Suddenly, Alternator Problems

"TwoFortified" (twofortified) 07/10/2014 at 01:25 • Filed to: RB20DET, Alternator, Charge System, parasitic, battery drain

0

18

I could use some advice regarding an alternator that won't charge. (I've got a TL;DR down there if you don't feel like reading much)

Prologue:

My battery gave up on me in April. I bought a brand new one. Shortly thereafter, I broke my main belt, boiled the bajeezus out of my coolant system, blew 2 hoses and a core plug, and proceeded to spend the next 2 months dropping the engine, replacing all of the coolant lines and vacuum lines, and putting the engine back in.

Independence from the shackles of a properly functioning charge system:

Last weekend was a 3 day weekend for me. I got home from work Thursday night, parked my car, and proceeded to party hardy for the weekend. Monday morning, my battery was dead. I didn't leave my headlights on or anything like that. The only culprit that I could summon was a !!!error: Indecipherable SUB-paragraph formatting!!! plugged into an always-on cig lighter. So, I stole my wife's car and went to work.

!!! UNKNOWN CONTENT TYPE !!!

Monday evening, I called AAA, and they sent out a battery tech. He gave me a jump, and proceeded to test my battery and whatnot. Battery is doing good, but there were inconsistencies with the charge system. He also mentioned how ludicrous my Belkin theory was. He recommended I pop over to AutoZone to get a charge system test done.

The lovely folks at AutoZone informed me that my alternator was dead (diode is good, but the voltage regulator was toast). So, I capped off my long Monday with an alternator removal.

Tuesday, I snatched my wife's car again, and brought the alternator with me. I cut work early and headed to an alternator refurbisher guy (My engine is an RB20DET. As far as I can tell, you can't buy a direct swap alternator here in the states, so I figured I'd get it rebuilt). The guy quick-like tested my alternator (on one of those neat belt-driver machines), and told me that I had an immaculately functioning alternator.

So, back home, I put the alternator back into my car. It's at this point that I found that I had missed one of the wires that went to the alternator plug. I had cut it when I dropped the engine, and forgot to hook it up. Lo and behold, I had also missed a ground strap (god I'm a noob). So, I soldered up the wire that I had cut, reattached the ground strap, hooked all of my belts and whatnot back up, fired her up, and read 13.5-14V at the battery. Huzzah!

There's got to be a morning after:

This morning (Wednesday), my car was dead again. This evening, I busted out the jumper cables and went to town. Here's what I found.

TL;DR My Findings:

1) My car is not charging. With the jumpers hooked up and both cars running, I have 13ish Volts at the battery. As soon as I take one of the jumpers off, the voltage hits 12, and immediately starts plummeting. Starved for electricity, my car dies after about 1 minute.

2) My alternator is wired up correctly. You'll have to trust me on this.

3) My alternator is grounded to the block. Continuity between the housing and the block is good.

4) My alternator is grounded to the chassis. Continuity

5) The chassis is grounded to the battery negative terminal. Continuity

6) My alternator 75 Amp fuse is not blown

7) The alternator pulley spins smoothly and easily. I had to pull a few bolts to get the belt back off, but I did, and the turning action seems fine.

8) The alternator belt has tension, and turns the pulley. Played the most dangerous game, in which you try to gently poke the pulley while the car is running. Yep, it's turning.

9) The alternator, as a whole, is not hot after running for a few minutes. Somebody somewhere on the internet mentioned that some issues cause the alternator to get physically hot. Not the case here.

And I think that's it. I haven't tested the hot side of the connection, because, frankly, it's more difficult than testing the negative side, especially since the car won't run on its own accord for more than a minute.

My current ideas:

1) Test voltage at plug while running. Difficult AND dangerous? Sounds like a plan.

2) Test continuity between the plug and the fuse

3) If possible, test continuity between the fuse and the Positive terminal on the battery (It's a long trip, my battery is in the trunk)

4) Take the alternator back out, and have it tested again. Maybe I broke it while I was installing it?!?!??!?

Beyond that, I'm tapped. Anything you guys can think of?

BTW, I get that my parasitic battery drain issue and my alternator-not-charging issue may not be related. It would be quite the coincidence if they weren't though.

Thanks guys :)

offroadkarter > TwoFortified 07/10/2014 at 01:29

0

It sounds like the regulator on your alternator is fried. this happened to my friends car, I put a volt meter on the car running and it was only putting out 12.5V which is way low, its supposed to be 13.6-14.3v for a car running.


how is the power regulated, does it have a module on the alternator? is it controlled by the computer?

GeorgeyBoy > TwoFortified 07/10/2014 at 01:45

0

Did you check the voltage at the alternator output terminal and ground or did I miss that?

Did you properly load test the battery?

My bet, the alternator.

They always seem to go in pairs. I watched my friend refuse to replace a bad battery. He finally did, shortly after his alternator failed. After replacing that his new battery failed because of that failing alternator. It was like a never ending cycle.

orcim > GeorgeyBoy 07/10/2014 at 03:15

0

I just went through the same thing on an old Camry.

orcim > TwoFortified 07/10/2014 at 03:21

0

There is a time when troubleshooting to the component level and using absolutely reasonable and justified logic just isn't worth it anymore. The system complexity can exceed "reasonable logic" in some cases.

If you are completely convinced that the alternator and the battery are good, then you have another wild animal running around that wire harness beyond what you've checked. If you aren't 100% on those components, I'd just start replacing them. What is it, maybe $300 to cover your ride with new battery and alternator compared with the time spent troubleshooting? Your choice, but unless I'm working with absolutely known good components, I take the FRU approach since that's exactly how these things were designed to be maintained in the first place (and one reason the dealer is so expensive.. )

deekster_caddy > TwoFortified 07/10/2014 at 07:15

0

So the alternator tested bad at autozone and good at the rebuilder's bench? sounds like something isn't hooked up right or there is a bad voltage regulator inside.

You already mentioned my other go to points - ground straps etc. make sure you take the ground cables off and rough up the surface where they connect. Also, quadruple check the wiring - especially the broken wire. Are any other wires brittle? If you jiggle the harness while the engine is running is there any change in output? Easy test - turn on headlights - when they get brighter, output just increased.

Charge the battery for a few minutes from jumper cables if you don't have a real battery charger. You need it to run long enough for some testing/tinkering time.

TwoFortified > GeorgeyBoy 07/10/2014 at 15:12

0

I checked the ground by testing between the body of the alternator and the frame, and the frame to the negative terminal of the battery.

I have not yet checked for continuity on the hot side of things. I'm gonna have to do that next.

TwoFortified > deekster_caddy 07/10/2014 at 15:13

0

The alternator tested "bad" while it was in the car. They ran this test by hooking leads to the battery terminals while the car was running.

When the alternator was benched, it tested good. This leads me to think it's a wiring issue.

TwoFortified > offroadkarter 07/10/2014 at 15:14

0

The alternator has an internal voltage regulator. I mean, I just had the thing benched on Monday, and it tested good. I guess there's a chance it got fried when I re-installed it, but damn.

TwoFortified > orcim 07/10/2014 at 15:18

0

Yeah, sure, fine, but what do I replace? The battery is two months old, and passed a load-unload test on Monday. The alternator failed the in-car test on Monday (with the leads hooked to the battery terminals), but passed a bench test on Tuesday.

At this point, I can pull the components pretty quickly, and get them benched again (battery and alternator), so I think that's where I'm at. I've been known to blindly replace components before, and can empathize with your point, but if I get the battery and the alternator bench tested, and they both pass, doesn't it seem counter productive to replace them anyway?

deekster_caddy > TwoFortified 07/10/2014 at 15:32

0

I'd tend to agree, wiring issue. You should be able to 'simulate' the correct connections on the voltage regulator - I forget which wires go to which (and you don't want to just willy nilly connect a few wires) but all you should need to make the alternator work is a good ground, and a positive jumper from the main positive to the positive input of the voltage regulator. All of the other wires are for things like key on/off or battery thermistors and such. It should work with just a basic positive, that's how they bench test it. If you know what the other wires are for (got a schematic or service manual?) you should be able to troubleshoot further.

offroadkarter > TwoFortified 07/10/2014 at 17:59

0

I'm not sure if the bench can act as a ECU and tell it how much voltage to put out. They might just check to see if the alternator puts anything out. I've never used one of them though so I'm not certain.

TwoFortified > deekster_caddy 07/10/2014 at 18:01

0

The one good thing that has come of this is that I've realized that my Alternator is about as simple as they come. It has a big ground wire, and a plug with two wires. A ~12 gauge wire that is presumably the hot wire, and a ~18 gauge wire that, I'm told, is a signal wire to the ECU/Gauge cluster, which will kick off battery idiot lights if it reads bad.

The signal wire was the one that was cut, and my car ran great for about a month...but now it has a sad.

Anyway, you're right, it should be simple. I'm gonna have to trace the hot side of the circuit, and make sure that's good to go. I'm fairly confident that it won't be...but if it is, the only thing left is to make sure I have immaculate grounds.

TwoFortified > offroadkarter 07/10/2014 at 18:11

0

Oh yeah, the bench thing was cool. It was a big machine with a belt on it, and you basically run a bolt through the swiveling bolt hole on the alternator, stick the belt on the alternator, tension it down, plug into the back, ground it, and then it runs the belt up to like 3K RPM, and reads the output on the alternator.

Basically, it simulates what your engine would be doing.

ECUs do have a hand in the alternator circuit, but it's pretty minimal, thanks in no small part to the alternator being an old design that's pretty analog.

offroadkarter > TwoFortified 07/10/2014 at 18:13

0

gonna have to bust out the voltmeter then and see where you have a break in a wire

orcim > TwoFortified 07/10/2014 at 21:51

0

Umm.. yes, and no. The piece that's missing in your response is the multiplication factor in complex systems which would defy common logic. An brief story:

Had a demo at EDS (GM Land) in Michigan in the mid 80's. We were gonna demonstrate our engineering workstations (sun microsystems.) Did the demo, everything fine, and went to lunch with an exec VP (Thorsen/on) at the time. He told a story of their computers in cars, how they'd replace them in warranty, but only if they were proven bad. The mechanics couldn't "prove" the units were bad, but they knew they were bad because when they replaced the unit, the car would run, so they'd run the units over, pound a screwdriver through them, saw them in half, etc. and send them back and say "See, this one is bad. Compensate us." The tests would all have returned them fine.

There are three issues here. Battery that's not holding a charge. Take it to some other place for testing thought it's probably good. Alternator that's not working - get a second opinion also - maybe it's just intermittent. And third, is the system these components are trying to operate in... (and I hate wiring harness issues.)

Example: My car started having intermittent trouble starting. Thought battery, thought alternator, but both bench tested fine. Kept getting worse, and finally decided on a circuit path teardown. After testing the resistance and connectivity from starter out, finally got to the cables leading to the battery and found that the big 3 fuse box was completely corroded inside, leaking voltage all through the three main circuits and not allowing enough amps to flow for starting. It was like a big plastic box filled with copper foam. The alternator and battery were fine, of course - the environment they were operating in was poison.

TwoFortified > orcim 07/11/2014 at 00:40

0

Interesting story.

Yeah, so, today, I tested continuity on the positive side of the circuit, and it's good (plug to fuse, fuse to +), so I, once again, decided the alternator was the problem. Now I've got two benches that say my alternator is pristine.

My last hope is the "Charge" wire...the "Exciter". Maybe the fuse is blown on that thing. It's late, and I'm tired of working on it, but tomorrow I check that. If that's not it, then I'm in deep.

Quattro-luvr, Powered by Datsun & Stinger > TwoFortified 12/12/2014 at 14:29

0

TL:DR

Legs is the answer.

youshiftem > TwoFortified 12/12/2014 at 14:35

0

Sounds like your alternator is not being "turned on," if showing that bottom picture to it doesn't work, try testing the "excitation" circuit on the car. It's the only thing the bench test would have that the in car test wouldn't.