On 01/12/2010 10:49 PM, Stephen A. Lawrence wrote: > > > On 01/12/2010 06:29 PM, Terry Blanton wrote: >> On Tue, Jan 12, 2010 at 4:05 PM, Stephen A. Lawrence <sa...@pobox.com> wrote:
> > The field of a permanent magnet must be either anchored in the magnet's > material, or knotted around part of the magnet, as in the attached > sketches. Oops, that's wrong. The field of a permanent magnet must be anchored in the magnet's material, period. The second sketch I gave -- "bad-toroidal-mag-field-1.jpg" -- is impossible to produce using permanent magnets. You can make such a field with a current ring, of course, but that's not at all the same thing. The idea I had was that you started with a bar magnet and milled out the center, and carved the rest into a smooth ring. But the field as I drew it is not what would result. The real thing would have the field going up through the magnet material, and down on the outside of the ring, *and* down on the inside of the ring. Sorry, no sketch, but hopefully the description is clear. (And I still don't know what Steorn's magnetic cores have for the shape of their fields. For that matter I don't even know what old computer magnetic cores had for a field shape -- sigh.)