I'm messing around with Faraday's Law of Induction, and I will be using two magnets attached either side to a small bar inside two coils of copper wire, which I spin to induce a current. The design is basically this: http://amasci.com/amateur/coilgen.html My question is, what shape for the magnet should I go with to maximize the current, assuming the loops of wire and rate of rotation of the magnets are constant? Should I go with long, skinny magnets? Short, fat ones? I know I need the poles of my magnets to change constantly in order to induce the current, and I've found that squarish magnets work a lot better than disk ones, at least by the tests I've done. At this point my available options are magnets with these characteristics (the strength is not constant here, neither is the geometry): 1. Grade N45 bar magnet with dimensions of 4 inches by 1/2 inches by 1/2 inches 2. Grade N52 block magnet with dimensions of 2 inches by 1/2 inches by 1/4 inches. I will build the device so the coils are as close as possible to the magnets. Assume the radius of the coils goes out 1 inch past the long side of the magnet (so in the first option the radius is 5 inches, the second it is 3 inches), and that the two coils of wire are each one inch to the side of the magnets. Which do you think will provide more current, assuming I spin both types at the same rate? Thanks.