Do a Google on single wire alternators: I came up with here.
"First off, realize that the alternator needs field power in order to
generate. Unlike an older generator, the alternator's tiny bit of
residual field is not enough to get things rolling. In a
conventional setup, power is applied to the voltage regulator from
the ignition switch. The voltage applied to the voltage regulator
and field is also the reference voltage that the regulator controls
to. The regulator excites the field at whatever level it takes to
keep the voltage at the input terminal at 13.8-14 volts. That means
the output terminal voltage of the alternator may be (much) higher,
depending on the voltage drop in the wiring between the alternator
and battery. If the vehicle has an ammeter, then there will be
voltage drop across that in addition to the drop through the
wiring.
The one-wire alternator must - by definition - regulate its output
voltage. The designer of the regulator has to allow for some
defined amount of voltage drop through the wiring to keep the
battery at the desired 13.8-14 volts. If the resistance in the
alternator circuit is higher than what was anticipated, then the
battery will be undercharged. If the resistance is lower, e.g., a
short, fat wire directly from the alternator to the battery, then
the battery may be overcharged. My experience with a number of 1
wire alternators is that the terminal voltage is set to 14.2-14.5
volts.
Two issues for manufacturers here. One, while an individual's
mildly over or undercharged battery may not matter much in the big
picture, for an OEM, having this happen to thousands of cars would
be a warranty disaster. Second, to keep the charging voltage at the
battery correct, a different regulator with different wiring
compensation would be required for each model. A logistical and
cost nightmare.
The other major consideration is field current control. A
conventional alternator draws full field current when the engine is
stopped. The reason it doesn't drain the battery is that the field
supply is switched off with the ignition switch. Since the switched
field power isn't available to the 1-wire alternator, engine stop
and start to turn the field off and on must be inferred from other
parameters. The 1-wire regulator detects engine stop by the
cessation of AC from the stator. This is reliable. Engine start
gets a bit more complicated. Since the alternator is not generating
until the field is applied, engine start must be detected by other
means. With the regulator I commonly use, this is done by looking
for the dip in voltage associated with engaging the starter motor.
If it sees a dip in voltage, it applies field and looks for stator
output. If no stator output, the field is cut off again.
The problem is, to be sensitive enough to detect engine starts under
all conditions (such as when the car is rolled off without engaging
the starter), the voltage dip detector has to be pretty sensitive.
In experiments I have done, I've discovered that the small dip
caused by switching on a single 50 watt driving light will trip the
field on. That means that the field will be momentarily turned on
from a wide variety of conditions other than engine cranking. It
won't be on long but it does consume some battery power.
Again, for the individual user, this isn't much of an issue. Most
1-wire alternators end up on hotrods and old cars with few
accessories and usually none that draw impulse current with the
ignition off. At most, if the dip detector ended up too sensitive
or the car has some load that trips it regularly, the only
consequence would be an occasional dead battery. Having this happen
over millions of cars would, for the OEM, again be a warranty
nightmare.
The 2 wire setup neatly addresses all these issues and so the OEMs
stick with that design."