A three-cylinder water pump with one dead cylinder can still fill a municipal water tower. A generator with one bad diode of the six will only be able to develop a maximum of exactly one third of that generator's rated output current. The standard generator for your car is a 55 amp unit which is very small by current standards, and was not much back in 1984, but it was sufficient when cars did not have a lot of unnecessary electronics. If your generator has a bad diode, the most you will be able to get from it during a full-load output current test is 15 to 20 amps. If the electrical system needs more than that, the battery will have to make up the difference until it slowly runs down over days or weeks.
All AC generators put out three-phase output current which provides a very stable and smooth voltage. The voltage regulator maintains that voltage between 13.75 and 14.75 volts. That is just the right amount to force electrons into the lead in the battery's plates, so they can be stored there. Just like you would get little pulses of varying water pressure from the three-cylinder water pump, you get little pulses of voltage variation from the generator. That is "ripple voltage". That cannot be measured directly. It can only be measured with a professional load tester found in repair shops. The testers found at some auto parts stores do not test for ripple voltage. That is why I send everyone to their mechanic for this test.
When a diode has failed, (those are one-way valves for electrical current flow), you lose one phase of output current, and during the time that missing phase is supposed to be there, output voltage drops a lot. That larger difference between the highs and lows is that high ripple voltage. My tester is sitting in the back of my minivan right now. I just performed this test on a friend's minivan, with the same results. The most we could get from the 90-amp generator was 28 amps, and ripple voltage was off the chart. Charging voltage was perfect at 14.4 volts, but the battery had to be charged with a portable charger about every other week.
The voltage regulator monitors system voltage, and most of them respond instantly to the low end of the ripple voltage by bumping up the target charging voltage. The regulator does not know when a diode has failed. It just assumes you're running a lot of electrical loads, and it needs to bump up the voltage to produce more current to meet the demand. The problem is when the next two working phases show up, output current is momentarily normal, plus the regulator is running the generator harder. It takes longer for it to decide to bring the target voltage back down. Before it does that, the next missing phase is detected, and the regulator continues to bump up the desired voltage. That is what you are seeing with the 14.75 volts.
This was one of my prepared "bugged" cars I built for my students to diagnose. The first time they saw the "battery" warning light on the dash, they became confused because charging voltage appeared to be fine. At the end of the learning experience they saw that charging voltage actually went up about 0.4 volts when the defect was switched in, and had they checked initially, they would have found full-load output current dropped from 65 amps to around 20 amps. Everything on the car worked fine, except for the "battery" light being on, because that car had very few electrical toys.
GM owners have a real lot of diode trouble with their 1987 and newer generators, and there are special considerations for those cars. For the rest of us, a diode can fail in any brand of generator, and while it isn't very common, that is the first thing we test for when we run into the symptoms you described. There is no value in searching for an elusive cause to a problem when a simple solution could be right in front of us.
Wednesday, April 19th, 2017 AT 11:43 PM