The homotenna (homopolar motor + antenna).

[broli]

Sometimes the boxing gloves are too hard to put aside.

This design is yet one more in the stream of homopolar concepts. I have thought about this design a long while ago but my knowledge of electro magnetism was fairly limited back then compared to now. So I just put it on the back burner until recently.

Theory:
For me the holy grail of a homopolar motor was to cheat and somehow break physics "laws" and come up with a single wire piece carrying a current. Specifically this would break Kirchoff current law as there will be current nodes that have current leaving without equal amounts entering and vice versa. But this is called the maxwell correction in "physics".

What i then realized is that a simple dipole antenna is really such a simple wire piece. The only problem is that current has to oscillate, if it didn't you would be stuck with a polarized antenna having an excess positive charge on one end and negative on the other. However there's a problem with this oscillation. More specific for the homopolar crowd, the forward current would torque the magnet one way, but the inevitable return current would torque it the other way...and you end up with a zero net energy from torque.

Concept:
The way I tried to solve this is by electromagnetically shielding one half of the magnet. Make no mistake the shield is not ferromagnetic. It is a simple but preferably good conductor that reflects the electro magnetic wave from the wire. Because the shield is a simple conductor, when rotated with the magnet the field remains is unchanged, as if the magnet was stationary and there was no shield. Because the conductive shield does not shield an unchanging field.

Problems:
The first notable problem is rotation speed. As seen from the animation the magnet rotation has the same frequency as the oscillating current. For good shielding to occur, a good conductor and high frequency are needed. However it is not always practical, perhaps impossible to rotate the magnet at 10 000 Hz (600 000 rpm) which is still relatively low for an oscillating current.

The next problem is a conceptual one. We know electromagnet waves can be reflected off metal surfaces. However in our case we have to ask whether they also have a physical reaction. The worst that could happen is that they somehow cause a counter torque. This means that the wave transfers its momentum in the worst way possible for this concept.

Solution:
The first problem can be easily solved by splitting the shield into more regions around the magnet. The second problem can only be uncovered through experimentation.

In conclusion I always hope I get to experiment with these concepts. But discussion on them can be equally important, I hope I shared something that can lead someone to somewhere.

[broli]

Some more input. What truly opened my eyes to em waves was falstad's applet, his famous one is the electronic circuit desinger. However he has many other incredible applets. One of them is this:

http://www.falstad.com/emwave1/

It allows you to truly see how em waves behave in realtime on a crappy pc. I truly respect that man for such advanced tools.

Now back to the concept. I took a plane source and moved it to the center. Then I decreased the frequency of it. Next I drew a closed box, assuming one half of the magnet is inside this box. In the image below from the sim you can see the setup. This should be seen as a top view. The red color indicates a magnet field going into the screen and green going out. What's also noticeable is the current in the wire and the induced one in the shield.
This setup has the wire radially approaching the magnet unlike the above one. Both can be used, but this one adds symmetry and better insight.

Now what can be seen is actually interesting. The reason why I thought about this is to eliminate the spinning circuit which makes a homopolar work. However the reaction of this is that that spinning circuit creates a back emf lowering applied current and thus reducing torque.
If the induced current on the shield is analyzed with a hypothetical magnet and spin direction it can be shown that this current will react with the magnet(like a conventional HPM because the magnets' field is not changing), but in the same direction of current. Thus it will help current flow so the shield does an even better job.

[exnihiloest]

...
Because the shield is a simple conductor, when rotated with the magnet the field remains is unchanged, as if the magnet was stationary and there was no shield. Because the conductive shield does not shield an unchanging field.
...

If there is no ferromagnetic shield, how are you expecting for a flux variation through the external circuit?
A simple conductor cannot modulate a constant magnetic field, expect if it is in a closed circuit.

[broli]

I'm not. What you mention is what exactly what should be prevented. The constant field of the magnet can go outside the shield. However the field of the wire cannot go inside the shield and torque the magnet.

You want a constant external field of the magnet. Or else that variation effects the wire and causes it to induce a current and you don't want that.

[gravityblock]

The shielding material will act similar to the braided or woven wire around the dielectric of a coax cable (B in the image below)?  We can use flat pieces of copper wire on top of thin pieces of tin foil which will be split up into regions around the magnet according to the osscilating current, since we won't need the flexibility of a braided wire, and this should offer better shielding performance.

I really like this concept.

GB