Since nobody has said its an old idea, or a terrible idea (yet), I will order the supplies to build the transformer today.  For the coils, I'm thinking of using ferrite cylinders, small, as well as ceramic magnets for a weaker field.  When playing with a halbach array in sims, it seems like the magnets need to be of roughly equal strength, so I think need to be able to gain the same strength from the pulsed cores.

Should be:
2 x ferrite cores, 1" diameter 1" length.
3 x 1" x 1" ceramic cube magnets 
3 x 1" x 1" neodymium magnets
A large quantity of 26 awg magnet wire

I'll try it with both ferrite and air cores, as well as both ceramic and neodymium magnets.

Here's an implementation idea

Quote from: void109 on July 12, 2010, 10:45:47 AM
Since nobody has said its an old idea, or a terrible idea (yet), I will order the supplies to build the transformer today.  For the coils, I'm thinking of using ferrite cylinders, small, as well as ceramic magnets for a weaker field.  When playing with a halbach array in sims, it seems like the magnets need to be of roughly equal strength, so I think need to be able to gain the same strength from the pulsed cores.
...

@void109

Wouldn't it make sense to initially use electromagnets in place
of permanent magnets? Precisely so you could make the field
strengths equal. This seems to have happened in the toroids.
Permanent magnets are quite powerful and it is hard to get a balanced
magnetic field strength system with them when using them with
electromagnetic coils. This might cut down on the net number of
coil turns required until you've proven operation.

One would simply not count the substitution coil current (power)
being dissipated understanding that these would eventually be
replaced by permanent magnets. These could be designed in
taking time to not have them affected by magnet erasability
issues of non-neo magnets. In NEO's the external fields do not
totally penetrate the interior of the magnets.

:S:MarkSCoffman

QuoteWouldn't it make sense to initially use electromagnets in place
of permanent magnets? Precisely so you could make the field
strengths equal.

@mscoffman

Great point!  Thank you!  That will make it much easier to determine whether or not this will produce useful output.

Hi Void,

Your last representation makes it clear now. This is a very interesting design.