Fred wrote:

If the unit is as you say a brushless unit it's unlikely to be self exciting
and therefore unlikely that any residual magnetism will be sufficient to for
the rotor voltage to overcome the forward diode drops. If it's brushed then
it is more likely since brushes have very little forward drop.

The one I looked after was definitely self-excited. The only battery
involved was for the engine starter motor.

I would have thought the AVR would have terminals for a battery for the
initial excitation? Given that this is a 1980's alternator I doubt the
electronics at the time would have stretched to supplying 71V and 18A.

Did you read my description properly? That level of voltage and current
are for the main alternator and the only 'electronics' involved is the
rectifier diodes built into the rotor. The PWM o/p from the AVR was in
the same 16 V, 1.8 A range mentioned by the OP, and was handled by a
single TO3 power transistor - possibly an OC35(!) although I can't
remember for sure.

--
Andy