I've spent a fair amount of the past summer working on my son's 97 EX, 5 speed. He was having trouble with it not starting. The local AutoZone said bad alternator. I put in one of theirs. Problem resurfaced. Someone else said battery, so we replaced that. Problem resurfaced. Replaced starter and it's been starting great ever since. HOWEVER, I also did other work, including installing a voltmeter on the A pillar, putting in a new power window regulator (again) on the driver's door, reconnnecting the a/c fan cable, etc. I drove it back to him with zero problems, volts reading right around 14 all 90 miles. We putter around town and I see the volts down to 12 or so, with it jumping back to 14 now and then. I verify the voltage reading with a digital gauge in the cigar lighter. Having nothing better to do with my time, I try turning various things off and on while my son drives. Lo and behold, the volts stay perfectly steady, at 14, when the a/c is on. Turn it off, it alternates between 12 and 14, tending to stay at 12 or so. Turn the a/c back on, it immediately jumps back to 14.

It does have an aftermarket alarm, installed by Circuit City. I think that's the only electrical modification other than the volt and oil pressure gauges. I disconnected the gauges, just in case my wiring of those was causing problems. No change. The alarm functions normally.

It's a 5 speed, 4 cylinder EX sedan, with about 150K miles. Runs fine otherwise, although I hate the low brake pedal! I'll whine about that in another post.

Thanks.

John Livingston
Newport News, VA

 

That may be the natural action of the voltage regulator? I never monitored my car's voltage while driving, so I am not sure.

It could also be from a slipping alternator belt. I would inspect the belt for cracks, etc, because it may need to be replaced.

I would clean all ends of both battery cables, then tighten them back down. Also check the electrical connection at the alternator for signs of corrosion.

 

The ECU has the ability to "switch off" the alternator completely, so it freewheels to improve MPG. That happens under the situation of:
- Engine is warmed up
- Battery voltage is good (I don't know the actual voltage criteria)
- No electrical loads measured by the ELD

The trick here is the electrical loads measured by the ELD. The AC will cause the radiator fans to run, so that load would be measured by the ELD.

If you want to do some testing just for acedemics sake, look for the "C" terminal of the alternator wiring. Disconnect that. Now try your measurement again.

You see, when the ECU wants to "shut down" the alternator, it grounds that alternator "C" circuit. If that circuit stays open, the alternator should keep charging.

HOWEVER... that alternator shutoff feature should not cause the battery to go deal & fail to start. Perhaps there's some insulation rubbed off that "C" wire allowing it to occasionally short to ground (shutting off the alternator).

 

I had no idea the ECU can tell the alternator to cut out if there's no significant load on it. The battery has been staying charged, I have to admit. I'll have him try it again with the a/c off and the headlights or, say, the rear defroster on, and see if that keeps the alternator engaged.

Pretty clever, these Japanese!

John