Alternator bad? Voltage regulator maybe?

This alternator has an IR. Problem probably is the alternator has lost a diode. Not worth tinkering with if you haven't rebuilt a few. You can use a much newer alternator if willing to change the wiring plug. Thinking 1994 was the year the plug changed.

-- Message posted using

information at

"johninKY" wrote in :

Sorry, but my primary career has been in computers. IR, to me, means InfraRed, as in the light spectrum used by communication transmitters/receivers. Maybe IR means Internal Regulator?

I just looked again at the alternator in my car and see nothing that is accessible from the outside to let me swap out a regulator module. I couldn't even find a tag that listed the amp rating. Any idea how many amps the replacement alternator should supply? I think it is 70 or maybe 80 amps. This car does have air conditioning (the compressor is still in place but the system sprung a leak after a recharge and hasn't worked in over 10 years - I'd like to remove the compressor to remove any drag, if there was any since it is never turned on, but don't see how to put a pulley there for the belt). Of the few places that mentioned a rating, I usually see 70A units although 50A and 90A were also mentioned.

Never rebuilt an alternator. My guess, based on the assumption that bad luck is likely, is that I would open it up and find I had to dismantle everything to get at a module that is riveted behind everything else. I found a diagram of the alternator at

Yep,the voltage regulator is inside (on item #2). Looks like it might bepossible to open it up to replace just the "plate" holding theregulator; however, the prices that I've seen for the regulator aloneare higher than the entire rebuilt alternator. Splicing in a new plug wouldn't be a problem for me. I'm very handy with soldering and heatshrink tubing to make a solid inline splice. The problem would be getting the plug. I don't know that a new alternator would include a matching plug. The rebuild that I bought last time didn't include one. If I don't get one spec'ed for my car, I'm also guessing the mounting plates and holes won't line up.

After checking some other maintenance receipts to figure out what the mileage might've been at the time I replaced the alternator, the 2nd alternator now in the car (which was the rebuilt replacement from the salvage shop with 1-year warranty) only lasted maybe all of 20K miles which seems damn short. The 72-month battery is only 16 months old. Is it possible a weak or defective battery would result in the voltage readings that I saw (12.5V across the battery terminals with the car running)? It'd be easier to dump in a new battery than take off the fan belts to remove the alternator and have to do it in reverse again. Even if the battery were defective, I still expected the voltage to rise when I revved the engine. Apparently to have the alternator tested at the parts store means having to remove it from the car; i.e., they can't test it /in situ/.

I have the following choices (for cheapest prices found so far), all of which are rebuilds:

- ProStart brand: $118.99 (PepBoys; also at Advance Auto but I can't get a price online and too late to call right now).

- Autolite (Pro-Tech) brand: $129.99 (Advance Auto or Checker Auto Parts since both use PartsAmerica.com).

- O'Reilly Auto Parts (no brand identified other than "Ultima"): $119.

All have lifetime warranties. The Autolite unit includes a roadside assistance perk (with a $60 max reimbursement). They didn't mention the amperage rating so I'll have to check that tomorrow. Any experience in which brand or store you like best for an alternator?

Hi, Sounds like your battery is drained by some unknown load or it is bad. Probably one cell gone bad. If you can't test battery, try another one then yo can decide. When charging the voltage should be ~14V. Is the battery warm to touch?

Hi, I'd rather think I in this case stand for integrated or imbedded. If alternator is really proven bad, it can be rebuilt if you want to. Or hen you bring bad one it's worh a few bucks in exhcange.

The only difficult part opening up an alternator is removing the pulley. Only thing I found that always works is an impact gun. Once the pulley is off just remove those long bolts running through the case and everything falls apart. Just mentioning this if you decide to open it to inspect the insides. When a diode fails the battery "back drains" through the alternator which I think is your problem. Or at least part of the problem. There are companies out there that sell new connectors but the easy way to get the newer connector is to just cut one from a newer wrecking yard car. You don't have a 90A alternator now. Use to know the amperage output but now guessing in the 75A range. I have never had good results with using rebuilds, or at least the kind sold at auto parts stores. Quality over the years has been very poor.

-- Message posted using

information at

May want to read this. Mitsubishi, not a Subaru, but essentially the same design.

-- Message posted using

information at

"johninKY" wrote in :

While I might attempt dismantling the alternator (although I don't have an impact wrench), the deterring factor now is the high price for the voltage regulator replacement part. It costs more than the rebuilt alternator, and the same price as a new alternator. No point in buying just the part when I can get the entire unit for the same or cheaper price. With a lifetime warranty (limited to the original car it was installed in) and the much reduced price of a rebuild and the ease of replacing the alternator (the belt and a couple bolts), I'll probably go with the rebuild.

"Tony Hwang" wrote in :

No warmth to the battery (when unused or with the car running).

Even if the battery was bad, wouldn't the voltage across it go up due to output from the alternator (when the car is running)?

Hi, No. because bad battery will load down the alternator like a partial short. It happens when one electrodes inside battery cell shorts out. Generally batteries cause more trouble than alternaotr/VG. Even brand new battery can fail. One factor of better battery is mechanical rigidity.

Took the battery to the car shop. The first handheld tester said Good (but a recharge was needed but I already knew that). That included a check for the short condition. Then they shoved in the big tester that first charges the battery for 30 minutes (ho hum, wander the store) and it tested okay. They also plug in the cold-cranking amp rating to check the battery can supply that load (although 85F outside and 70F in the store isn't really "cold"). Couldn't find my trickle charger at home so picked up a 3-stage one at a different store where it was cheaper and am charging the battery tonight. That's because this car shop says they can test the alternator while in-place but it requires the car can startup on its own while testing. They have an emergency jump kit to let me start and drive back home but they need the car to be able to start itself using the battery and alternator without being jumped. So tomorrow I find out more.

Ok, just my 2 cents.

Before killing the alternator, a battery with a dead cell can prevent the voltage from the alternator from going to normal charge levels when measured at the battery.

Clean the alternator connections, make sure that the Alt S fuse connections are clean, make sure that the connector that plugs into the alternator is clean. Make sure that the fusible link and main fuses are clean. Mind the ground connections and battery terminals as well.

Also, make sure that the v-belt is in good shape and that i has the proper tension. I have seen them slipping with no squeal.

Charge the battery and measure the voltage at the battery when cranking. It should be more than 9 volts. When the motor is running, a properly charged battery should see 14.2 to 14.8 volts in a properly regulated system (those that allow batteries to last for 7 years, ;) )

Good luck!

VanguardLH wrote:

Been busy replacing air conditioners, rebuilding windows, and cleaning the yard of debris and hauling it away. Just haven't gotten to working on the car yet this week beyond getting the battery charged. My buddy is coming over (maybe) this Saturday or Sunday to help with more debris (which reminds me that I have to get busy with the chainsaw) and getting rid of a removed water heater. Afterward, he'll accompany me when testing the alternator just in case I need a ride back or a battery jump (although the store says they have an emergency boost kit). While some car shops say the alternator must be removed, this one says they can test with the alternator in place but the battery must be in the car and capable of starting the car itself (no boost, no jump). So, for now, it looks like this job waits until this next weekend.

Hmm, come to think of it, I should put the charged battery back into the car and connect it up. If there is a leak in the wiring or other component (or even through the alternator) then I might see this by next weekend (although the battery might still be good enough to start the car so I'm not sure this test will reveal anything).

Before you reconnect the battery, put an ammeter in series with the positive cable. You will see any current draw. If you get a current draw pull the plug on the alternator and see if goes away.

Good idea. I've got an old ammeter (Amp Probe) that I inherited from dad.

Did as you mentioned. The old ammeter didn't budge when I put it across the "+" on the battery to the positive cable (with the negative cable already connected to the "-" battery post), even on the 6A range. That seemed weird since I know the radio is still hooked up and has a clock in it. So I got out my DVM which has a 2A scale and it showed 0.438A, or a little under half an amp. Odd part was that it showed -0.438A (with red probe to cable and black probe on "+" battery post), as though current was running to the battery instead of out. In any case, there is a little bit less than an ampere showing on the DVM from the battery.

I left the battery disconnected until it's time to take in the car to have the alternator tested. My buddy isn't showing up today so maybe tomorrow.

Hi, Simple way to see excess ghost current is using a small 12V bulb in seiries with battery cable. There is a very very small amount of current flowing at all times for clock and such.

The DVM reading (-) is correct. The battery being the source would read drain current as positive if the red lead was connected to the battery + post.

As far as the magnitude goes, did you read with all the accesories off? It strikes me as a tad high, was the dome light on? something else on?

Good luck.

VanguardLH wrote:

Yeah, I figured that after thinking about it. The current would go out from the "+" battery post so the red probe on the DVM on there would've been backwards to the electron flow.

The only device that I know of that would've been on was the clock in the radio. No lights, including dome lights, were on. The key was not in the ignition switch so it wasn't even partially turned to enable some of those circuits.

I figured if a diode in the rectifier were shorted then that might be the cause of the drain through it. I found a basic schematic at:

If a diode were shorted, wouldn't that mean that there would be current flowing through it (back to ground)? I always thought a diode would blow open due to the heat of the leakage current but then these are high-current diodes and the leakage current is only half an ampere.

The reason the amp draw may be a little high is because of the keep alive memory in the radio and computer. Here's how to get around that. Hook up your amp probe as before between the +post and +cable. Now take a jumper wire and place it between the +post and +cable. Leave it there for at least

15 seconds (some take up to a minute). At this point your amp probe will read zero. Once you remove the jumper wire the probe should read much lower than in your initial test.

All that is moot though as I agree with JohninKY's original post, you have a bad diode since it's only charging 12.5 volts. BTW... that shop that says they can't jump the car, then test the alt is suspect. It sounds like they have an idiot proof tester of some sort and only have idiots to test with. ;-)

NAPA sells a very good reman alt with a lifetime warranty. Be sure to ask for their premium line as they also have a cheaper line with a 3 year warranty. They also sell a new one for many applications for not too much more.