> So this is interesting... the old-timers did this all the time to 'check an
> alternator', and I've done it  a few times myself and to my knowledge I've
> never toasted an alternator because of it. In fact, I drove one car for
> about an hour once like that and the alternator survived.
> Having said that, I haven't done it in many years and I have noticed in many
> vehicle owner's manuals I've read, particularly from later model cars, a
> warning not to do it and so I've wondered 'why?'.
> What problem does it cause? Is this something that is more prevalent in
> later cars that didn't exist in older ones? Did I just get lucky, or what?

From what I understand, the trouble is that modern alternators depend
somewhat upon the battery to keep the field steady. Since it generates
AC which is then run through a bridge rectifier in the regulator, it
doesn't generate perfectly clean DC. The battery acts a bit like a
capacitor and keeps the voltage steady. Since the regulator watches
that voltage and manages the alternator field based on that to keep
the output steady, if you pull the battery off it has a tendency to
get erratic. The small variations on the output side can be amplified
by the regulator into bigger swings in the field coils, which can (in
theory) burn them out. In the olden days, when they used permanent
magnets and they were called "generators", this wasn't a problem--
generator output is mostly unaffected by feedback. And really, an
alternator with a good regulator will probably be fine. I think the
admonition is mostly in the vein of "unwise to press your luck". I'd
hate to have a weak (but working) regulator turn into a dead regulator
and/or a frotzed alternator.

--
John Bange
'90 Vanagon - "Geldsauger"