Battery, Alternator or Something Else?

sm625

Diamond Member
May 6, 2011
8,172
137
106
99 Chevy Cavalier. It has a 3 year old battery and a 9 month old alternator. Usually it starts and runs fine.

Since the car has no battery gauge, I placed a multimeter across the battery terminals. The battery voltage reads about 12.0V when off, 12.6V when idling, and about 13.5V when revving. All of this varies a bit depending on electrical load. Sounds about normal right?

At random times as soon as you start it, the meter reading rapidly drops to under 10V. The engine will stall if its not shut off. Revving the engine does not help. However, as soon as you shut it off, the meter returns to 12V. The car will usually restart without requiring a jump, if not left running in this "shorted out" state for too long. Then the next time you start it, everything will be normal. Then you shut it off and restart it and it might go below 10V. It seems completely random, although heat does seem to be a factor. The longer you allow the car to cool, the more likely it is that it will start normally.

Here is one other interesting piece of information. IF you start the car and the meter drops below 10V, all it takes is a quick jump to bring it up to 13V. I mean literally tap the jumper cables onto the battery terminals for one second and the meter jumps from <10V to 13V and everything seems to be normal.

The auto parts store checked the battery and said it is dead, it is at "65%", whatever that means. I've seen plenty of dead or dying batteries but I aint never seen one behave like this.

For it to drop below 10V so incredibly fast, it takes a 50+ amp load to do that. Maybe even >100 amps. I've driven for over 45 minutes on battery power alone; it takes a long time for a normal daytime load to drain a battery to the point where the car dies. But when its in this condition the car dies within 20 seconds. It sounds to me like a serious electrical problem, and I dont want to just start replacing everything if it is going to be some intermittent short somewhere that I wont discover until after I've spent $200 on parts I didnt really need to replace. Is a 65% battery really that dead, or does it still have a couple years left in it?
 
Last edited:

MixMasterTang

Diamond Member
Jul 23, 2001
3,167
176
106
If you are taking the reads directly from the battery, then I would guess it's a battery issue. Even without a working alternator the battery should be able to keep the car running for a minute or two at minimum.
 

PottedMeat

Lifer
Apr 17, 2002
12,363
475
126
intermittent connection between the plates in the battery?

what other high current paths are there from the battery out? maybe the cable to the starter is damaged and coming in contact with ground.
 

phucheneh

Diamond Member
Jun 30, 2012
7,306
5
0
Check the output at the alternator post while running. Note DC voltage, then switch and check for presence of AC voltage.
 

JCH13

Diamond Member
Sep 14, 2010
4,981
66
91
Is some high-load device turning on when it's dropping to 10V?

Sounds like a dogshit battery.

FWIW the auto parts store probably means the batter is at 65% capacity. 80% capacity is usually considered "needs replacing".

slalifecycle.png


Basically 80% capacity is half way down to "totally dead" and is usually when batteries start to degrade rapidly. They might have a float voltage that's just fine, but discharge VERY quickly. Also, FWIW, a good lead-acid float voltage (fully charged, no load) is 13.1-13.6V. If you're floating at 12.6V that's not a great battery.
 

phucheneh

Diamond Member
Jun 30, 2012
7,306
5
0
The one time someone actually has a meter, they still choose the 'don't listen; replace a part' comedy option. :(

Unless your battery is causing wonky charging output (possible, not probable), you have an alternator or cabling issue. Put one probe on alternator output (big cable on post), one on the alternator case. Be careful that the metal part of one probe does not touch both the post and the case, or you'll get a wee bit of sparking and maybe some slight welding action.

Report back with voltage numbers.

If you have a 600cca battery (random guess), 65% would be 395cca, which is pretty weak. But I would charge and retest before replacing it. People will tell you that conductance testers work fine on partially discharged batteries, but it's only true in a really vague sense. The cranking amp number WILL go up on a charged battery.
 

sm625

Diamond Member
May 6, 2011
8,172
137
106
Dammit I knew it seemed like more than just a battery issue. Replaced the battery and it still does the same thing. It runs fine then out of the blue the voltage drops liek there's a huge load on the battery, ie a short somewhere. The alternator is under warranty so all I will lose on that is my time.

Put one probe on alternator output (big cable on post), one on the alternator case. Be careful that the metal part of one probe does not touch both the post and the case, or you'll get a wee bit of sparking and maybe some slight welding action.

Report back with voltage numbers.

How is that different from measuring across the battery terminals? The output of the alternator is wired directly to the battery +, and the alternator case should be at ground.

I'm going to try disconnecting the alternator while the meter is reading 10V to see if it jumps back up to 12V. (My current best guess is that the alternator gets very hot and randomly decides to short out and sink current instead of producing current as it normally should). After that I start pulling fuses I guess.

Because its been hot and this failure coincides with a major heatwave, I wonder if it could have anything to do with the A/C. Is there a large electrical load that the A/C exerts somewhere, aside from the fan?
 
Last edited:

phucheneh

Diamond Member
Jun 30, 2012
7,306
5
0
The alternator is not mounted to the battery. There is wiring in between. You check at the alt to take that wiring out of the equation. More than likely, your car still has the 'old school' charging system setup of alternator -> starter -> battery (a lot of new cars go from the alt to fusebox). Plus your battery grounds and your engine-to-chassis ground. That's a decent amount of connections and cable, and any one of them could have a loose terminal, oxidation, or some other flaw inhibiting the flow of electricity.

In addition, AC voltage may be smoothed out when you measure across the battery terminals. To see if there is a leaking diode inside the alternator (regulator and rectifier are internal), you check for AC voltage at the alternator post.
 

LTC8K6

Lifer
Mar 10, 2004
28,520
1,575
126
Drive it to the auto parts store and have them check the whole starting and charging system.

Do not disconnect and reconnect the alternator/battery cables while the engine is running.
 

phucheneh

Diamond Member
Jun 30, 2012
7,306
5
0
So instead of doing a better job of checking it, as I am attempting to instruct him, you would recommend he get someone to do a worse job of checking it?

That box at the parts store is not magic. It will do a conductance test on the battery and then check charging voltage (at the battery). That's it.
 

LTC8K6

Lifer
Mar 10, 2004
28,520
1,575
126
He is free to follow whichever recommendation he chooses, or none at all, as you can see.
 

sm625

Diamond Member
May 6, 2011
8,172
137
106
I tested the alternator voltage at the output of the alternator to the alternator case as per phucheneh's instructions. The voltage read 12.16, exactly the same as it does at the battery terminals. When I switch to AC I get 26VAC. Strangely, I get the same 26VAC even after turning the car off. So I have to question whether the AC part of the multimeter is working. (It does read 120VAC when I stick the probes in an outlet.:confused: ) At any rate the alternator does not appear to be charging. Since it is under warranty, and easy enough to remove, I'm just going to take it off and get a new one.
 
Last edited:

LTC8K6

Lifer
Mar 10, 2004
28,520
1,575
126
Maybe you should have changed the warranty item first. :p

Anyway, you will now have a replacement battery and alternator which should last you a while.
 

phucheneh

Diamond Member
Jun 30, 2012
7,306
5
0
Yesterday you said it was between 12.6 and 13.5...?

If you're reading straight battery voltage at the alternator (as opposed to higher than battery voltage, but less than optimal charging voltage), I would check for voltage at the plastic connector. There should be two wires; one is the power for the field coil, one is the voltage sensing wire. If battery voltage is absent on either, the alternator will not work, and you will see battery voltage at its output (the post).

He is free to follow whichever recommendation he chooses, or none at all, as you can see.

Unfortunately, the recommendation of 'ask the guy at Autozone' is about as helpful as 'burn it for the insurance money.'

edit: and I got no clue what is up with your 26v AC reading. That's, uh, strange...what kind of meter?
 

sm625

Diamond Member
May 6, 2011
8,172
137
106
Yesterday you said it was between 12.6 and 13.5...?

That is in the "normal" state, when everything appears to be working. Sometimes it is normal, and the alternator is charging. When its not normal, the old battery quickly dropped to 10V while running. The new battery drops to 12.2V while running. But I can shut the car off, wait an hour and it will read 12.6V while idling (13.5V when revving). Then I'll shut the car off, start it right back up and it will be back to reading 12.2.
 

phucheneh

Diamond Member
Jun 30, 2012
7,306
5
0
If you don't care about warrantying out the alternator, I guess it doesn't hurt to try it. Sounds like it could be a regulator issue. It could also be an intermittent wiring issue in those two wires I mentioned, though. Particularly the 'sense' wire not getting an accurate battery voltage.
 

LTC8K6

Lifer
Mar 10, 2004
28,520
1,575
126
12.2/12.6V while idling normally means the alternator is on vacation. Should be in the low to mid 14V range.

Maybe you also have a wonky multimeter, though?
 
Last edited: