O2 sensor placement on an EJ22T ?

[legacy92ej22t]

Well I've been having some issues with my idle and thanks to vrg3 we discovered my O2 sensor isn't working and I was leaned way out at idle .

Some of you are familiar with my turbo-back and the fact that I am running my O2 sensor just before the mid-pipe cat, which is really not a very good idea, instead of on the DP like it should be. I had extended the O2 sensor wiring using the double connector method and it seems that the wiring is now having problems as well. I figure when I was driving through some deep slush I damaged either the connectors under the car or the wiring. Now instead of fixing the wiring and having a sub-par O2 sensor setup still, I guess it's time to move the sensor.

Now here is my dilemma, I have a ceramic-coated CES DP and I don't want to screw with it really. I also don't trust the local shop around here to do a good job on it and not mess up the ceramic coating and cosmetically ruin it (petty I know but it's soooo damn pretty!).

So after all this here's my short question, could I put the O2 sensor in the up-pipe or is this ridiculous?

[vrg3]

Like I said to you in person yesterday, I think it ought to work...

It'll see higher temperatures than usual, so it may have a somewhat shortened lifespan... and you'll have to route the wiring very carefully to avoid the heat...

I think you're supposed to mount oxygen sensors so that they're aimed downwards a little for proper drainage. That may be a little bit of a challenge but you ought to be able to do it, if only a few degrees.

Now I'm looking at a parts diagram for the WRX, and it looks like the WRX doesn't actually have a sensor at the top of the downpipe at all -- it looks like their front wideband oxygen sensor is actually in the passenger side exhaust manifold! Maybe you could put your sensor there. Just angle it a few degrees downward for drainage. Since WRXes have them there from the factory I'd say it's probably safe.

Hmm... I wonder if you could pick up a passenger side WRX exhaust manifold cheaply from someone on NASIOC who installed headers and just use that.

[legacy92ej22t]

Like I said to you in person yesterday, I think it ought to work...

Hehe..I know and I trust your judgement. I just wanted to start a discussion about it, that way we can get some more feedback, ideas and opinions.

Now I'm looking at a parts diagram for the WRX, and it looks like the WRX doesn't actually have a sensor at the top of the downpipe at all -- it looks like their front wideband oxygen sensor is actually in the passenger side exhaust manifold! Maybe you could put your sensor there. Just angle it a few degrees downward for drainage. Since WRXes have them there from the factory I'd say it's probably safe.

Really? That's interesting. That might be easier then trying to put it in the
up-pipe.

Hmm... I wonder if you could pick up a passenger side WRX exhaust manifold cheaply from someone on NASIOC who installed headers and just use that.

That would be pretty cool. Do you think it would bolt right up to the front joint pipe and up-pipe?

[vrg3]

Yeah, we should definitely get some people who've actually mucked with the pre-turbo exhaust to chime in...

I don't know if it will bolt right in. I do know that the driver side manifold is actually the same part number for WRXes and turbo Legacies. But the cross pipe is a different part number, and of course the part number is different for the passenger side manifold.

Some relevant threads:

http://bbs.legacycentral.org/viewtopic.php?t=1434
http://bbs.legacycentral.org/viewtopic.php?t=1550

Kelley says he's using WRX headers from HKS that fit perfectly. Mike says he's using GT Spec headers that are shaped a little differently, making the turbo sit 4mm lower, but he doesn't say what car they're meant for.

Don't WRX owners sometimes upgrade to Legacy Turbo up-pipes?

[vrg3]

You should PM this guy on Nabisco: http://forums.nasioc.com/forums/showthr ... did=462401

And see if he'll give you just the passenger side manifold cheap. It looks like he's willing to give people good deals.

[legacy92ej22t]

Sweet, thanks for the heads up. I just read those links and it sounds like it will bolt right up if I use my oem up-pipe . I'll probably have to wait till after the holidays to try and get it though . Or maybe I should save up and try to get a nice aftermarket header. Hehehe, now that'd be REALLY nice.

[Matt Monson]

I have done a turbo back install on my GF's '04 WRX and can tell you the first O2 sensor is in fact in the driver's side manifold and the second one is after the cat in the DP (or in the midpipe on a stock one)

I am a little confused by the diagnosis of your problem based on my recent experience. Why? Because I have been running for a week without any O2 sensor since I installed my DP with no driveability concerns to speak of. My DP has a bung for it, but since I haven't lengthed the wires yet, I have been running without and with the bung wide open leaking exhaust.

If anything, it runs a little rich at idle based on the smell. Otherwise the car drives fine. I does throw a CEL about half way from Boulder to Denver, but nothing changes in the driveability and the engine sounds totally fine. When I turn it off, the CEL away and only comes back on with drives over 20 minutes. Obviously, I won't run like this more than a few more days but it makes me wonder if moving your O2 sensor is really going to fix your problem. Or maybe you guys can edumacate me a little more since I am by no means an expert on O2 sensor function...

[vrg3]

If you're running without an oxygen sensor and the ECU realizes it, then you'll just end up running a little rich with conservative ignition timing.

Matt's problem was that there was some kind of short in his wiring that made the ECU think the sensor was there and reading rich all the time. We resolved that by disconnecting the non-stock wiring, so his car's running okay without the sensor.

Ideally you want an oxygen sensor, though, to get good fuel trims and to avoid a needless trouble code. Matt (the original poster)'s car has no good place right now for a sensor. The only bung is way down in the midpipe, which means its readings will be stale.

Matt (Monson), from your observations, does it look like the WRX manifold will bolt up to the Legacy Turbo with the stock crosspipe and uppipe?

Oh, and you can use a big spark plug to plug that hole in your exhaust if you like; the threading is identical to that on oxygen sensors.

[legacy92ej22t]

Vrg3 can explain this all much better then me because I didn't understand everything he showed me exactly....

I had extended my O2 sensor wiring to the WRX turbo-back location and we think it was shorting or something causing my ecu to lean way out at idle. Vrg3 hooked his scan tool up to my ECU and it was doing some wierd stuff. The O2 sensor wasn't acting right but at the same time wasn't throwing a CEL or code in test-mode. It was constantly sending a signal to the ECU like it was shorting. If you aren't running an O2 sensor or are in open loop (I think) you would get a 0.0 reading on the scan tool and I was getting something like -21 which wasn't good. When you run without an O2 sensor you run on your baseline maps. We first tried to unhook the O2 sensor connector that is by the sensor to get my car to throw a code and run on baseline maps but it didn't change anything (which is another reason we think the wiring is bad) so we unhooked the connector in the engine bay and that did the trick. Once we did this and reset the ECU the car ran a whole lot better. The IAC valve seemed to work better (refer to thread of mine in engine forum) and my idle smoothed out a lot. I'm now running without an O2 sensor right now and under WOT my car isn't as fast on the baseline maps and runs richer but probably much safer. I shouldn't run my O2 sensor so far back in the exhaust and don't want to put the bung in my bueatiful DP so we were exploring the idea of putting it pre-turbo in the up-pipe or passenger side exhaust manifold.

Wow that was kinda all over the place, hopefully it makes sense

[legacy92ej22t]

Vrg, you posted while I was typing. I need to learn to type faster I guess

[vrg3]

Hehehe, yeah, but you brought up a good point that we should tell everyone about.

There are basically three states the sensor can be in:

When it's not warmed up to operating temperature, its signal is floating.
When it's at operating temperature and the mixture is rich, its signal is less than about 450 millivolts.
When it's at operating temperature and the mixture is lean, its signal is above about 450 millivolts.

In Matt's car, there was presumably a short between the signal wire and ground (somewhere on the car's chassis). So the signal line was not floating, and had about 0 (milli)volts on it. This made the ECU think it was running way rich, so it kept making the fuel trims more and more negative, until it hit its limit (20.3% less than the base map).

So whenever the engine was running closed loop, it was running really lean. This includes idle.

I don't know exactly what effect these trims were having on open loop operation, if any.

Moral: It's bad to let your oxygen sensor signal line short to ground! You'll run lean!



(That's me and Matt above)

[legacy92ej22t]

So whenever the engine was running closed loop, it was running really lean. This includes idle.

I don't know exactly what effect these trims were having on open loop operation, if any.

I guess I got the closed loop and open loop mixed up.

(That's me and Matt above)

[Matt Monson]

Ok,
That all makes sense. Thanks for the primer on O2 function. As for the questions on the up-pipe and cross-pipe fit, my recent expereince installing my turbo back suggests that you will never know until you try it...

Personally, I would put a bung in your downpipe. If it is done right it will be fine. And I would get an EGT bung installed why you are at it. But that's just me. I totally understand your desire to keep your DP pretty and clean...

[Legacy777]

I'd just go for the dp too....but that's me....

[vrg3]

Wait a minute -- I had that backwards above. Rich mixtures yield voltages above ~450mV, and lean mixtures yield voltages below. I have no idea what kind of a short Matt's car must have had.

About placing the sensor in the downpipe -- the problem with that is that it'll cost a lot of money to have someone do it, and it'll cost even more to have someone who won't screw it up do it. Lots of people think they can weld, many more than actually can. And this isn't an easy job -- a stainless steel pipe with a ceramic coating.

If the WRX manifold fits or can be made to fit I think that's the way to go. You'd even get faster response out of the sensor. The extra heat would also mean you could probably easily get away with a 1-wire sensor if you wanted to cheap out on it.

[Opie]

Hmm... I wonder if you could pick up a passenger side WRX exhaust manifold cheaply from someone on NASIOC who installed headers and just use that.

That would be pretty cool. Do you think it would bolt right up to the front joint pipe and up-pipe?

The WRX passenger side manifold WILL bolt right up and the flanges mate up to the up-pipe and center flex pipe perfectly. I installed an entire 3-peice WRX exhasut manifold on my Legacy today. My stock manifold and heat sheilds were rusted nasty so I picked up all 3 pieces from a WRX owner for free. The O2 bung would work perfect for what you are needing, it points out towards the wheel well where the manifold meets the up-pipe. I had to find a plug to plug mine!

[Opie]

Here's a pic:


The bung is in through that hole! Hope this helps!

[ciper]

Never thought Id see a picture of Opie's Bung Hole

What a great idea. I had planned to put my 02 sensor in the DP as well. I guess aftermarket WRX headers would provide the same result.

[Opie]

[legacy92ej22t]

Sweet! That takes the risk out of getting them, thanks! I will be doing this very soon.

[Legacy777]

uhhuhuhuhuhu......you said bung hole! uhhhhhuhuhuhu