Thermoelectrics seem quite appealing for this sort of thing. After all, a thermoelectric generator using exhaust heat can be made decently light. And it has ZERO moving parts. So it should be quite reliable. Unfortunately, the efficiency of current thermoelectric devices is quite low. But this is not a big problem from an energy standpoint, as you are using heat that would otherwise be completely wasted. BUT, this could make a thermoelectric generator of the size needed prohibitively large. However, it IS possible that it could meet SOME of your electrical needs
Aside from the cost of solar panels or cells adding to the roof of the vehicle and boy my xB has a big roof! By adding an energy capturing device like solar cells you ADD energy to the vehicle instead of trying to capture energy from the burning of the fuel you are loosing. Of course if you had room in the clutch area of the transmission you could add a thin motor/generater on the input shaft of the transmission and use that to capture downhill energy with the clutch held in and no ICE engine braking plus if it was also a motor have yourself a hybrid.
I assume the auto manufactures create altenators to fit their application needs. Thus with this assumption in hand, there changing the belt ratio to provide higher revs with less torque would yeild the same electrial output. So the real issue boils down to finding the right balance between torque reqirement for maxium load and the correct gear (belt) ratio to obtain the maxium load.
alternator voltage output is proportional to RPM. amp output is proportional to torque. they're generally geared to produce about 14v at idle and have to be limited (voltage regulator) at any higher rpm. changing the gear ratio so it's 12v at idle might reduce some frictional and inertia losses but not much.
1991 Toyota Pickup 22R-E 2.4 I4/5 speed
1990 Toyota Cressida 7M-GE 3.0 I6/5-speed manual
mechanic, carpenter, stagehand, rigger, and know-it-all smartass
"You don't get to judge me for how I fix what you break"
I think possibly the best way to do this would be to have this thing activate itself only during braking, then free-wheeling friction would be the only counter-active force and that I do believe is minor.
Unfortunately I think that's what is currently known as regenerative braking.
A FE gauge should be standard equipment in every vehicle.
If you wanted to do something like regenerative braking you could put an electric motor as a part of the drive shaft (assuming RWD) and upon braking have the motor load and charge some capacitors or quick charge the battery for you in lieu of using actual brakes, for the first maybe 10% of pedal movement before the normal braking begins. Possibly in that area when you're pressing the pedal down and the brakes haven't engaged yet.
The only problem with this is by the time you've integrated it you might as well load up some batteries with a control circuit to hybridize the car since you're halfway there anyways.
Personally, I'd pull a drive shaft from a junked car like yours and wind my own rotor windings and everything to build the alternator as an integral part of the drive shaft assembly. If I was doing it, while I was at it I'd also make sure to wind it to put down power too. Probably wind it like some capacitor start motors where at low speeds the shaft motor is strong but not fast and then a centrifugal switch changes the motor winding setup to make for high speed, good hp.