Weirdness

[entirelyturbo]

Alright, my gauges at night have their usual brightness, and then occasionally, they are exceptionally bright. It's no misconception, it's very noticeable. I see no correlation with anything else.

Now, I must note that on my way home from work tonight, I was approaching a red light and my gauges were at this brighter-than-normal state. I declutched and my engine immediately stalled. It didn't make the slightest attempt to idle, it just went straight for 0.

I haven't had any CEL's so I haven't felt the need to pull any codes, but I will try it tomorrow.

I'm thinking a ground wire might need to be checked and/or cleaned. What does everyone else (this means you Vikash ) think?

[scottzg]

the dimmer switch. Mine doesnt work anymore (i just remembered!) but it did that a few times before it died. Now it's just at full brightness, a terrible terrible safety risk, but one im willing to take.

As for your engine dying, i think it's a timing problem. You have to time the clutch depression to be BEFORE you've come to a full and complete stop.

[vrg3]

What does everyone else (this means you Vikash ) think?

Hehehe...

This is just a random shot in the dark, but I think it's time for you to install a voltmeter in your car (that could just be your handy digital multimeter masking taped to the dash) and watch what's going on with system voltage. Maybe your alternator's acting weird and that's annoying the ECU somehow?

But you say you don't notice anything else happening when your gauges get bright... like the headlights don't get a little brighter or anything? Hmm. It does sound like what Scott suggested there. Either the dimmer switch or the illumination control unit. If you want to know how to test the dimmer switch I can find out...

When your gauges get super bright, what happens when you turn the dimmer back and forth?

[entirelyturbo]

Forgot to add that the alternator is less than 2 years old and the battery about a year old...

[vrg3]

Yeah, I remember that your charging system is fairly new... but it's still something to check.

[entirelyturbo]

Okay, I maintenanced the ground wires today (I'm guessing I got all of em, the one on the - battery cable of course, and the two on either front strut tower).

I took the bolts out, took them to a surplus store to reference them so I could get new bolts and washers, then thoroughly cleaned the contact area with brake cleaner and a wire brush, as well as cleaned the O-connectors on the wires thoroughly. I coated everything in a generous amount of dielectric grease and put it back together with the new bolts and washers.

Now I've learned not to jump to conclusions anymore, as I used to in the past, but from where things are right now, this seemed to make a significant improvement. First off the car seemed much more eager to start, and ran a lot better than normal right after startup. I proceeded to drive it around for about a half-hour, doing as many different things as I can remember that made this stalling problem happen. It didn't even try once at all. It idles solidly at 750rpm now and if it does dip below it for a second, it seems like the ECU actually notices that and immediately brings it back up.

I didn't notice any power increases or anything like that, but it does seem to run slightly better and this idling problem is fixed for now.

I'll drive it around tonight and see what this did as far as my gauge-dimming weirdness and report back.

[vrg3]

Hm, cool...

There are several other factory ground points in the engine bay. Look on the vertical faces of the wheel aprons, and near the driver side headlight, for example. There's also one on the intake manifold by the rear left runner; this one is used for engine management.

I think most, if not all, of these use M6x1.0 bolts.

[entirelyturbo]

and near the driver side headlight, for example.

Isn't this the one off the - battery cable?

Yeah, they're all M6X1.0 bolts...

I'll check the others as well...

[vrg3]

Oh yeah, the one near the headlight is the small leg of the negative battery cable...

I'm very curious to see if your occasional stalling actually has to do with a grounding issue. How did the bolts and terminals look when you removed them?

[entirelyturbo]

They didn't look terrible, but there was a fair bit of corrosion and gunk around and on the terminals, and the bolts were old and rusty. I cleaned it off thoroughly and just replaced the bolts and washers (a whopping 75 cents).

I actually did see the idle dip a little too low at one point where I was sitting and it was dropping to 500rpm a couple times. But it wasn't for long and I still have not had anything even remotely close to stalling. I got on it hard at one point and declutched at 70mph at 3k in 5th, and it actually went to a high idle at about 1250rpm, so I think it knew I was moving along.

I might replace the O2 sensor later if this doesn't do it, as that's still original and the wiring looks pretty nasty (dirty and gunky, not frayed though) on that sensor.

[entirelyturbo]

It seems that it has pretty much fixed the gauge-brightness thing too. It's now permanently at the brighter setting! I definitely know it's brighter because I actually notice it everytime I get in the car at night, without even thinking about it.

I'm still having some slight idle stuff, upon declutching a couple times, it did drop a bit too fast, to where I wondered if it was gonna stall, but it was able to catch it before it got too low. It did try 500rpm a couple times again. But the improvement is significant and noticeable.

Wow. I guessed something right the first time, and what's more, it was electrical!!!

[vrg3]

What happens when you adjust the rheostat?

Maybe before your gauges were always just a little dimmer than they were supposed to be.

Did you reset the ECU after improving all the grounds?

[entirelyturbo]

Well, I didn't fix the stalling problem Happened again on the way to work.

I think I'm gonna see if any codes are stored tomorrow, although I have no CEL at all, and haven't had one in 2 years.

The gauges started playing the dimming game again. It's brighter more often, but it's still dim most of the time.

I'm starting to think that this little ground-cleaning adventure only seemed to do good for one day

[vrg3]

D'ooooh... Yeah, a lot of people who do the "grounding mod" say it seems like the car learns around it after a short while and then goes back to how it was.

I wonder if this could be something like a clogged fuel injector.

I'll ask you a third time, and then no more : What happens when you try to adjust gauge brightness when this gauge problem is happening?

[entirelyturbo]

Haha. The gauge brightness adjusts just fine, the switch works fine, no glitches in that at all. Doesn't matter if it's super-bright or a little dim, it always adjusts in the four increments, in what seems to be fairly equal changes between them all. And the dimmest one is always the same.

The car really ran terrible today. It had absolutely no idea where it wanted to idle at, and it stalled on me yet again.

However, I did try pulling codes today. Put it in D-check mode and went Xmas shopping (yeah I'm draggin my feet this year ). The first time I started it today, I was a little ways away from the house and I was momentarily distracted by traffic, so I only caught the end of the first code, which was a 5. Then it showed me code 35 which I find strange, since I haven't had a CEL and my CPV is only about 3 years old. And then it just went to the steady flashing of no more codes. The next 2 times I started the car today, the codes would not come up, it just flashed no more codes like it did before.

Referencing the codes, there are 3 codes that end in a 5: 15 - #2 injector abnormal output, 35 - CPV, and 45 - Atmospheric sensor.

I believe I'm correct in saying that only turbo cars have atmospheric sensors, so we can eliminate that one. I don't know for sure, but I doubt that it showed 35 twice in a row, so it's possible I have an abnormal injector output code stored.

However, I don't want to just start ripping my fuel system apart on a hunch. Is there any way I can check said injector, without a bunch of fancy equipment?

Vikash, you've mentioned making a Select monitor of sorts out of a laptop computer. Can someone like me do the same with a friend's computer? What wiring and software would I need?

[vrg3]

Hmm, so the dimmer always works, but the intensities change across the board when they change at all? It kind of sounds like a voltage issue.

My car occasionally gives me codes 35 and/or 45 if I try a D-check while the engine isn't warmed up yet (which you're technically not supposed to do). Once the engine warms up a little, it changes to the "all clear" flashing. It sounds like the same deal is happening to you.

It might have flashed 35 twice if that was the only code. It would just keep flashing it over and over until it switched to "all clear."

Your car has an atmospheric pressure sensor; it's built into the ECU on 90-92 non-turbos. Turbo cars have a different pressure sensing setup, and 93-on non-turbos don't do any pressure sensing.

I don't know how you can check a fuel injector for clogging while it's still in the rail. In fact, I don't know an easy way to do it anyway.

Yes, someone like you can build a scantool with a friend's computer: http://www.graphics.cornell.edu/~v/bcbfscan/. Be aware that on your car it can't measure injector pulse width at high resolution.

I think you should go to Advance Auto Parts and buy the $16.99 voltmeter over in the rice section, and then see if you notice anything interesting. After my alternator died the other day I picked up that gauge when I picked up the new alternator, and I just ziptied it to my cupholder and wired it to a cigarette lighter connector. Ghetto as all hell, but it's nice to see what's happening.

You can also monitor system voltage with the scan tool, but it's hard to safely watch a laptop screen while driving.

[entirelyturbo]

Okay, for shits and giggles, I unplugged the O2 sensor before I left work tonight. I know for sure now that I no longer have a heating element in my sensor, because it ran a lot better when it was cold that it normally does. It's usually VERY grumpy when it's cold. So as I figured, I could use a new sensor, although my gas mileage is fine and I have no codes for it.

It did absolutely nothing for the idle problem though, it actually exacerbated it if anything. It didn't stall on me, but it had no idea where to idle at all. So while an new sensor might slightly help the situation, it's definitely not going to solve it, as I hoped it would

And these cars really don't give a rat's ass about the O2 sensor either. It took 7 miles of driving before the CEL finally came on. Didn't run the slightest bit different, except at startup.

[vrg3]

Hmm, it shouldn't have run differently when cold if the heater's broken. When the sensor's not warm it has no output signal, unless something's wrong.

Did you reset the ECU after disconnecting the sensor?

[entirelyturbo]

I have yet to do an ECU reset at all. I was hoping I'd find a cure for this without it.

It would make sense that there's no output signal from the O2 sensor when it's cold. That's, like, kinda why the CEL didn't come on until 7 miles later

I dunno, I think I'm gonna run another bottle of fuel injector cleaner after this. If it does get better, I might just take the plunge and condemn all 4 injectors and replace them with new ones

[vrg3]

Well, until you reset the ECU, it's still remembering its old fuel trims. To really test it running without an oxygen sensor I think you should reset it.

Buying four new injectors seems like a big jump to make... We don't really have good reason to suspect the injectors. Before doing that I'd just have them cleaned in-car with one of those injector cleaning machines.

[entirelyturbo]

So you think I should do a cold battery-unplug, and then start the engine with the O2 sensor unplugged and let it get up to temp? Or should I pull it after the reset process completely? I would think it would get pretty pissed if I started it without the sensor for the reset process.

I think I'm going to do a couple other things like maybe clean the contacts on the IAC valve plug (but nothing else!! I'm not going through that again!), and maybe test fuel pressure to make sure it isn't the pump in its earliest signs of failure. At least this way, I'll use the process of elimination to determine what is NOT causing this.

[vrg3]

So you think I should do a cold battery-unplug, and then start the engine with the O2 sensor unplugged and let it get up to temp?

Yes, I think you should do that.

I would think it would get pretty pissed if I started it without the sensor for the reset process.

Nah, it should be fine. It probably won't even throw a code.

[entirelyturbo]

Should I plug the sensor back in after I turn it off to complete the reset? Or leave it unplugged?

[vrg3]

Leave it unplugged, and see how the car runs without it and without any memory of ever having it.

[entirelyturbo]

Update:

I don't think any of this stuff will be necessary, I'm pretty sure what's wrong now.

I started it after work earlier this week, and it had its dash lights at the dimmer setting like it normally does and was idling at about 1700rpm like it does when it's cold. Then, the dash lights and the headlights got significantly brighter, and the idle started dropping quickly. About a half-second later, the dash and headlights dimmed again, and the idle went back to 1700rpm.

This is also after I've run almost a tank of fuel injector cleaner with still the same occasional incidents, so I feel confident saying this isn't a fuel delivery problem like I had originally supposed.

I'm going to finish up reconditioning the remaining grounds on the rest of the car, and do an ECU reset. Then if that doesn't make a difference, I think I'll rig up a multimeter into the cig lighter and see what the system is doing...

I will report back soon.