Hi Jamie -
"Before jumping in with both feet and building a design of my own, I'm going to try a DanB 17ft machine.
Dan posted in his notes that with 16 3x1.5x.75 inch magnets on a 16" rotor left only 3/8" between magnets at their base.
I'm thinking of using 17" rotors to help with this problem but then the coil edges wont be touching in the stator due to the slightly large spacing. Will this present a problem?"
It wouldn't hurt - it'll give you a bit more space, you could enlarge the center of the coil. In the end it will change things a bit - you'll use more wire in the coils, resistance will be a touch higher but perhaps with the more optimal arrangement you could use a few less turns. I doubt going up 1 inch will give you room to use larger wire. The main reason I built that with 16" rotors was... 8" is the largest radius my drill press could deal with, so I made that compromise so I could be sure to fit the rotors on my drill press. In the end though - I had all the work done with a water jet so it did'n't matter. I don't think I'm losing a great deal packing the magnets in close like that - maybe a bit.
"Also the rotor plates are 1/2" thich. I used 1/2" rotors on my last machine that had 2x1x.5" n40's and i could not a bit of flux leakage on the backside of the rotors. won't this be much worse with 3x1.5x.75" n40's ??
should i use thicker rotors? if so any suggestions?"
I think half inch is fine, but it's probably at the limit... there is a slightly detectable magnetic field on the back side of the rotors.
"Dan, did you notice any cancellation in the coils with the close proximity of the magnets?"
Not really. There must be a moment of cancellation on the inner parts of the windings because there is a moment when each leg of the coil is seeing both a N and a S pole. It's not nearly as noticable on that machine as it is on my old 14' machine and MattB;s 15' machine where we used large wedge magnets packed in pretty closely on 14" diameter disks - there you could see a noticable defect in the sine wave on a scope. It's a bit of a hard call for me to say if you'd be better to use a larger disk (and more copper) and optimise the magnet rotors a bit bit more - or leave it as is, save on some weight and copper, it seems to work pretty well.