Exhaust Leak and Oil Leak P2096

Hi,

 

2005 Legact GT. Stock. Engine threw a code a days ago during a bout of freezing rain. The code is P2096 Post Catalytic Fuel System too lean bank 1 and it is accompanied by rough idle. I have experienced some rough idle/poor performance just when first starting up in the past when it has been cold/humid at the same time.

 

So I decided to take my car into the dealership for diagnosis. I asked them to determine what was causing the code (I think maybe an exhaust/boost leak) and I have also been having some oil leak issues so I asked them to look into it at the same time.

 

Well, they gave me non-specific analysis, "oh there appears to be some leak around your exhaust where it goes to turbo" and "yes you appear to have some oil leaks". Well instead of telling me where they were, they just recommended an entire engine re-seal at a cost of $5500. I don't think this is necessary as I just had new shortblock installed 100k and less than 5 years ago and all the gaskets/seals were done at that time. Obviously I am not going to go this route so I want to troubleshoot a little myself.

 

The exhaust leak is most concerning because there are performance issues as well as a code. I have my TMIC out right now and am looking around and going to perform a boost leak test. Any tips on where and what to look for would be appreciated.

 

While the TMIC off, I did notice oil pooling on the top of the engine, so wherever my leak is, it seems to be really high, above the block. The oil seems to be coming from somewhere near the bell housing and I have attached a picture. There is also a pipe or hose that looks pretty gummed up directly below the turbo heat shield. That might have something to do with the exhaust leak also. I have also attached that picture

 

So a dual ask.

 

 

1. Any insight or tips on where to look for the boost leak.
   
2. Anybody know what could potentially leak oil high up on the engine like that. Its either above the block, or right at the top of it not far from the turbo inlet.
   

 

 

Much Thanks for any help rendered.

1. http://legacygt.com/forums/showthread.php/intake-vacuum-leaks-and-you-pics-inside-235626.html

 

2. You might need to replace your PCV hose assembly and it looks like possibly your turbo oil line copper seal washers. KEEP THINGS CLEAN!! Don’t want crap getting into any of those lines. The washers shouldn’t be THAT bad...but your PCV hose assembly is going to be a pain. One of those hoses is connected to the turbo intake hose and has a plastic lock that is hard to get to.

I’m guessing you disconnected the hoses from your blue t???

Yes,

 

I pulled them out to test them. Haven't had a problem with them yet.

I am trying to perform a boost leak test however the procedure I am reading I think is for a different engine. https://forums.nasioc.com/forums/showthread.php?t=2205161

 

Does anyone know what hoses I have to remove and plug on the 2005 LGT? I will be hooking up at the inlet hose. I have the TMIC off.

http://legacygt.com/forums/showthread.php/1st-time-doing-boost-leak-test-help-124680.html

OK, first thing I did was cap my inlet pipe, remove little hose from BOV, blow on it and then put thumb under it. It held air easily for 10 seconds. Does this test vacuum or boost? I believe I am looking for boost leak probably in the exhaust so not sure if this would test pre or post turbo.

 

If I do the boost test, does this test the same thing or different as what I just tested?

I'm fairly confident that an exhaust leak isn't going to give you a "lean" condition. Your A/F sensor is prior to the turbo so there will be a slight positive pressure. If air is getting sucked into your exhaust...then your car has bigger issues.

 

It is odd that your trouble code doesn't show up in the manual...or on the sticky at the top. But whatever, here's things to think about:

1. Possible major air leak between your MAF and engine...your intercooler should be connected for the leak test, or your can put everything back together and connect a vacuum gage to a vacuum drag line after the throttle body and make sure the vacuum is within spec

2. Your A/F sensor could be bad (O2 sensor prior to the turbo)

3. your fuel system isn't able to deliver the volume needed (need to see what your fuel pressure is at idle)

 

You can spend a lot of time and money chasing this blind...or you get a code reader that will read real time data. Plus it will help others better help you.

 

If you don't want to go out and buy a code reader that can read real time data...then I would be jumping on the PCV hose assembly first. They are close to $60 wholesale. It looks like oil is coming from there...and it could kill two birds with one stone. It is a pain and very frustrating to replace.

http://legacygt.com/forums/showthread.php/trouble-codes-cels-26113.html

 

I would replace the rear O2 sensor.

 

Front O2 sensor would be the codes up around P0133.

 

Your results from post 7 confirm you do not have a boost or vacuum leak. If it held pressure for 10 seconds. If you had a leak you would hear it.

 

I most case you would hear an exhaust leak.

http://legacygt.com/forums/showthread.php/trouble-codes-cels-26113.html

 

I would replace the rear O2 sensor.

 

Front O2 sensor would be the codes up around P0133.

 

Your results from post 7 confirm you do not have a boost or vacuum leak. If it held pressure for 10 seconds. If you had a leak you would hear it.

 

I most case you would hear an exhaust leak.

 

 

Derr...not sure at what time I went full retard to skip over "POST" catalytic.

There was a change in the frequency of the exhaust several months ago. Should a little like loose heat shield or hole in the muffler.

Well we found a missing bolt that might explain the exhaust sound change. Oddly, the P2096 code is gone and has been replaced with P000 and P003. However the codes are memorized, not temporary.

 

Car still running rough so I datalogged it. Maybe someone can take a look and see if anything looks untoward.

 

Took a log when starting, driving cold, warming up and then idling.

20180203_095021-warmed up idling.csv

cold car driving_20180203_093920.csv

longer drive warming up-20180203_094358.csv

starting up_20180203_093913.csv

LVs look really good for a stock tune. None of the log links worked for me.

 

What kind of scanner are you using to pull codes? Not only is P2096 invalid for a Subie, P000 and P003 are invalid, period. Have you tried the car's built-in code reader?

If the problem is not the scanner, it's the ECU or its wiring.

I am using learning view and the blue vagcom cable

Try the car's built-in code reader or any standard OBD2 code reader. Auto parts stores will pull codes for free.

I should note the P003 is a typo. It should be P0303

Looks like it was a typo...the pic shows P0303 and P0000.

So looking at the ROM raider logs, I am seeing some misfires on cylinder 3. On the longer drive warming up there are 12 or 1784 intervals with a misfire count of 1. After the drive, I let it idle for a few minutes and did a log. While idling there was 129 of 429 intervals showing a misfire count basically starting at 1 and going up to 5. 1,1,1,1,1,2,2,2,2,2,3,3,3,3,3,3,4,4,4,4,4,5,5,5,5,5,5. What exactly is this misfire count? Is this an average of several previous intervals or something else.

OK, didn't see the codes on the screen shot. Standard diagnostic for single-cylinder misfire: swap coils 1<->3, see if misfire follows. If not, swap plugs, then injectors. If all 3 check out, move on to compression.

Will do this. I logged another drive.

 

Have noted that misfires only occur when engine coolant temp is > 190 F and also that OCV duty cycle is always 9.41%, whether idling or cruising or accelerating. Is this normal? I have read some posts that 9.41 is normal at idle but should it always stay there?

OK, didn't see the codes on the screen shot. Standard diagnostic for single-cylinder misfire: swap coils 1<->3, see if misfire follows. If not, swap plugs, then injectors. If all 3 check out, move on to compression.

 

P2096 Post Catalyst Fuel Trim System Too Lean (Bank 1)

P0303 Cylinder 3 Misfire Detected

Also pay attention to the spark plugs to see if one of them is having a different color in addition to compression test. If in your case the #3 plug looks like it's running hotter or colder than the others it may be an injector too that's bad.

 

And even if the compression test checks out the valve clearance may have to be checked because when the engine is warm then the clearances may prevent valves from closing causing misfires.

Oh, and the coil connector pins can get loose causing the coil to misfire. A tiny twist to the pins with needle-nose pliers will fix that.

thanks,

 

Want to make sure the OCV question doesn't get buried. I looked at an old log. OCV was 9.41 at idle and something else when cruising or accelerating.

 

Is it weird that now both OCvS stuck at 9.41 regardless of what the car is doing?

thanks,

 

Want to make sure the OCV question doesn't get buried. I looked at an old log. OCV was 9.41 at idle and something else when cruising or accelerating.

 

Is it weird that now both OCvS stuck at 9.41 regardless of what the car is doing?

 

Log your AVCS cam angles both sides at idle, cruise, and pulls.

Cam angles look very similar. OCV also looks OK. It might just have been in open loop when I logged before.

 

Have swapped coil and changed plugs now. Misfire staid in cylinder 3. Log shows roughness only occurs between 2000-2500 RPM which seems weird that it doesnt happen throughout RPM range.

 

Any ideas there?