Did you measure the voltages I asked about?
Ripple voltage can be misleading. Digital volt / ohm meters measure AC voltage accurately only at 60 HZ, which is house current. Their accuracy drops way off at higher frequencies. To add to the misery, I've been doing these tests for well over 35 years, but I never used a tester that displayed a voltage for ripple voltage, so I don't know what "normal" is. Every tester I've ever used shows the relative ripple voltage as "low", "high", or somewhere in between. It's not common to find something in between. Normally ripple voltage will be low AND full-load output current will be close to the alternator's rated capacity, or ripple voltage will be high, AND the most you can get on the full-load test is exactly one-third of the alternator's rated capacity. That is caused by one failed internal diode of the six, but that won't result in the severe over-charge condition you found.
I can see that I added some confusion to the story. The diagram shows three wires, plus the output wires, at the alternator. This circuit is different from the simple and common Chrysler circuit that worked well for decades. You have a Mitsubishi alternator with an internal voltage regulator. The regulator can be replaced separately to try to solve the over-charging condition, but the Engine Computer is involved too. In the Chrysler system, the regulator is built into the Engine Computer so it can modify charging voltage according to a number of variables. The Engine Computer in your car also has some say in the desired charging voltage, but I don't know what voltages to expect on the wires. I DO know the computer would never request 17.8 volts. That leaves the voltage regulator or entire alternator as what needs to be replaced.
Saturday, April 8th, 2017 AT 7:26 PM