I don't mean to take over this conversation, but I can add a few comments of value. Everything you've said regarding charging voltages is correct. We like to see between 13.75 and 14.75 volts, (depending on which textbook you read), but the low 15s is okay, especially when starting to recharge a run-down battery.
My biggest concern is with the AC voltage test. Without going into a lot of boring theory, your voltage regulator works by switching the field winding fully-on and fully-off roughly 400 times per second. It varies the percentage of on-time to off-time to adjust the average strength of the electromagnetic field to adjust average output voltage and current. Abruptly switching current off through a coil sets up a huge voltage spike from the rapidly collapsing magnetic field. That is highly desirable and necessary in an ignition coil, but even in a small relay, their coils can develop as much as a damaging 300-volt spike if it isn't shorted out with a diode
At its maximum output, the field coil only passes three amps of current, but even that normally requires a rather beefy switching transistor. By using this "switch mode" or "switching" power supply, a very tiny transistor can handle the job. On "DC Volts", your meter will read the average charging voltage while it filters out any AC voltage that would affect the reading. On "AC Volts", there are a number of problems with this test. First, it will respond to those voltage spikes and show an abnormally high reading if the battery is not damping and absorbing those spikes. Lead flakes off the plates in all batteries over time. What remains becomes less and less until there isn't much capacity for the battery to absorb those spikes or crank the engine. The manufacturers know this can't be avoided, and they provide warranties as long as possible to cover that. The spike-damping ability goes away long before the battery is unable to start the engine. That's why we always recommend replacing it when the generator has to be replaced on an '87 or newer GM vehicle.
The other concern has to do with the limitations of digital voltmeters. The "AC Volt" ranges are only accurate at 60 hertz, (house current). Some special meters are accurate up to 400 hertz for military use. AC generators, aka "alternator", typically have twelve output coils around the stationary "stator" winding, and there's three of those groups because these are always three-phase output. Then, those output currents are "rectified" turning AC into pulsing DC, so there's twice as many pulses. That comes out to 72 voltage pulses per revolution. Crankshaft pullies are roughly four times the diameter as a generator pulley, so you'll get, uhm, . . . math, . . . 288 pulses per crankshaft revolution, but they idle at around 800 rpm, so now we're up to 230,400 hertz if I did that right. That's a long way from 60 hertz, so while the meter will show something, it won't be accurate. I suspect what you're finding is more a result of the normal switching action of the voltage regulator. I know there's something wrong with my math because when there's excessive ripple and you can hear it on AM radio, it's in a range we can hear, which is up to only 20,000 hertz. That's still well above 60 hertz.
There's two things you can do next. One is to unplug that smaller connector from the generator and see if the symptoms clear up. Even easier is to observe if there's symptoms with the ignition switch on but engine not running.
Did you figure out the age of the battery? If it's old, consider connecting another one with jumper cables. If the symptoms clear up with that extra capacity, replace the battery.
The better way to test the system is with a professional load tester. You've already done the first test correctly by measuring charging voltage, but passing that test only means it is okay to continue with the rest of the tests. This AC voltage test Ken had you do is just a preliminary step. Professional testers measure this accurately. It's called the "ripple voltage" test. I put this sad drawing together to explain a different problem, but it works here too. On top it shows the rectified output from the three phases. Ripple voltage is the difference between the highest and lowest voltages, in this case, half a volt. It is very easy for the battery to smooth that out.
A somewhat common failure is for one of the six diodes to be defective, in which case you lose one phase. That would look like the lower drawing. Here, ripple voltage is 5.2 volts. Even good batteries will have a hard time cleaning that up. A common symptom is that is heard as an annoying whine on AM radio that changes pitch with changes in engine speed. Additionally, with one failed diode, you lose exactly two thirds of the generators maximum output. A common unit today is a 100-amp generator. With one bad diode, all you'll be able to get under the "full-load output current" test is roughly 33 amps. That's not enough to run the electrical system under all conditions. The battery has to make up the difference as it slowly runs down over days or weeks. In some cases an obscure additional symptom is charging voltage can be a little high as the regulator responds to those dips in output voltage and tries to bump it up.
The next concern has to do with digital instrument clusters. If your vehicle has that, there will be an additional wire in the four-wire plug on the generator. Normally system voltage is monitored right at the output stud on the back of the generator. Minor fluctuations that result in other places are insignificant. However, digital dashes are extremely intolerant of unstable voltage, so the voltage regulator monitors it right there where it's most critical. This is where if you unplug the smaller plug on the generator, only the battery powers the dash and that is very stable.
What this all boils down to is that professional test on the charging system. It only takes a few seconds; less time than it takes to connect the tester's cables. It will measure the charging voltage, like you already did, maximum output current, and ripple voltage. Output current will be close to correct and ripple voltage will be "low" if all six diodes are okay. High ripple voltage with only one third of expected output current indicates a failed diode. It's not practical to try to replace parts on this design. Replace the generator instead. If the generator can develop the full current it's designed for, the diodes are okay, but if ripple voltage is still "high", suspect the battery.
Be aware too there are a few tester models that can make paper printouts of the results. These will list ripple voltage as an actual voltage. (I don't know what "normal" would be). Most testers just show ripple voltage as "low" or "high" with a series of flashing lights. The results have to be written on the repair order by the mechanic. This is the best way to test generators. The least-preferred way is on a test bench. Those don't include an aged battery, and the electric motors that drive the generators aren't strong enough to get them to reach their full potential.
I'll go back to waiting to learn the solution. I hope you come back with some good news.
Feb 20, 2026 at 7:00 PM
