I'm familiar with the heat and freeze technique, but it won't work in this case because the sheave is made out of cast aluminum -- or aluminium for our friends in the UK. ;)

Woof, getting a piece of steel behind the magnets won't be possible since the poles are going to be in the same plane as the disk/ring itself. I've done some testing on one coil, 600 turns of #26 wire and I can get about 12.5 volts at roughly .75 amps just cranking it by hand. With 36 coils (18 on each side of the disk), it'll more than meet my needs for this project.

Mike, I thought about using individual plates -- still am, actually -- my only concern is that a magnet will undoubtedly want to be in a space occupied by one of the spokes, and that, as they say, is a "sticky wicket."

I took a look at an 8" sheave made by the same manufacturer (Congress) and it turns out there is a narrow flat spot at the bottom of the V-groove, and I can probably manage to get a long, skinny screw or brad into it. What I might be able to do is first drill a series of holes around the perimeter, then "pin" the ring to it via some kind of fastener through those holes.

Another thought I had today is to get some of that red rubber gasket material in the form of a strip, its length being that of the sheave's circumference. The ring's inner circumference would be just a hair larger than the sheave's, and in this way I could achieve a press fit; it would allow me to get the ring positioned so that it spins true, and then fasten it using the technique mentioned above and/or with a generous amount of adhesive, probably urethane glue. (Whew, now that's what I call a run-on sentence!)

I'm still open to suggestions. I still think this method will work out better than having to drag out the router, cut a circle, make a hub, yadda-yadda-yadda.

Progress: I ordered the 14" sheave from Grainger last night so it'll be here in a few days. Also decided to go with fine grain particle board for my ring since I'm more of a woodworker than a metalworker. I think the idea of a press fit will work out; there'll be just enough room to make the necessary adjustments so that it runs true.

Tim, I know from your previous posts that you're into this sorta thing so I'll just say this much and I know you'll understand: I don't intend to kill the dipole. I've already done a great deal of preliminary testing and I'm amazed by the results. When it's all said and done, I'll post everything here. In the meantime, if you want to chat about it via email, drop me a note at gspirit at trailnet dot com.

Mos, as for the brakedrum, I did one of those a while back -- quite an awesome beast but waaaay too much weight for this particular application.

Thanks so much for all the feedback!

Cut some tabs from a length on angle iron (1x1 or 1.5x1.5 1/4" thick), next drill 2 holes on one side large enough for screws to attach to the disk.  on the other face, drill a single hole to be tapped for 5/16 or 3/8 NF thread.  Locate this hole a little less than half the v-groove's width from the inside corner of the angle.  Get 1 bolt and one lock-nut for each angle iron tab.  Install a nut on each bolt, as far as the theaded length allows, then grind a hemisphere onto the end of the bolt (or use bolts with a "dog-point" already machined on them).  Now run the nut over the ground area a few times to clean the threads, then insert the nut into the angle iron tab from the outside of the angle (so the dog-point is on the inside of the angle).  Attach a few of these to the ring's back (sheave) side with equidistant spacing.  Locate them so that the corner in the angle iron overhangs the edge of the ring by just a little bit and the bolt/locknut is not aimed at the disk, but at the sheave.  You should now have a clamp on ring that can be attached to the sheave by way of the clamping bolts seating into the v-groove.  Radial run-out can be adjusted by loosening a bolt and tightening the one (or ones) opposite from it.  Adjustment is easiest if you just use 3 bolts until runout is perfect, then seat any other bolts that you may have included for extra strength.  Make sure to tighten the locknuts.  If the machining of the angle iron puts the bolt hole's centerline a little bit short of the v-grooves centerline, there should be no axial runout, as the bolts will pull the ring tight to the side of the sheave - if there is axial runout it is either built in to the sheave, or something wasn't drilled right, or the runout is present in the ring, no matter how you mount it, or the ring is floppy material, and the bolt tension distorted it - in this case see if you can back off the bolts a little bit without getting any slipping.  

Think of it basically like a spoked wheel where the spokes push against the hub, instead of the spokes pulling against the rim.

Also -- in this particular project (non-wind), the metal backing for the magnets would only get in the way. I may end up using something akin to laminations on the coils if I can find a suitable material. Silicon steel seems to be quite difficult to find, not to mention that it's pretty pricey even on the surplus market.

Nack -- that's an interesting approach. If I end up building another one of these I'll probably go that route or something very similar. In this particular case, I'll be taking advantage of both sides of the magnets, i.e., coils on both sides of the sheave, so what I really need is a ring that fits over the sheave; the magnets will go completely through the thickness of the ring.