I'm not sure where to put this as it's not pertaining to LS vehicle. I feel you guys have more experience with odd GM problems than any other group of auto guys.

I have a '98 Cavalier. I had the battery go dead overnight. The battery is only six months old. The alternator is 16 years old and rebuilt by a local friend that owns an auto electrical repair shop. I made the bone head decision to buy an alternator and throw it on the car thinking I had a bad diode.

Here's how it played out. I needed my car, so I jump-started it using my Furd. I drove the car for a half-hour. When I shut down the car, I did a quick voltage check and the car was losing voltage pretty quickly. I have jumped off my other vehicles and had the same issue happen. The jump start fried the diode in a couple of those alternators. So, I toss an alt at it. The new alt is charging just as good, if not better than the old alt, but the battery light is on. I break out the GM service manual spec'd for my car. I run through the charge system tests and PCM tests. Everything checks out. Take both alts to my friend. He ran them both on his test bench underload and both checked out perfectly. So, I reinstall my old alt. The alts have three wires, charge, field, and the battery light. Why would the battery light come on with the new alt(not reman) and not the old one? I've had two other experienced auto techs, with a background in electrical work, look at the alt and test it. All the tests come back fine. None of us can figure it out. Why would the battery light come on with the new alt and be fine with the old alt?

I have the battery drain problem sorted out. It seems the battery was drained and needed a good slow recharge. I did find the clock spring was putting a minor drain on the battery due to a short when the steering wheel was in a certain position. It was by accident that I found that. Had the battery been fully charged and holding steady around 12.5 volts, I doubt I would have seen the drain.

 

The car doesn't use the battery to run the car. The alternator picks up that duty. The battery is only really there to start the car and give the electrical system a little balance while driving.

So, if you can drive fine, and not start the car, you have a battery or battery cable problem.

The battery light us usually a low voltage warning. A bad battery could suck this down or you could have wiring issues. A lot of times, when changing the alternator, folks will damage the exciter wire, which creates voltage problems. You might put the old alternator back in to see if it's really an old vs. new thing or if the low voltage condition is a more recent global development.

Putting a voltmeter on the system while the engine is running might tell you some things.

 

I've done all those checks. There's no low voltage issue with either alt, only the battery and that went away after a good bench charge. I've owned the car since new. Other than the clock spring there are no electrical issues. Battery cables are clean and fine with no resistance. I've checked the grounding points in the car and they're fine. The old alt works fine and no light. Only the new unit gives the battery light. I went back to running the old alt after I had it tested and it passed. The voltage holds steady at the battery and alt when running,14.45v. There has to be something off with the new alt. It's the only thing causing the issue. I hate to scrap or rebuild a new alt for nothing. Maybe I'll make a DC generator out of it for charging my equipment batteries.

 

Do you have a voltage gauge on the dash or just the dummy light?

 

I only have the good and proper idiot light. Its a base model.

 

To figure out what's really going on, you'd need to figure out where the voltage is measured for the light, check it there, and then work your way back. It must be a voltage complication.

 

I did that. It was part of the alt tests laid out by my GM service manual for my car. All checks fine.

 

These may not be complete and would assume that all parts are wired correctly on the inside.

When you perform the alt tests, are any with the car running and the battery disconnected?


If the battery is hooked up, you are measuring the whole system at one point. The alternator is not isolated and the gauge light takes its measurement at a different point in the system, where the voltage could be a different value.

 

I test voltage at the alt and the battery. 14.45 at the battery and 14.67 at the alt. If I disconnect the battery, the car dies. All of the Cavaliers and other OBDII cars I have worked on suffer this condition. The alt tests good on the car and on the bench, even the connection for the battery light is checking out as good. The only thing my way of testing tells me is the voltage drop across the charging wire to the battery and that the battery is getting a charge. I'm wondering if there is something in the alt that turns the light off if the voltage is in range.

 

I wasn't sure if the car was old enough for this to work.


Check for codes and you might look for the Charging System Description Section in your Factory Service Manual. That Description, and possibly some of the descriptions for specific PCM error codes may give you ideas of other things to test and places to probe beyond what the basic procedures in the manual instruct. You may also look for schematics that indicate how the dummy light is hooked up and find the voltage meter that feeds that light input. (Then probe there.)


Another thought - When you have the dummy light on, have you checked voltage at your cigar lighter to see what your accessories are getting outside of the engine bay?

Other than that, you have a condition that doesn't make any sense. To understand it, you'd need to research the specific function of your alternator (and its wires more) or start and the dummy light and trace the wires back to the problem it's measuring and then to the source. Alternatorman on youtube has really good videos on most common alternators.

 

Thank you wssix99 for trying to help. I appreciate your efforts. We've determined the problem lies within the new alt voltage regulator. There are no codes in the PCM. I haven't checked for voltage at the accessory port/ cig lighter. The old alt works and is in the car. I was hoping to fix this new one and keep it for a spare. I have an upcoming project that I can use the alt on. I will have to remove the internal voltage regulator and wire the alt for and external regulator. This will eliminate the troubled internal voltage regulator and allow me to use the alt.

 

I was thinking that you might have a bad regulator which is throwing a bad wave form, but with a good average voltage that shows OK on your voltmeter. The dummy light circuitry may be fast enough to pick up the spikes and dips in the voltage where some voltmeters may not be. (I expect that there are some descriptions and passages in the FSM that would clue into this.)

A oscilloscope will indicate this, for sure, but not too many folks have one laying around! Some scan tools have this function, but they are just as expensive.


Check out the Alternatorman videos on Youtube. He shows how to replace the internal regulators and I think he also runs a store that sells them.

 

Thanks. No oscilloscope, unfortunately. I had one back in the '90s that was owned by the owner of our shop. Neat scopes. I'll check out the videos. Replacing the VR isn't a problem. I've done quite a few. I needed to get that pinned down before ripping the alt or car apart.