The patient is a 2010 Dodge Avenger R/T with the Federal-spec 2.4L I4/4 speed automatic. It has ~117K miles.
My son's original complaint was that the battery light had come on while driving to an appointment. I had him drive the car back home and lent him my truck so he could get where he needed to go.
Thinking I possibly had an alternator problem on my hands, that evening I checked the voltage at the shock tower connections and found it to be around 13.6V. I performed a voltage drop test between the positive shock tower terminal and the B+ terminal on the alternator while it was running, and only found about 50mV of drop. I also used my digital oscilloscope to monitor the PWM waveform of the voltage regulator signal from the PCM and observed a crisp square wave. The drive belt for the alternator is good, as is the tensioner pulley. The alternator seems fine.
Testing the battery, both with my own electronic tester as well as one at the auto parts store with a carbon pile resulted in a "replace battery" message. I verified there was <80mV of voltage drop across the battery cables between the battery and the shock tower connections while the engine was running and battery voltage was being maintained by the alternator at 13.6. So I replaced the battery. The terminals on the cables as well as the new battery posts were shiny clean when I was done. There is no corrosion on the cables themselves, and the insulation is intact.
Today my son reported the battery light had come back on. He said that he was accelerating hard when it happened. He also had discovered that restarting the engine cleared the battery light, but the problem would later reoccur on another hard acceleration event. (He's 20 and "reboots" things reflexively, I guess. He is also a little hard on cars.)
Once again, when he got home I checked the basics: charging voltage at the shock towers ranging between 13.6V and ~14.5V. I also reconnected the oscilloscope and observed a nice square wave on the alternator field control circuit at the back of the alternator.
Having a little more information this time, I took the car out and beat the snot out of it. Sure enough, after a hard acceleration, when I lifted off the accelerator the battery light came on. It starts happening after the engine is warmed up to full operating temperature.
When the battery light came on, I immediately pulled into a parking lot and checked battery voltage with my DVM. 12.2V and slowly dropping. I also checked the waveform of the voltage regulator field control circuit at the alternator, and found nothing but some low voltage AC noise on the circuit.
Stopping and restarting the engine returned the charging system back to normal operation. The DMM showed charging voltage between 13.6V and 14.5V, and the square wave returned on the field circuit. I drove the car home.
After some digging I found a wiring diagram and traced the field control circuit back to pin 51 of connector C2 at the PCM. I did a simple continuity check from the PCM to the alternator and verified I had the right circuit. I set up a back-probe with a T-pin and a short length of wire soldered to it, secured it under the hood where it couldn't short out, and took the car for a ride.
I reproduced the problem, drove the car back home (it was dark by now) and checked the field circuit with the oscilloscope at the PCM connector. It was showing the same low voltage AC noise as before, and the battery voltage was at 12.0V. Again, stopping and restarting the engine restored the charging system to normal operation.
Do I have a bad PCM? It sure seems like it stops turning on the alternator when certain conditions are met. There are no untoward voltage drops, obvious mechanical problems, strange noises, or visible damage to any components.
