want to have a temporal accuracy sufficient to measure the field velocity over a short distance, we need fast transitions and therefore short wavelengths compared to the distance, and consequently we generate EM waves.

[Hakasays]

Same remark as to Smudge. Such an experiment proves nothing, because of the acceleration of the electrons and the radiation of an EM wave, and all the more so as the pulse has a steep front.

I'm only talking about the instantaneous collapse of the electrostatic field, magnetism need-not enter into the picture yet.
(though an EM wavefront would inevitably be produced as an byproduct of this collapsing static field, that would propagate near C)

I understand that such measurements may not be practical to engineer, which is why I stated as a though-experiment.

This leaves no doubt that at constant speed, the coulombic field accompanies the electron, it does not propagate from the electron.

Let me rephrase by asking if there is a maximum velocity at which this accompanying 'coulombic field' can expand/contract?
I understand the electrons themselves are limited to C.

[F6FLT]

I'm only talking about the instantaneous collapse of the electrostatic field, magnetism need-not enter into the picture yet.
...

And how do you make an electrostatic field collapse without current?

Let me rephrase by asking if there is a maximum velocity at which this accompanying 'coulombic field' can expand/contract?

It does not expand or contract. It follows the charge, and if the charge accelerates, the field at a distance d updates with a delay d/c.

[Hakasays]

And how do you make an electrostatic field collapse without current?

Conduction current or displacement current?
Conduction current would expand from the dissipating conductor outward, while displacement current I think acts across the entire dielectric symmetrically.

Actually that lends itself to a potential experiment by measuring the electrostatic gradient across different segments of a very thick dielectric during a transient charge/discharge.

[F6FLT]

Conduction current or displacement current?
...

There is no displacement current if there is no displacement of charges somewhere, since it is a way to call their influence at a distance on other charges, through the vacuum or an insulator.

And if there is a displacement of charges, then there is a current, therefore there is a magnetic field, and therefore electromagnetic waves if the current is variable, which is the case when we collapse an electrostatic charge.

Q.E.D., one cannot do an experiment on Coulomb field without being polluted by EM waves when it changes, and the faster the variations, the greater the pollution. I have been looking for solutions for a long time in vain. EM waves have a much more powerful impact on measuring devices than the coulombic field.

[partzman]

That 28ma is for that first cycle.  You don't say what resistance you give L but whatever it is the load current will not be constant, it will rise over many cycles, the time constant of that rise being L/R.
Yes, C1 now sees a short (the zero impedance source) so it discharges to zero volts and zero current over the same CR time constant as its initial charge.  L1 is clamped at that 56mA level but in practice that would decay over the L/R time constant.  With that short across the LC circuit it can no longer resonate.

The dcr of L1 is .89 ohms and the esr of C1 is 10 ohms.  Actually the load current to V1 as defined by L1+C1, decreases until 40ms or so where it reaches 100's of ua.  C1 has reached zero volts at the end of the half cycle and the displacement current in IC1 goes to zero because it is cut off from IL1 IMO.

The initial peak current of L1 reaches ~56ma due to it being the first cycle but after ~40ms of stabilization, the current in L1 will be ~ +-28ma.  What really puzzles me here is the fact that IC1 reaches ~28ma peak nearly instantly especially with the esr removed, which will eventually be the matching current to IL1.  I know the sim math provides this result but there seems to be such a small amount of time and signal amplitude for this result to be calculated!

Now we have a different regime, when the infinite impedance current source is switched off the LC circuit can continue to resonate as shown.  At the start the rising sine wave current gets shared between the L and the C with initially C taking all the current.  The IC1 and IL1 current waveforms there so not follow the sine function.  If you did not switch off the current at 13.3uS the L and C currents would gradually build up until they do each follow a sine function that is 90 degrees shifted from the drive current, they would reach peak values Q times the drive current.  That step function change in IC1 is the current in L suddenly not being sourced from the input drive, so it switches to being sourced from C instead.

From bench experiments I have run, I find that a capacitor's displacement current can be re-gauged instantaneously at any time without any loss from an inductive source.  This really makes sense because an inductor's voltage can be changed and in that instant no energy is required so, IMO the current in IC1 is instantaneously brought to equal the current in L1 without any apparent glitch in L1's current waveform because no energy is consumed.
   

The displacement current through C1 is the IC1 current.  Tapping into it in the circuit is not going to lead to anything magic.  The only hope for anything unusual is getting inside the dielectric and doing something their.  Perhaps one type of experiment worth pursuing is to have a large ring core with a toroidal winding on it.  Place circular electrodes each side of the hole in the donut so that you have a working space in that hole.  Connect the electrodes together with a wire that passes round the outside of the core so the system becomes a transformer with a one turn secondary driving the capacity between the electrodes.  Now we can place a test device inside the donut gap that responds to the displacement current flowing through it.  That test device could be a smaller toroid that would show an induced voltage from the displacement current.  Not sure this would lead anywhere but it would be an interesting demonstration.

I haven't tried your exact experiment as you described above but I have many other combinations for inductive elements sandwiched between plates of a capacitor and I did not find any evidence of induction via displacement current in many different configurations that were tested.  The only time I saw any induction was if a flat coil was used as one of the capacitor plates but this I discount as not meeting the true criteria.

I would be interested to know if anyone has been successful in creating induction from displacements currents!

Sorry to be so late in responding but have had other stuff to take care of.

Regards,
Pm

Smudge

[Grumpy]

I'm only talking about the instantaneous collapse of the electrostatic field, magnetism need-not enter into the picture yet.
(though an EM wavefront would inevitably be produced as an byproduct of this collapsing static field, that would propagate near C)

I understand that such measurements may not be practical to engineer, which is why I stated as a though-experiment.

Let me rephrase by asking if there is a maximum velocity at which this accompanying 'coulombic field' can expand/contract?
I understand the electrons themselves are limited to C.

Nikola Tesla claimed to experience a super-charging effect on the closing of a switch, observing large DC dynamos as I recall.  Master Ivo on YouTube has a few videos where he explores it and then does some interesting work with the "extra coil".

My suggestion is to apply a high voltage unidirectional impulse to an air coil of very high self-impedance and then look for a charging effect from the longitudinal wave that is produced.  You will find that there is no magnetic field with this wave.

Tesla claimed to have emasured their velocity at 1.5c.  There are many, many papers end experiments if you look for them.

[Centraflow]

Grumpy, if you look at the switching of the mosfets in STEAP you will see the abrupt switching of the C1 charge.

Regards

Mike