George White wrote in a message to Greg Mayman:

 GW> I can't think of any designs that were star, all the ones I came
 GW> across were 3 phase delta, 6 diode bridge for the main output and a
 GW> smaller 3 diode bridge in parallel with half the main bridge for
 GW> the field excitation.

 GW>                 ___________________Main
 GW>                 Ý       Ý       Ý
 GW>                 -       -       -
 GW>                 ^       ^       ^
 GW>                 Ý-------+-------+---->Ý---Ý
 GW>                 A       Ý-------+---->Ý---ÝField
 GW>                 Ý       B       Ý---->Ý---Ý
 GW>                 Ý       Ý       C
 GW>                 -       -       -
 GW>                 ^       ^       ^
 GW>                 Ý       Ý       Ý
 GW>                 -------------------Ground

 GW> Where A, B and C are the three points of the delta.
 GW> I suspect a star winding is better where there is a potential
 GW> inbalance between the phases (say the power grid) and a delta
 GW> winding where there is balance between the phases (as in a vehicle
 GW> alternator).

Hm.  This has some implications with my thoughts regarding tapping into an
alternator before the diodes,  and coupling it to a transformer.

By necessity it'd be only one phase,  and that would tend to imbalance
things. I've been toying with the idea for some time now but am not sure as
to how well it'd work considering the frequency a typical alternator would
be operating at...

---