legacycentral bbs

The symptoms: When parked and idling, or cruising slowly around town, according to Vikash's Scantool, the O2 sensor works fine. Goes rich when I gas it, goes to 0 when I come off the gas, bounces around when in closed loop. If I do a decent pull while under boost, and then stay on the gas (merging on the highway), I eventually get a CEL. The Scantool shows the voltage slowly dropping to 0 while on boost (slow enough that I can see each value well), and the only way to get the O2 sensor to respond again is to spend a couple of seconds off the gas, so that the mixture is nice and lean. Then its fine until I hit boost again.

What I know: Last week, the sensor was fine. No problems since I bought it new from the dealer last July. Friday, had to remove it because it was in the way (I'm running a Brullen turboback, and the O2 bung is at the bottom of the downpipe). It sat on a workbench for 2 days, before I reinstalled it.

What I've tried: 1. Suspecting the extension I had to make to reach the sensor was messed up, I replaced all the wiring up to where it joins the main loom above the transmission. I now only have 1 splice, instead of the 2 before. No change.
2. Reset the ECU, thinking the messed up wiring affected the fuel trims enough to continue causing the problem. No change.
3. Thinking today, after watching how the voltage reacted, that the sensor's ground was messy and being affected by the increasing EGTs, ran a separate ground to the external body of the sensor. No change.

What I'm thinking now: 1. The body of the sensor outside of the exhaust isn't connected to the internal side of it, and my ground did bubkus. I didn't get a chance to check the resistance to ground.
2. The exhaust leak I have at the turbo / downpipe flange is affecting it. But I can't see how that would, as I had a leak before, and it was fine.
3. Something got into the sensor while it was sitting on the table, and when exposed to high EGTs, causes the air around the sensor to read lean. Very possible...
4. The carbon buildup is different on the 2 sides of the sensor (one side's exposed to the exhaust gas, the other side is somewhat protected), and the sensor is reinstalled differently. So the carbon is heating up and messing up the mixture around the sensor. I dunno...
5. Vacuum leak. Vacuum readings are fine, it idles and drives fine, and wouldn't a leak of some kind, when on boost, cause it to run richer?
6. Last one - My brother wonders if it was damaged in the removal/sitting/reinstallation confusion, and is damaged just enough that when it gets really hot, it stops working. Also very possible. I hope its not this, because like I said before, its OEM and new from last July. I really don't want to have to buy another one of any kind already.

My biggest problem with all of this is that it was fine last week, and I really haven't changed anything. Also, I've gotta do an emissions check sometime this month, and with the single hiflow cat, I'm slightly concerned about the results of the test.

Sorry for the long post, but I wanted to cover as much as possible right off the bat. So... Any thoughts anyone?

The heat sounds very plausible. Have you checked your wiring extension?

Just beware that an O2 sensor extension started me down the road to trouble with the O2 sensor wiring harness on the car side, which was the main contributing factor that eventually led to my demise.

Mine was a freaky weird problem though that defied many minds it's secrets.

I'd unhook it though, except when trouble shooting, just to be safe.

Be carefull with your O2 wiring everyone, it's a major factor in fueling....

Yeah, I redid and checked the extension yesterday. Thats fine. Its doing the same thing now as it did with my old extension. And its fine when the car's idling...

I can't see how the car's wiring could have been affected by removing and reinstalling the O2 sensor.

Thanks tho...

mikec wrote:
I can't see how the car's wiring could have been affected by removing and reinstalling the O2 sensor.

Thanks tho...

On mine it was when driving in slush turned icy conditions. My wiring extension ran down under the car to a bung in the cat. It was scraping along for about an 1/8th of a mile and it shorted or something and screwed up my O2 sensor wiring really bad. Car drove fine, hit sluch/ice road and car immediately drove wrong, after turning off, from then on for the better part of a year. It would read really weird and cause extrememly lean conditions so I ran with it unplugged for months. I tried to fix it numurous times to no avail. The O2 wiring harness finally had to be bypassed completely, all the way back to the ecu to fix the problem.

I'm sure that's not what's happening here though. I'm just passing iit along because I'd feel bad if I didn't.

Good luck.

Hrm... I haven't pulled the sensor to check it at all since reinstalling it. That's going to be the next thing I do.

Thanks Matt for your help

you've just described exactly what happens when you have a perfectly good sensor, and an exhaust leak above the sensor. at wot or close you are pulling in O2 through the leak, causing the sensor to see way too much oxygen and thus give an out-of-bounds low voltage, which sets the CEL. if that oxygen were coming from a lean mixture you'd be knocking like a bastard.

you are going to have to fix that leak first

Sorry to interupt, but my CEL light comes on first thing when the car is cold. It is O2 sensor related because the codes tell me it is. After the car is warm the light goes out. But comes on again at various times. I checked it at out Subaru and they said the O2 is fine if it is reading 0-1, but that makes no sense, why would the light still come on? Its driving me crazy!! Any ideas on this aswell?

professor wrote:you've just described exactly what happens when you have a perfectly good sensor, and an exhaust leak above the sensor. at wot or close you are pulling in O2 through the leak, causing the sensor to see way too much oxygen and thus give an out-of-bounds low voltage, which sets the CEL. if that oxygen were coming from a lean mixture you'd be knocking like a bastard.

you are going to have to fix that leak first

Cool! Guess I'll try doubling up the gasket for now, until I have time to get the flange milled flat (its slightly warped).