Your vintage Baron has two regulators-- one is in use and the other is a backup. The one regulator in use simultaneously supplies the field current to both alternators. The field current creates a magnetic field inside the alternator which is used to generate output current. The higher the field current, the stronger the magnetic field and the greater the output of the alternator. The job of the voltage regulator is to adjust the field current so the alternator output is at a constant voltage (nominally 28.25V). If more output current is being consumed than is being generated, the voltage drops and the regulator increases the field voltage to keep up with the demand. The same happens in reverse-- if the alternator output current is more than is required by the electrical system, the bus voltage increases, which causes the regulator to decrease the field current.

The left and right alternator outputs get joined on a common bus bar under the nose baggage compartment and this is where the voltage regulator senses the bus voltage. This type of setup is not a true paralleling of the alternators but is instead two separate half systems (a left and a right) tied together. The most efficient half system will drive the bus voltage and the other half will "be along for the ride".

If the half systems are perfectly matched, both alternators will have identical field vs. output curves and will be outputting exactly the same current at exactly the same voltage with exactly the same input field current and the alternators will share the load equally. But if there is any difference in either half system---for example, a higher resistance on one side in the loop from the voltage regulator field, to the alternator, to the bus bar, back to the regulator, or if one alternator is stronger than the other---then one alternator will supply more of the load. This is the case with almost every Baron in existence because things are rarely perfectly matched (one can spend a lot of time getting it close). Because the currents in these loops are large (tens of amps for the alternators and 3 or 4 amps for the field), milli-ohm differences can result in large voltage changes.

Sometimes, as might be your case, the half system imbalance is so large that one alternator dominates and supplies all of the required current and effectively keeps the other alternator off, which the system senses as a failed alternator (the alternator out light will stay off as long as there is voltage at the aux pin of the alternator; when the alternator output is low or zero, that aux voltage is also low or zero). When that dominant alternator can no longer can keep up with the current demand, its voltage will start to drop, which will cause the regulator to output more field current, which will in turn cause the second alternator to start to output current and take up the load (the added field current will have little effect on the previously dominant alternator because that alternator is at, or is reaching, its limit). This could be what is happening when you turn on your landing lights.

Since the regulator is common to both alternators (both fields are supplied from a single regulator pin), the problem you describe with a single alternator is not likely due to the regulator. I would start looking at every connection in the common bus bar-->regulator-->field-->alternator-->common bus bar loop, (ground connections, too). If everything there is clean, then the next step is to look at your left alternator.

Here is some Bonanza VR troubleshooting advice from Derek for a fellow Beech owner's alternator woes:

John's alternator behavior sounds a little different, and I agree with Stuart that the root cause is probably in the field segment.  I do not think glass fuses are your problem as I believe those are only in the 'alt out' light circuit and your voltage reading shows your alt is indeed out.

But first, it wouldn't be Friday if I didn't include some theory---

As everyone knows (or should know), electric current is produced when a magnetic field moves past a wire and, conversely, a magnetic field is produced when current flows through a wire.  Adding more wires or more current adds to the effect, and an efficient way of putting a bunch of wires
in a small space is to wrap them in a coil.  An alternator uses both of these traits to produce current:  first it flows a relatively small current (the field) inside a coil to make a magnet, then that magnet is spun inside a larger coil of wires to produce a bigger current, which is the alternator output. 

The stronger the field current in the magnetic coil, the stronger
the magnet and the higher the alternator output.  The voltage regulator is there to adjust the field current to keep a constant voltage on the bus.

An aside-- it is sometimes said that an alternator produces AC and a generator produces DC, but that is not exactly true as both inherently produce AC.  The generator rectifies the AC mechanically using a commutator, whereas the alternator does so with diodes.

Oh, and another thing-- calling it a generator is technically imprecise because an alternator is also a type of generator.  What is traditionally called a generator on vehicles is really a dynamo.  Both alternators and dynamos are generators that use a coil to create the magnet.  A third type of generator uses a permanent magnet instead of a coil magnet; this is
called a magneto.

Something else to know is that the Beech DC generation system has a weakness in that the voltage regulator senses the bus voltage at the end of the very same wire that is flowing the field current.  The field current could be a couple amps, which means a very small resistance---only a couple hundred milliohms---is all it takes to totally screw up the voltage control loop.

So, WTF does any of this have to do with John's problem?

I think (and Stuart might be on the same page) that John has a high resistance somewhere in the DC generation loop, quite possibly in the field supply part of it.  I further think that the intermittent piece of it has to do with heat, i.e., at first the resistance is small enough for the system to stay within regulation, but because that resistance has a small voltage drop, the VR must demand more field current to supply the load, and that higher amperage causes the resistor to heat up, which increases the resistance, which causes more current to be drawn, which causes more heat, which causes more resistance.... until the power loss across that resistor does not leave enough for the field and the alternator output sags enough to turn on the 'alt out' light. Turning off the alternator momentarily lets the resistor cool, and the cycle restarts from the beginning.

There are lots of ways to debug the system, which include bypassing certain circuit elements, or flat out replacing them, etc., but I think the most quantitative way is to measure the resistance of the field loop and eliminate the offender(s).  For this you need a milliohm meter, which is nothing more than a Kelvin bridge that anyone can cobble together using simple components, such as this one:

http://www.aeroelectric.com/articles/LowOhmsAdapter_3.pdf

Use that to measure the resistance between the alternator output and the field input.  If it is more than, say, 300 milliohms, look for the circuit element that has the high resistance.  It could be wires, crimps, connections, CBs, switches, etc.  Check also the alt grounds to aircraft ground (e.g., engine block) and airframe ground to aircraft ground.  Replace or fix the connections that are bad, and don't be surprised if it is more than one (and I am not sure I agree with Stuart that the alt switch is a breaker.  If that were true, there would be no need for a separate field breaker.  And incidentally, there *is* a way to "pull" the flush field breaker:  ground its output and turn on the master.  <g>)

At least you have it easy with all the circuit elements within reach, try measuring resistances between the left and right nacelles of a Baron!

Good luck and try not to let the smoke out of the wires.

This Baron voltage regulator is typically a Delco PN: 9000591 which consists of very old school components that have likely not aged well. Here is a new solid state version that can be found, the Transpo PN: D591S and shown below. The Transpo lists itself as a replacement for the Delco 9000591.

Amazing that the small potted section of the Transpo is all that's required. Look for the D591S on eBay. It holds a rock solid voltage!

Click HERE for the Transpo Catalog which shows the Delco 9000591 interchange/compatibility.

Click HERE for a $70 (1/4/2022) source for the Transpo D591S.

Be sure to consult your A&P/IA for replacement suitability.