I could be wrong...  this is just kind of my guess.  I think ideally - to make best use of the magnets and the wire, we should probably have about half the rotor filled with magnets.  And I believe - that filling perhaps 3/4 of the rotor with magnets will cost us very little.  But I've not tested all this... just speculating.  If it were me, I'd not go with 13" rotors unless you also went with larger magnets and larger coils.

Now some of my questions. Why would I get an increased output going from 12" to 13"? Is this due to the increased magnet velocity? It was stated that the coils would have to be bigger using a larger dia. and that I'd have lots of "dead" copper adding resistance. Is this due to the magnets being moved out which effectively changes their spacing which would require the coils be changed in width to match the magnet spacing?

And finally, this ties into another posting of mine regarding the leading and trailing edges of adjacently placed magnets as they pass over a single coil. Seems like if you have 12 magnets and 10 coils aligned over the same dia. this isn't going to happen regardless of the dia.

Using the same coils spaced out to 13" would result in no significant difference.

It may be possible to use slightly thicker wire for the same number of turns but the increase in length will nearly offset the reduction in resistance from the increase in section.

... Voltage increase with 13/12 in each coil and is linear with velocity increase.

Actually the approximate center of the magnet, rather than the edge of the disk, is the appropriate radius to examine.  Daleh007 didn't say what the magnet diameter was, but assuming it's about 2" (putting the center about an inch in from the edge of the disk) the correct ratio would be 12/11 rather than 13/12.

However, when you place the coils on a larger diameter and the coil shape isn't change then in theory 1pi/12pi of the circumference is a dead spot if the flux is very concentrated ( dual magnet rotors).  Then the new total Voltage will be 48.(13/12 - (1.pi/ 12.pi)) = 48 Volts.

But that assumes he's interested in an average voltage.  I'm presuming this is for a battery-charging service, so the peak is more interesting and the (1.pi/12.pi) term goes away.  That makes the voltage of interest 48 * 12/11 = about 53.3.

That assumes the coils have the same number of turns but are expanded to match the overall spacing of the magnets, so that the incoming magnet still starts entering one side just as the outgoing magnet starts leaving the other.

If the width of the winding is kept the same as the OLD gap between consecutive magnets, and the gap between the rotors is kept the same, you'll get that break in the waveform.  But if you spread out the coil to match the new gap, while you get slower rise and fall slopes you'll be able to make your coil thinner.  This will let you decrease the gap between the rotors, which will increase the field.  That will raise your peak voltage further (although the gentler slopes will reduce the fraction of time per cycle that the voltage is near the peak).