I suppose I should point out that this is not a true three phase alternator design. If the coils move clockwise over the magnets, the red phase will generate a negative voltage pulse due to the decreasing positive flux through the red coils, the blue phase will generate a negative voltage due to the increasing magnitude of the negative flux through the blue coils, and the green phase will generate a positive voltage due to the decreasing magnitude of the negative flux through the green coils (a negative times a negative gives a positive in this case). If one connects the three phases in series, the voltages from the blue and green phases will nearly cancel. Since the magnitude of the flux through the blue coils increases faster than the magnitude of the flux through the green coils decreases, the magnitude of the voltage from the blue phase will be larger than that of the green phase resulting in a net negative voltage. This voltage will add to that from the red phase resulting in a net negative voltage when all three phases are connected in series. There will be a net voltage waveform when the three phases are connected in series no matter how they are connected. A true three phase alternator will always have zero voltage when the phases are connected in series. That is what makes possible the delta configuration for the three phase alternator. So, this would not be the most effective use of the magnets. This design suffers from the same issues as the (rotating) single magnet three coil design.
