I replaced my '86 battery about two months ago. Last week, it started showing symptoms of the previous battery (hard starting, followed by starter chatter and the need to charge the battery).

I checked the forum and realized I should get the alternator checked. NAPA checked it and said it needed replacement. Another local auto parts shop checked the same alternator and said it was fine (putting out 14.7 volts).

I swapped the tired 2-month-old battery today at NAPA (no cost; under warranty) and completed an initial round of checks based on previous forum advice (good stuff from 86Pacer, Agent86, CFI-EFI, jfb and others). This includes cleaning the ground strap mounting points, cleaning alternator bracket, checking continuity of cables, etc.. I charged the battery on 2 amp charger before I started the car. Engine fired right up. The numbers:

With the engine off:
volts across battery = 13.19
volts at alternator connection = 13.15
resistance through cable from alternator to positive terminal = 0.000

With the engine running (cold):
volts across battery = 14.61

With engine running (after 20 mile warmup with headlights on):
volts across battery = 13.89

After shutdown, I measured the volts across the battery again: 12.97. Not bad, but lower than it was right after charging.

Knowing that I have a fully charged battery, I'm going to pull the alternator and take it in again for a bench test. I want to check the alternator now before I drag the battery down.

The big questions: What tests should the shop run? How do I know they're checking all the correct elements (such as voltage regulator)?

 

Most shops should have a vat40. They can load the charging circuit with the machine and see if the alternator can put out the propper amperage. It's a lot bettery way than slowly spinning the alternator at the parts stores.

 

Amperage test: got it.

Is that enough to tell me if the voltage regulator is crapped out?

 

Your #'s all sound good. I don't know if parts stores can test the regulator...?. Usually they only bench test alternators to check output under a load. Advance Auto has some cool tester they can wheel out to your car to test your charging system. It's a free test, you could ask them what it checks.

 

Make sure they check the diodes. Mine was showing good, but had a bad diode. So, instead of just rebuilding or replacing with the same alternator, I went with the higher amperage unit. He said it's the one used for those running big stereo amplifiers. Now it always starts great and displays plenty of voltage.

 

How often do you drive the car?

If it onlys sees the road occasionally, its time to buy a battery tender.

 

This is the best way to test the charging system. The VAT40 will test alternator output under load and it will also do a battery load test to check the battery performance.

Most chain parts stores do not have this piece of equipment. The bench testers you find at some parts stores typically have a fail/fair/good meter and that is not a good indication of alternator performance especially when there is no load applied to it.

An alternator is desigend to produce enough amperage to keep the battery charged when a load is applied (such as the various systems running when the motor is running). A good alternator will not generate extra amperage overand above the load requirements. Too much voltage and you can easily cook a battery or components like computers.

 

I'm lucky enough to live in San Diego; I drive the Vette a few times a week...even in January.

Alternator is in the local repair shop for a bench load test; they also rebuild if necessary. Follow-on plan includes an in-car check of the charging system.

Thanks for the feedback! I'll update when I have results.

 

To test an alternator you need to be able to put a maximum load on the alternator (around 100 amps, requiring a high power 0.12 ohm resistor) while measuring the output voltage. You need to look at the output waveform with an oscilloscope to determine if you have one or more open diodes. An oscilloscope will also tell you if the voltage regulator is oscillating or not. No parts store alternator testor does this, they only measure output voltage and current.
If you want your own tester, I can tell you where to buy a surplus resistor cheap! Scopes aren't cheap though and you can't borrow mine unless you live close to me.

P.S. You don't need to remove the alternator to test it, you can load the alternator and look at the waveform in the car.

 

Local shop ran a 3 stage load test with alternator in car; I haven't seen specs but this is what I got over the phone:
The shop says everything looks healthy. I'll ask about the output waveform & voltage regulator health when I stop by this afternoon.

To refresh, I replaced a five year old Diehard in early December with NAPA 8475 Select. Symptoms were starter chatter and failure of battery to hold charge. Late last week, the battery started with the same symptoms, so I started looking at other causes. Starter was rebuilt 15 months ago and only balked when the battery was dying.

Of note, NAPA didn't blink twice when they handed me a fresh battery under warranty. Maybe they've seen a lot of them cross the counter!

Thanks to all for input. Beyond keeping a watchful eye on battery, any other ideas?

 

there are a few things you need to know.
1. the alternator is a constant potential charging device. That means there is a preset voltage limit in the alternator voltage regulator.
The alternator tries it's mightiest to achieve that at all times.
2.When you first start the car, the alternator wants to get to that voltage, and it kicks in more field voltage to accomplish this.
The field voltage is a function of the voltage regulator based upon it's constant battery voltage sampling.
3.so, cold, the battery has sat all night, it's achieved it's lowest unloaded voltage, and then just started the engine. The voltage wire on the plug on the alternator feeds the sample voltage to the regulator (CHECK THAT PLUG) then if the voltage is lower, the reduced battery voltage biases the voltage regulator, and the alternator starts out at 105 amps. As the battery voltage comes up, the field voltage goes down, ending up with the load and the battery setting the alternator load at any given second.
4. The alternator generates heat.
a. the heat of the alternator is respective of the load.
b. Load is based upon battery voltage
3. there is a thermister in the regulator that reduces the amount of voltage sampled by the alternator, reducing the voltage and the load.
4. reduced voltage and load reduces the battery voltage when the car is hot. The hotter the alternator, the lower the voltage to a point. The battery could overcharge at elevated temps and sulfate excessively, so this is a good thing.
5. The alternator doesn't produce much output below 550 RPM.
6. The alternators don't do well at extremely high temps.
7. Notice the voltage cold, and the voltage hot. it may vary by 2 volts or more.
8. overhauled alternators may be compromised by the overhaul methods.
9. there are aftermarket methods of reducing the alternator temp.

 

Measure the battery terminal voltage before you attempt to start the engine. Battery voltage tells you the state of charge in the battery. 12.0 volts and below, discharged, and 12.9 volts and above, fully charged and linear in between. During cranking, the battery terminal voltage should not fall below 9.0 volts or the battery is discharged, battery cables and posts need cleaning and replaced, or the battery is at the end of its life. Since you have trouble with 2 month old batteries, I believe you let your car sit for long periods and you do not use a battery tender to keep it up. You might also have leakage current higher than 50 milliamps which GM says should not be exceeded. My 87 has 27 ma. of leakage. A discharged battery cannot supply enough voltage to keep the starter solenoid closed during cranking and this is why your starter chatters. Pull the negative cable and place an ammeter from battery terminal to negative cable and when the courtesy lights time out, switch down your ammeter to small full scale current and read what the leakage current is.

 

Got it; I've been keeping notes on terminal voltage, but not during startup. A good chance to let my wife help me work on the car!

Car is driven 2 times a week for 9 mile commute; I'll keep my eye on voltages as mentioned above. I don't use a battery tender right now, and haven't needed it in the past three years. However, to keep from returning batteries, I'm considering it until I track down root causes.

Leakage check is next on the worklist.

Thanks to all for persistence in helping me track this down. I've got a short trip to Japan over the weekend, so no updates until next week. Thanks again.

 

To recap: battery & alternator were suspect for about two months. Checks on alternator appear normal (see above) and battery has been replaced for second time in two months, so installed battery appears healthy also. I started checking for current leaks following recommendations.

I disconnected underhood lights and set up to check battery leakage:

1. With all fuses in place, after lights time out, steady draw of about 8mA.
2. Removed CTSY/CLK fuse to make checks go faster and discovered something interesting. When all other fuses are in, current first shows 170 mA for about a second, then settles down to 8 mA.
3. With CTSY/CLK fuse still removed, I pulled all fuses one at a time. The 170mA/8mA pattern repeated for most fuses. If I removed the leads for a few seconds, the 170mA/8mA pattern would repeat. However, if I held the leads in place, the current remained at 8mA for well over a minute.
4. The two fuses that showed a different result: the LCD and CLSTR fuses. When I removed each of those one at a time, there was no 170mA "spike" and the steady draw remained at 8mA.

I've been keeping the battery on a 2A charger overnight for the past two evenings.

Any ideas on the current draw (in particular the momentary 170mA spike)?

 

It's been a while since I did those current draw tests, but I believe your spike in draw is coming from the delay timer for the interior lights, even if the lights are off.

Also, it's not uncommon to get a new battery off the shelf that's bad, or even a bad batch. Most places will load test them first just for that reason.

 

I would think the ECU has something to do with the 170 mA spike. It senses the change in voltage and probably resets its configuration.

 

Thanks. It is a temporary load, which quickly drops to 8 mA. Per earlier posts, 8 mA isn't likely to cause significant drain on the battery over a few days.

Still looking for reasons why a relatively new battery (NAPA Select 8475) croaked after two months. Having replaced it, successfully load tested alternator at the shop, and eliminated significant battery current leakage, I'm wondering if it might have been as simple as a bum battery (the previous battery was a five year old Diehard Gold).

In the meantime, I'm keeping it on a 2 amp trickle charge at night.

 

the draw comes from the circuit trying to make a connection as you pull out the fuse

it is basically arcing across to the fuse for the 1/10 of a mm where it is still close to the contacts when you are pulling it out

thus the sudden jump

as it jumps across