Fluctuating (low) voltage: what does it mean?

Status
Not open for further replies.
gathermewool

gathermewool

Site Donor 2025
Joined
Jan 9, 2009
Messages
10,248
Location
New England
97 Subaru Legacy GT with 206kmi (owned since 167kmi)

Reman’d alternator, but not sure when or by who. Looks pretty new (compared to the rest of the engine compartment) but that may not mean anything when it comes to quality or age.

About a year ago I bought a scangaugeII. I didn't always have it on Voltage until recently, but never noticed anything less than 13.8 V on the highway and 13.4 at idle, even when loaded.

I was flipping through the gauges and noticed, with the headlights, radio and heater on, that the voltage was reading around 12.6 at idle. After some monitoring, I noticed it would regularly go down to the mid-to-low 12's loaded at idle (low 13s unloaded) and low to mid 13s at highway RPMs. It would even go below 12V on occasion, though that was only for an instant before popping up.

After a quick look I noticed the insulation to the connector had crumbled away about six inches, leaving exposed, corroded-looking conductor. I had an electrician buddy of mine replace about two feet of wire that had brown, over-heated insulation with new stuff. The wiring past where it ran over the engine looked good all the way to the fuse box.

I noticed an immediate improvement, but it’s not where it used to be. I’m getting a fluctuating voltage on the highway from 13.4-13.9V regardless of load, and anywhere from 12.8-13.3V idling, varying with load. I’ve seen as low as 12.4 still, with the lights on and A/C on high.

It may be coincidence, but the problem seems to be worse when it rains.

Also, the RPMs at idle when this occur stay very steady at 650.

Is low voltage or varying voltage a definite sign of a failing alternator or could it be caused by a small ground?
 
Unless it is very very loaded, it would not be caused by a small ground. It could however be caused by a bad ground. Sometimes in older vehicles the chassis ground gets corroded and just needs to be cleaned up. What exactly do you mean by small ground? You dont have some hoopty coat hanger used as a gnd wire or something do you?
 
Maybe a loose belt, rain would make it slip more.

I don't think your voltages are that bad, a bit low, but I have seen worse on a failing alternator (and/or battery).

Just take it into your local Auto Parts store, most will do a free test of the charging system.
 
Have the alternator load tested and see how many amps it's pushing. It sounds like an alternator dying.
 
I was being generic when I said "small." I meant a short to ground that might be overloading the alternator to the point that it was showing a lower than usual voltage. All ground wires are stock, and have been cleaned up - no change after cleaning those and battery terminals and connectors.

I guess the next step is to have the alternator load tested. I'll updated with results when ever I get around to it. Thanks for the posts, guys.
 
sorry, i misunderstood. The chance that it is a wire shorting somewhere is pretty low, unless it is a main cable from the alternator or battery, because they are not fused and could potentially dump 100s of amps to ground, in which your alt could not supply that much current and you would see the corresponding voltage drop. What was the original burnt, brown, over heated battery cable burnt from. That maybe a place to start if you feel it was not done from an external source like actually touching the engine, manifold, or something else very hot. I think other guys are correct, go to any parts store and they will measure the system for you for nothing and let you know if all is well. I did have a case once similar to yours where i had a bad voltage regulator in the alt that was intermittent. I had to have it tested a few times before it finally acted up under test.
 
Idling low voltage is not rare.
Before you do anything, clean all of the cables you see, both ends, and where they seat - nice and shiny.
You HAVE to have this right for bunches of reasons.
A dirty commutator or weak or worn brushes in the alternator are possible suspects.
 
Since it's reman'd I'm assuming it'll most likely be the voltage regulator. I'll search online to see if this comes with a rebuild kit or whether it's best to replace with another reman'd alternator (should the load test come back iffy or UNSAT)

All contacts have been cleaned thoroughly of corrosion and reconnected. There was no difference after this, so I'm doubting this was the cause. Also, what jstutz said makes sense in that it would have to be a major short to ground to really affect voltage this much.

The wire insulation was light-to-medium brown, and looked either like it had been subject to too much heat over a long period of time. The wire runs directly over the boxer engine, which tends to get pretty hot. I think it's just a bad design and, over the years, has slowly caused the insulation to degrade, become brittle and crumble away. The sheathing crumbled away as I tried to cut it back to see how far back the insulation was gone. As soon as we got more to toward the fuse box, away from atop the engine, the wire insulation was whiter and the sheathing less brittle.

I wouldn't be so concerned with low voltage, but, to me, low 12s at idle is too low. Also, kind of like UOA's, I look for trends. When I first got the car all I'd see was mid to low 13s at idle, depending on load. Now I frequently see mid to high 12s at idle and mid 13s at higher RPM's. This shows me that something has changed. What, I'm not sure.

One part that I did leave out is that the previous owner installed JDM headlights, which is pretty common, and a really good upgrade over the stock headlights. I checked all of his wiring, from battery and ground to relay sense and power points, and everything looks good. I do see some corrosion, and when I took a sniff near the relay it smelled slightly like over-heated insulation. I also, according to the scangauge, notice a 0.4-0.6 drop in voltage when I turn the headlights on. When I turn them off it goes back to normal. What I normally see, at idle, is a drop from ~13.4V to ~12.9V.
 
The ScanGauge is presumably measuring voltage at the OBD port. It that really the best place to assess "system" voltage? Only if the relevant wire goes directly to the alternator output without being loaded along the way.
The alternator output voltage at idle is going to be very sensitive to system loading, the condition and charge state of the battery, and the exact idle speed. As such, I don't think you should read too much into it.
 
I had the battery tested today, and it passed.

I've loosened the alternator pivot bolt - to tighten the belt - but never fully removed that fastener. I plan to spray it with deep creap and let it sit over night, then remove the alternator and have my lady drive me up to have it tested tomorrow. I was tempted to do it there at AA, but, with my luck, the bolt would snap or I'd cross thread it, and be stranded. haha

As far as how accurate the scangauge is, I have no idea. I haven't ever hooked up a reliable, calibrated Fluke or any other multimeter and noted the reading compared to the scangauge. maybe I'll rig something to check it out. Either way, I wouldn't assume there would be that much of voltage drop from the output of the aternator to where the OBDII picks up. Also, like I said before, I'm not going based on single data points here, but trends. I observed predictable readings before, and have only recently been noticing changes and variations from before; what I'm calling normal operation.
 
Last edited:
Check the battery voltage with a real voltmeter. The point at which the ECM monitors voltage may have a bad connection. Of course if the ECM is also the voltage regulator, that might be a problem and the battery might actually be getting OVER charged.

First step- diagnose with an independent tool.
 
Found the multimeter. With the car at idle - 700 RPM - heater on high and headlights on, voltage was 12.4V. Multi read 12.34V. Heater off, up to 13V. Heater and lights off, up to 13.5V.
 
will a throttle blip to 1200 rpm settle out the alternator at a decent voltage? Say 13.5+ volts. This is very common on the Bosch alternators in old BMWs
 
With the lights on and the fan on high, the voltage this morning read 12.4V. I blipped the throttle to 1250RPM and the voltage read between 13.2V and 13.4V.

I'm going to try to get the alternator load tested some time this week.
 
I think you either have a bad connection somewhere or a failing alternator. Anything less than 13V with the engine running is too low. If you don't mind the labor, pull the alternator and have it tested. You might also turn on all the accessories and when the voltage drops, wiggle wires going to the alternator and battery and see if it momentarily improves. This may help you find a bad connection.
 
Good suggestions. I'll most likely take the alternator in this weekend to have it tested. I 'm giving the car to my little sister, and I need to have everything good to go by next Tuesday. Oddly enough, the car seemed to stay above 13.4 all night tonight. I should have tested with the fan on, but usually the lights will cause idle to drop below 13V. Hmmm.
 
Status
Not open for further replies.
Back
Top Bottom