| The flux doesn't follow a wire round and round as implied by the looping.
| Take a cross section of the wire core - it appears as a number of parallel
| flux paths. Ideally it would be nice to have the ends of the wired joined as
| when they aren't the flux coming to an end has to cross over to adjacent
| wires so for a short distance the flux distribution is not uniform. In
| practice this is negligable with many turns of wire. Note that with
| laminated cores, the laminations are not continuous but the layers overlap
| so flux has to cross over from one layer to the next through the varnish gap
| between layers.
So with a bunch of wires, which would be more distinctly separate parts
of the core, the flux is always going to cross over anyway? So it really
won't matter if this is 1000 tiny steel wire loops, or one long wire that
loops around 1000 times.
| The main problem with a wire core is that good magnetic material is also
| relatively brittle. A thin flat strip is easier to roll than a wire of the
| same cross-section in that case and (I haven't worked this out) will have
| lower eddy current losses (longer eddy current path and higher resistance
| for a given cross-section and voltage induced in the path.) Use of wire
| wouldn't reduce hysteresis loss per se but it would reduce eddy current
| losses compared to a solid core but possibly not as much as with a rolled
| strip or conventional laminations. These are likely the historical reasons
| why wire cores are not seen nowadays.
My ultimate experimental idea is to have a topology where the core itself
spirals around the windings, which also spiral with it. It would be kind
of like those double-pretzel sticks where 2 pretzels wrap around each
other. Or like how certain snakes make love. This would then be wrapped
around to connect on the ends. Topologically, the windings do wrap around
the core wires, so there should be an induced magnetic field. Then the
field itself would be a spiral.
Then the next trick would be to use insulated steel wire and make it wrap
around itself in a double-Mobius fashion. So if you follow a bundle of
wires around, you go around twice to get back to the same spot. This
whole circle would be all a bunch of spirals. The steel wires would be
both the electrical winding and the core at the same time. I have no
idea if that could possibly work. But mentally picturing the topology
tells me it should. Of course steel is not as good a conductor as copper
so this is not likely to be of any practical value on a large scale.
| Phil Howard KA9WGN (ka9wgn.ham.org) / Do not send to the address below |
Click to see the full signature.