Bryan Welch

My 1991 944S2 has been eating batteries. I add a new Interstate battery, it runs for either a day or few days, then it's dead. It works again after charging it, but after one or two charges the battery is dead completely. I'm on battery number four now. After tests at Interstate and my mechanic, I'm told the charging system seems fine, but most systems don't test everything. I'm wondering if the battery is being overcharged by a defective rectifier.

Please help me figure out what to test and order. I have Iceshark's full boat headlight upgrade and think I have an upgraded alternator. Because of this, I don't know what alternator or rectifier to order as a replacement. My mechanic gave up and there are no other Porsche mechanics in my area.

thanks,
--Bryan

944Ross

If all is not well with the rectifiers, you'll see an AC waveform at the alternator output. Set a digital VOM to AC and stick it on there, or at the battery. If you see more than millivolts AC, there's a problem. But I don't know that that would kill a battery on its own.

It sounds like your battery may be discharging thru the field coils, draining it. At any rate, removing and disassembling your alternator will tell you exactly what you've got and whether it's bad.

ZR8ED

+ on removal and inspection of the alt. Batteries aren't cheap!

Have you tested for excessive battery drain? Power cables/dead short somewhere? Do you/did you have any other electrical problems? Additional electrical devices ie amps, extra lights etc.? How about main battery cables? are they good? are you sure? they may be internally corroded, and adding a lot of resistance Cheap and easy to replace.
Just trying to throw out some ideas.

JohnKoaWood

My bet is a slow drain from the battery...

Disconnect the terminals, and measure resistance accross them while fuses 11 and 36 pulled (these are always powered, pulling the fuses should remove these paths..)

Do you have any aftermarket bits installed, that might be wired directly to the battery (had a local car with an installed NARROW band O2 gauge, the gauge was wired directly to the battery for power, and this SMALL load was sufficient to kill the battery in two days.. removed the gauge, and all has been right since!)

You dont want ANYTHING wired directly to the battery other than the OEM harnesses, everything else should be wired via switched power... check for items that sould be off (dome lights, glove box light, etc..) but are on all the time..

What does the battery gauge show when the car is running?

FWIW, it takes one heck of an AC waveform to kill a wet cell lead acid battery, while that could be the issue, I am more inclined to believe it is a slow drain killing it over the course of a couple days!

Bryan Welch

The only additions to the car are a new stereo and Iceshark's headlight upgrade package. I'll pull the fuse for the stereo and get some measurements as soon as it stops raining.

Is it normal for something draining a new battery to ruin the battery that quickly?

For the last couple Interstate batteries, I would get no more than a few hours of driving time before the battery was flat dead. I'd install the battery, drive 30 minutes home, and park it overnight. Next morning or next week, it might be fine for another short drive, or it might be completely dead. If dead, I'd charge it for a few hours and it would be fine, drive again, park. It may sit for a week and be fine, or it may be flat dead after 30 minutes. Once I charge it (with my Die Hard charger that's worked well on other vehicles), I can drive the car again and it's likely it won't start again. No charging works after that one charge and Interstate tells me the battery is defective after that. That's happened with two Autozone and two Interstate batteries so far.

The charging needle was fine and when I measured with a voltmeter or with a tester at a parts store, they said things were fine. Voltages all seem proper which it what has led me to think of overcharging killing the battery.

I'll get some voltmeter measurements and post again. If that doesn't make something obvious, I'll pull the alternator and see about testing it separately.

944Ross

If your battery went flat, to where it wouldn't turn the engine, it would take almost a full day to charge it with the typical home battery charger, maybe longer (at say 5 - 10 amps). If it isn't getting charged by the alternator, and you only charged it a couple of hours before driving it again, you're taking it all the way down, which is real bad for batteries. I don't doubt you could get it to turn the engine over and run after a few hours charging.

If your new stereo has a separate amp, make sure it isn't powered up all the time.

Bryan Welch

Test results:

I started with a new battery today and did some testing. Measurements were with an old Micronta (Radio Shack) auto-ranging meter.

Pulling fuse 11 and 36, I tested resistence across the battery terminals with no battery installed. Resistance was 30 Ohms. I left those fuses pulled for the following tests.

Testing the new battery, it started at 13.07 volts and dropped to 12.99 volts when installed in the car.

Testing parasitic drain, I disconnected the positive terminal and applied the meter between the battery and positive terminal. With the car locked or unlocked, the drain was 0.1mA.

Testing the battery again after applying the terminals, it now read 12.85 volts. I started the engine and the voltage went up to 14.08 volts.

I then measured AC voltage across the connected battery terminals with the engine running. Watching the meter, it between these values: 0 volts, 3 volts, 10 volts, 18 volts, 0 volts, 3 volts, etc. It took about a second for each AC voltage to change to the next value.

Turning the engine off, I checked battery DC voltage again. It started at 13.35 volts, but a minute later it was down to 13.10 volts.

At that point, it started to rain, so no more testing today.

So, does this rule out or indicate anything? What more tests would make sense?

JohnKoaWood

Indicates a drain on the battery when off...

disconnect the terminals at the rear of the alternator to isolate it... other potential short would be the starter, but the alternator is easier to get to..

disconnect the positive battery terminal, and measure resistance between the positive cable to ground.. disconnect the alternator / starter / fuses till there is infinite resistance showing..

IF you cant isolate the short, it is a harness issue!

944Ross

LOL, I have the exact same Micronta DVM! Are you sure the resistance across the battery terminals w/no battery was 30 ohms, and not 30 KOhms? Big difference. 30 ohms means 12/30 = 0.4 amp drain, which is significant. 30 KOhms, insignificant.

The AC voltages you saw could indicate failing diodes, which could allow the alternator coils to drain the battery. I'd repeat by measuring AC voltage across the alternator.

Bryan Welch

I did more testing today. Measured resistance across the battery terminals with no battery installed and first got around 330 K Ohms. I did experiment with the meter and am sure of these readings, but the previous reading of 30 Ohms may be a mistake.

Battery disconnected read 12.76 volts. Connected battery, started engine, ran for 30 seconds, turned off engine. Did this three times. When running, the voltage was at 14 volts. After turned off, voltage was 13.1 volts.

After each engine run, I disconnected the battery and measured resistance. After the first, second, and last run, the resistance was infinite.

Does one disconnect the alternator from underneath? I'm not seeing them at all from the top.

A few years ago, the started stopped working and my mechanic managed to rebuilt the starter somehow so it worked again, though it sounded a bit different. I'm wondering if that might be an issue. Then again, this alternator has run for 205,000 miles. It's time might be up, too. In a way, I'm tempted to replace both to rule them out. For over a year now, I can't drive the car anywhere without the chance the battery will be dead when I try to start it again.

944Ross

It sounds like your alternator is working, but it's clearly on borrowed time. (are you sure it's original?)

Doesn't seem to be any parasitic load, either. Just to check, next time you're measuring resistance with the battery disconnected, open a door and see what it says (assuming your dome lights are set to come on).

Battery voltage right after being charged is not a valid indicator, it will have a "surface charge" that settles down in about 15 minutes or more. Should be closer to 12.6.

Bryan Welch

How hard is job is replacement of the alternator and starter? Any tricks? It looks like I could do the alternator mostly from the top, so that sounds good. I haven't replaced a starter since I worked on a 1977 Honda.

944Ross

Both are pretty easy, the starter especially. Just have to get under there, although I can't say specifically for an S2

Bryan Welch

Did a bit more testing this afternoon. With fuses 11 and 36 pulled, doors closed, no battery, resistance between battery terminals was 810K Ohms. Add fuse 11 and it drops to 530K Ohms. Battery was 12.82 V after sitting an hour after the three starts.

Today doesn't seem to have told me much. Time to review some guides on replacing the alternator and starter.

944Ross

Was the alternator connected when you did that? I think the high resistance (but not infinite) may be the resistance of the diodes in the rectifier bridge.