I welcome any input on this topic, as I may have the entire thing wrong. I started this thread because I keep seeing people explain the waveform output of an alternator as "pulsed DC, alternating voltage with same polarity," etc.
I am pretty sure it is a misconception that a dual-rotor configuration consisting of all North poles on one and all South poles on another will produce a waveform above the zero reference (pulsed DC). The reasoning is simple.
If we are using a coil on a stator, it lays flat, and the magnet must pass through both sides of the coil. Using the left hand rule, it is easy to easy that once the magnet, regardless of its polarity, passes over the left side of the coil, it will induce voltage in one direction. Once it passes over the other side of the coil, it will induce voltage in the opposite direction.
A loop of wire is typically wound in a circle that lays parallel to the stator medium. As such, the wire winds up on one side (in the direction of current) and down on the other. Thus, the wire "sees" an alternating field, not only in magnitude but also polarity.
Now... alternating poles versus fixed poles...
I'm going to take a shot at some figures here. The arrangement is hypothetical, of course, and is not a recommended configuration. It is assumed that the coil winding is seated on one fourth of the stator, perpendicularly, at 90 degrees between windings 1 and 2.
Fixed Poles:
Magnets: 2 (same polarity, arranged directly across from each other on rotor)
Coils: 1
Turns per coil: 1
Induced voltage per winding pass per magnet: 1
RPM: 60
RESULT
Cycle analysis (passes in 2 rotor cycle): +1v, -1v, +1v, -1v, +1v, -1v, +1v, -1v
Voltage: 2 Vpp, 1 Vp, 0.707 Vrms
Frequency: 2 Hz
Alternating Poles:
Magnets: 2 (alternating polarity, arranged directly across from each other on rotor)
Coils: 1
Turns per coil: 1
Induced voltage per winding pass per magnet: 1
RPM: 60
RESULT
Cycle analysis (passes in 2 rotor cycles): +1v, -1v, -1v, +1v, +1v, -1v, -1v, +1v,
Voltage: 2 Vpp, 1 Vp, 0.707 Vrms
Frequency: 1 Hz
It is noteworthy that between each of these passes, the voltage will drop to zero, giving something more like this:
Fixed: +1v, 0v, -1v, 0v, +1v, 0v, -1v, 0v, +1v, 0v, -1v, 0v, +1v, 0v, -1v
Alternating: +1v, 0v, -1v, 0v, -1v, 0v, +1v, 0v, +1v, 0v, -1v, 0v, -1v, 0v, +1v
If we plot both of these waveforms, it is easy to to see that the fixed arrangement will produce a sine wave more or less, while the alternating arrangement will produce a distorted waveform. Even though the full-cycle frequency differs between the two waveforms, the net change in amplitude at each quarter cycle is the same--1 volt. Thus, both waveforms SHOULD (please correct me if I am wrong) have the same total power.
The point is, you cannot get "pulsed DC" out of a coil/magnet arrangement unless you completely remove one of the winding directions from the field. Each wire would have to be wound in a given direction, looped away from the field, and then wound again in the same direction.
I suppose you could create a doughnut-looking stator that encapsulates a rotor in its complicated winding arrangement, but why? Just use a rectifier.
The reason alternating poles are used is not because it is the only arrangement which will produce AC. It is used because, arguably, it creates an elegant and efficient NET flux path between all the magnets.
Sorry for the rant, but I have seen this argument go on so many times in these forums that I felt compelled to register and offer some insight.