buffington

That makes sense. Unfortunately it doesn't look like I can log the A/C voltage (unless it's totally hidden in Dashcommander somewhere). I also think that the DC voltage it records is the DC voltage at the ODB2 port - not exactly what the alternator is putting out. There's a lot that can happen to that voltage between that port and the alt so it's probably best not to rely on that reading. I've been hunting around for an alternator in the Salt Lake City area - so far the only sure thing is a new OEM alternator from a Subaru dealer. At $500 I'm much more motivated to go digging for one in a pick-a-part yard. My current alternator is still under warranty as well, so I'll probably get that swapped to see if I end up with one that actually works. That's assuming I can find my damn receipt..

The issue occurred again tonight while I was logging. I thought for sure I'd see high voltages in the logs, but it never climbed above 14.2. My first instinct was that it's possible the logging resolution wasn't enough to capture a spike, but the timestamps show about 20 entries per second. The brake/ABS held solid this time for several seconds - surely enough to show up in logs. There have been times when that combo of lights barely flickers on too - those not showing up would make sense, but not a several second light up. Anyhow, I'll be finding an OEM alternator this week. I'm really hoping it's the cause, because otherwise, I don't have a clue as to why this machine is haunted.

My ODB2 scanner arrived and I was pretty surprised to find no error codes. Maybe that helps support the idea that the alternator is bad (and not the ABS system as dash the lights would suggest). I also drove around the block and logged the voltage. It ranged between 12.8v and 14.2v over a 2 mile run with AC on, lots of acceleration. There were no crazy dash lights this time so stuff was working fine. I'll probably try it again later this evening to see if I can get it to do its funky business. It's kinda hit or miss. I'll probably also compare what the scanner reports to what a multimeter is measuring in real time. Probably not safe to assume they'd be identical. I'd at least hope they'd be consistent.

I thought I knew how to test the A/C ripple, but quickly discovered that measuring between the alternator post (is that the same as the "hot lead") and body is the quick way of letting the smoke out of your cheapo multimeter. That happened this morning, pre coffee, and not really thinking it through. No huge loss though - I still have my good multimeter. After thinking it through I realized I had the positive test lead in the 10VADC port. While the advice was good the follower (me) of advice was not. A/C ripple out of whack or not, I've had enough people suggest that 15v is too high that I'm convinced the alternator is a turd. I'm going to start working on finding a replacement that works. For the sake of learning, let's pretend the alternator was good. Is there any other circumstance where I'd see the 15v I saw? For example, if some other component were malfunctioning, could it be responsible for the measured 15v? I'm guessing no since the alternator is the only component in the system that can produce voltage (if you exclude the battery, which I don't think counts by my definition).

I'm still troubleshooting this. I'm not convinced it's the alternator or the battery. I also checked brake fluid - it's where it should be. After letting the car sit for a few days I tested the voltage early in the morning when it was "cold" (though 55F probably isn't really cold when it comes to batteries). Battery voltage was 12.7v. Later in the day I turned the car on to test it while running. With A/C on, radio on, headlights on, radar detector on, all inside lights on, trunk open (for the trunk light), engine at about 1500RPM, the voltage varied between 14.4-14.5v. Turning the A/C on or off did little to the running voltage. With all lights off, A/C off, stereo off, etc., running voltage was 15.2v. Immediately after turning the car off, the voltage was 13.2v. I also checked for bad ground connections or voltage drops between the alternator and battery. The voltage between the battery negative and the alternator body was around 0.04v with every accessory and AC on. Between alternator mounting bolt/bracket, and between main engine ground to battery was also 0.04v. I checked the positive battery terminal to the alternator positive post bolt, the nut, and the lug on the same post: all were between 0.14v to 0.20v at around 1500rpm, with accessories. Turning the A/C on and off showed no significant changes to that reading. While under full accessory/AC/lighting load I also disconnected the positive lug from the battery and there was no change. Not a surprise since the alternator is obviously producing voltage, but figured it was worth trying anyhow. Based on these tests I'm pretty certain that the battery is charging fine, and that while running, the alternator is at least functional. That said, I know that the tests I did won't tell me if the voltage coming from the running alternator aren't having huge microsecond dips or spikes. For that, I need to do some tests with my oscilloscope, which is what I'll probably do next. I'll also be pulling any error codes in a few days when my scanner arrives. In the meantime, does it still sound like an alternator issue?

The battery and alternator were installed at the same time. It's been a few months now, and the battery voltage appears to be totally stable. Perhaps the battery condition is such that it's masking an issue with the alternator? In any case, I hear what people are saying about the alternator. I'm going to learn how to test it properly so I can feel more certain about it being the issue or not.

I'll check brake fluid. Is it worth testing the alternator beyond ensuring it's putting out the right voltage when running?

I'm baffled - I have a 2006 WRX STI that has very recently started to do some very weird things. Here's what happens with the dash lights: While driving at any speed the emergency brake and ABS lights will flicker on for a split second. They don't stay solid, just blink on then off. Sometimes it's just one blink, sometimes it's a few in a row, but not at any discernible pattern. I'm about 85% sure it happens more when the car hits smallish bumps, but it has also blinked on when I know the road is smooth. There's another set of lights that I just noticed today too: emergency brake, battery, and rear diff light. Those seem to flicker when the car hits bumps, and doesn't seem to have any other pattern. The brake/ABS thing seems to happen more often when it's a bit cold. Perhaps related - when starting the car, it sometimes acts nuts. With the key in the on position (prior to turning over the engine), things will seem normal, then everything can totally die, then be right back to normal again. Prior to all this happening I replaced the alternator with a refurbished one. I replaced the battery at the same time. Here's what I've been able to test: - after sitting for a few days, I've checked battery voltage - 12.7 volts. - I've checked for parasitic draw using a multimeter in circuit with the car completely off, doors closed (aside from open hood). There was minimal draw. A few milliamps at most. - I've removed an aftermarket stereo install, suspecting grounding issues - I've checked ground connections on the engine and at a few places inside - I've checked fuses - none are blown. - I've checked voltage while the engine is running. If I recall, it was close to 14 volts. I'm pretty sure I also checked current while idling but can't recall for sure. What should I do next? I don't mind taking it to a dealer if I need to, but I want to have a good idea of what the problem could be before I go in. If their diagnosis is similar to ideas folks might have here I think I'll trust the dealer more.