Magnetic Diode

[Smudge]

The subject of a magnetic diode that only transmits magnetic energy one way has been raised recently.  If this could be achieved we have the possibility of a transformer where the primary and secondary coils are at different positions along a core, and the load current in the secondary cannot influence the primary because the backward transmission can't occur.  F6 recently posted a link to work creating this effect using mechanical movement of the coupling medium.  That got me thinking and I wondered if we could use electron movement to do the coupling.  One experiment would be to use the electon movement along the neck of a CRT that has the two coils wound round the neck at separated distance.  The electrons would need some spin-polarization provided by another coil wound around the whole length of the neck, that coil driven with DC.  Or do the same thing using a DC energised neon tube.

Another possibility could use a bundle of thin insulated Fe wires as shown in the image below.  With primary and secondary coils located at each end of the bundle there will be some magnetic coupling between the coils which needs to be cancelled.  A small cancelling coil wound on the secondary in series with the primary will do this.  Now if a DC current flows along the bundle the spin polarization occurring due to the primary signals will propagate toward the secondary increasing the coupling to the secondary.  This would create the non-reciprocity we are looking for.  Although conduction current flow is very slow, this velocity will not produce a large time delay; it is more a case that the electron flow from the primary position alters the fast-acting magnetic coupling.  The actual time delay will come from experiments.  Experiments could also include the effect of applying a static magnetic field using permanent magnets at each end of the rod.

If someome takes up this suggestion they could gain some notoriety in creating a magnetic diode.  Whether this leads to OU is another story.

Smudge

[F6FLT]

In the paper I submitted, as I recall, coil 1 induces a current in a solid copper cylinder moving toward coil 2, with the two coils already having some coupling between them.
The coupling will therefore be favored from coil 1 to coil 2. But the drive of the current loop induced in the copper causes flux variations in the coils, so that this type of diode cannot generate free energy; it is the equivalent of a 3-coil coupling, the fictitious and intermediate one being short-circuited and mobile.
So, can we build a "solid-state" system based on this principle? I'm afraid we'd end up back in the case of the Hall effect or your other idea involving the magnetic gradient, resulting in negligible currents and voltages.

Indeed, in the case of this magnetic diode, the induced current loops are mechanically displaced perpendicular to the induced current. In the case of a bundle of iron wires, the current loops would be canceled out in the same way as they are in the laminated core of a transformer. Even if we used a solid metal core, the speed of the electrons along its length, due to the current, would be considerably lower than in the mechanical experiment, so the coupling would hardly change. The idea that spins play the role of current loops in each wire would hold up if we were talking about direct current, but here we're talking about varying induced currents, and barring some fantastic discovery, we can't induce them in the spin.

In a vacuum tube, it would indeed be possible to accelerate electrons that would also form current loops. But this brings us back to a familiar and well-understood field: the magnetron. The electrons would spiral, subjected to both the induced current and the accelerating voltage. It is likely that we would have better coupling from a coil near the cathode to a coil on the anode side than the other way around, but as in the mechanical case, I don't see where free energy could come from.

[Smudge]

In the paper I submitted, as I recall, coil 1 induces a current in a solid copper cylinder moving toward coil 2, with the two coils already having some coupling between them.

Let's be clear about this.  It was not a cylinder containing cylindical coils on the same axis.  The two coils were not on the same axis, their axes were parallel but separated.  The moving coupling was a U shaped channel of conducting material moving along the direction of the coil's separation. 

The coupling will therefore be favored from coil 1 to coil 2. But the drive of the current loop induced in the copper causes flux variations in the coils, so that this type of diode cannot generate free energy; it is the equivalent of a 3-coil coupling, the fictitious and intermediate one being short-circuited and mobile.

I think that is an over-simplification that could apply to axial coils in a moving cylinder, but not to their U shaped moving channel.  I don't have access to the full paper, only the description in the link you gave, so I haven't studied their math but I would not be surprised to find that the U shape is important compared to having two conducting strips containing your fictitious short circuit coils moving past the pole faces of the two coils. 

So, can we build a "solid-state" system based on this principle? I'm afraid we'd end up back in the case of the Hall effect or your other idea involving the magnetic gradient, resulting in negligible currents and voltages.

Indeed, in the case of this magnetic diode, the induced current loops are mechanically displaced perpendicular to the induced current. In the case of a bundle of iron wires, the current loops would be canceled out in the same way as they are in the laminated core of a transformer. Even if we used a solid metal core, the speed of the electrons along its length, due to the current, would be considerably lower than in the mechanical experiment, so the coupling would hardly change. The idea that spins play the role of current loops in each wire would hold up if we were talking about direct current, but here we're talking about varying induced currents, and barring some fantastic discovery, we can't induce them in the spin.

You seem to have conflated my experiment with your perception of moving current loops.  That is not the case.  You mention spin.  Spins are magnetic dipoles.  In magnetic permeable material that is conductive the positive ions and the conduction electrons both have spin.  Huge relative permeability occurs in such materials and the thermal agitation movement of the conduction electrons at huge fermi velocities play their part in how a varying induction in one position quickly transfers field to another position along a core.  The Unruh/Coler stromerzeuger uses such material and passes DC through the core, so I think it possible that DC movement of the electrons can influence how the core acts magnetically.  Not connected with the slow net speed of them but changing the effective permeability.  Perhaps this would be better investigated using a ring core with small windings diametrically opposite, as shown in the image below.  The high permeability core will give coupling between primary and secondary close to 1, say 0.95.  Can the DC current increase that to say 0.99 or even higher?  Has anyone even tried this? 

In a vacuum tube, it would indeed be possible to accelerate electrons that would also form current loops. But this brings us back to a familiar and well-understood field: the magnetron. The electrons would spiral, subjected to both the induced current and the accelerating voltage. It is likely that we would have better coupling from a coil near the cathode to a coil on the anode side than the other way around, but as in the mechanical case, I don't see where free energy could come from.

The magnetron has a magnetic field crossed with the electron beam whereas mine has the field along the beam.  The Travelling Wave Tube would be a better one to compare.  Again you are talking about current loops, whereas I am considering electron spin polarisation as the carrier.

Smudge

[F6FLT]

...
You seem to have conflated my experiment with your perception of moving current loops.
...

That's quite possible. You seemed to have described your experiment as a variation on the magnetic diode, replacing mechanical motion with electron movement. Wasn't that your idea? In which case, it would be mine .
Regarding the magnetic diode, it's the principle that needs to be understood, as the implementation can vary considerably. The coils are partially coupled to each other by a current loop induced in a moving copper mass, which acts as an intermediary for the coupling. The orientation of the coils is irrelevant as long as they have a common coupling via the induced current loops moving from one to the other. Therefore, the experiment could perfectly well be performed with parallel coils, or coils facing each other, and even the copper mass could be replaced, albeit with less efficiency, by a short-circuited moving coil. In principle, it's basic electromagnetism, but with a clever engineering twist.

...
The high permeability core will give coupling between primary and secondary close to 1, say 0.95.  Can the DC current increase that to say 0.99 or even higher?  Has anyone even tried this? 
...

The effect of a current or voltage on a magnetic field through certain materials is a hot topic in research, but in the field of nanotechnology. I have already tried to observe a difference in the magnetic effects of a ferromagnetic core with or without a current flowing through it, specifically using iron wire and conductive ferrites. To no avail. The only effects of this kind I have come across in the literature required exotic materials.

Again you are talking about current loops, whereas I am considering electron spin polarisation as the carrier.

I explained why in the context of the magnetic diode, where the currents are variable. The spin magnetic moment, however, is constant; the spin cannot move while carrying a variable current loop. It can be used, but not in the same way.

[chief kolbacict]

The subject of a magnetic diode that only transmits magnetic energy one way has been raised recently.  If this could be achieved we have the possibility of a transformer where the primary and secondary coils are at different positions along a core, and the load current in the secondary cannot influence the primary because the backward transmission can't occur.

Can a mechanical device perform this function ? 
Even if it is low-tech and noisy. 

[Smudge]

Can a mechanical device perform this function ? 
Even if it is low-tech and noisy. 

The first posts on this bench discuss such a mechanical device.  F6 posted a link on another thread, here is the link.
https://physics.aps.org/articles/v11/s134

You might note that F6 posted that link in reply to your question here.

Smudge

[F6FLT]

The full article (pdf)

Replication of the device here and more (no moving conductor)

"The objective of this phase of the investigation is to obviate the use of a moving conductor by replacing it with a static device through which electrons can freely flow, according to the suggestion of Prat-Camps." That's what we were talking about without realising it a few days ago. But the currents must be very strong.

[chief kolbacict]

The rotating transformer from the video heads of the VCR also has a U-shape.
Can I use it ?