Weird fuel issue

I don't know where to start with this issue so thanks in advance for your help!

 

My car is an 05 Legacy GT with Perrin IC, DP Invidia UP, Stromung Mid and Cans, AP with Cobb stg 2 shelf map. It's got 150,000 on it.

 

The problem: When I first drive after a cold start, my car seems to have a "sticky" idle after I let off the gas. It acts like it's trying to compensate for too much air by adding extra fuel for a few seconds between shifts. The problem stops after the engine is considerably warm. Also, there is no CEL.

 

I originally thought IM gaskets but if that were true it would lean out. Now I'm thinking boost leak somewhere or maybe a sensor.

 

Please help me figure this out.

Check for leaks around the manifold. Also check the PCV valve since it may sometimes be a culprit in incorrect behavior.

 

The gaskets between manifold and cylinder heads can be suspected, and as the engine gets warm they start to seal and when the engine cools off they starts to leak again.

yeah I bought the gaskets for the TGV/Head joint, but decided to replace the IM/TGV gaskets first sincce that's usually a common problem. They're also much less beefy than the TGV/Head gaskets so I thinking it would make sense if it was the IM gaskets

during warmup the ECU calls for a higher target idle, as the engine temp gets to normal the rpm resumes ~750rpm. you can post a long term fuel trim table from the accesstuner program and that will tell us if you DO have a fuel/air issue.

I knew about the higher initial target idle, it just seems like it's really going overboard on fuel delivery for several minutes when it's cold out. Thanks so much for your input fellas. This issue doesn't EXACTLY line up with the other guys who have IM gasket issues, but it's close enough symptomatically, a cheap place to start, and they need done anyway. I'll wait until this mild air moves out to get some cold start fuel logs.

during warm-up fueling is also rich. again, get a long term fuel trim read-out in the cobb software to determine if you have a fueling or air abnormality. or a learning view.

Here's the latest data done today...any thoughts?

12.6.12 Datalog.pdf

that log tells us almost nothing. you need to log knock parameters. you're DAM isn't 1, so we know you're knocking a little. long term fuel trim is fine

Sorry, didn't have my AP updated and didn't realize what I needed, here's a fixed log

12.6.12 Datalog.pdf

ewwwww 4* of learned knock, ewwewewwwww

Yea I don't know what's up but it just doesn't feel the same.

though I thought if it returned to zero (which it does) it's not of great concern

you're thinking of feedback knock? getting feedback during driving/pulls is the ecu's signal from the knock sensor to pull timing NOW. learning knock is past knock that the ecu now expects to see on a regular basis(after the car actively knocked) thus pulls timing so the car doesn't actively knock every single time you hit that particular rpm and load point. fine learning knock will slowly reduce back to zero if the engine will take it, same with the IAM. this is subaru's way of dealing with a tank of bad gas....the car will knock and the ecu pulls timing so it doesn't blow itself up during the entire tank of bad gas, it will slowly add timing back in- hoping there's decent quality gas in the tank now. but, bad gas is certainly not the only reason these cars knock. so seeing an IAM or DAM less than 1 on a regular basis means it's not the gas, it's mechanical or a tune issue.

 

 

maybe you meant since the fine learned knock decays to 0 by the end of the pull? that is irrelevant. it just means your car wasn't knocking in that range(row). you still were knocking at the previous points where you see learned knock being pulled. remember, when the ecu thinks there's knock, it pulls timing in that particular load and rpm area. that's why 3rd gear pulls are done from 2k rpm to redline....so we can see if the engine knocks and at what particular point.

Thanks for explaining that to me...I was just about to go on a lengthy self-teaching romp. It's possible it could be a bad tank of gas (93 that I got from a Getty Mart/Exxon in a tiny town with one stoplight), which I'll try to eliminate as variable next week, but I'm leaning towards the tune. It's a Cobb stage II map that the previous owner had for his mods (which happen to be the same mods I have). I probably just need a good tune. I might run another pull just to check validity. Thanks so much for schooling me!

what you should immediately do is load the 91oct cobb stg 2 map. then re-evaluate. it runs a little less timing and is less prone to knock. get a decent tank of 93 in there, and see if it still knocks. cobb OTS maps do knock, but usually not 4*...could be the combination of bad gas and mediocre OTS map.... just load the 91oct one and the car will be happier.

Done and fixed (AFAIK)! thanks again

New issue, fuel related. After switching to Cobb 91 Shelf map, I'm getting a CEL for P0171 (system too lean bank 1, sensor 1). I'm going to start with the basics and work my way up...

 

A code P0171 may mean that one or more of the following has happened:

 

The MAF (Mass Air Flow) Sensor is dirty or faulty

Note: The use of "oiled" air filters may cause the MAF to become dirty if the filter is over-oiled. There is also an issue with some vehicles where the MAF sensors leak the silicone potting material used to protect the circuitry.

There could be a vacuum leak downstream of the MAF sensor

Possible cracked vacuum or PCV line/connection

Faulty or stuck open PCV valve

Failed or faulty oxygen sensor (bank 1, sensor 1)

Sticking/plugged or failed fuel injector

Low fuel pressure (possible plugged/dirty fuel filter!)

Exhaust leak between engine and first oxygen sensor

 

I'm thinking it could be a few things but least likely a boost leak since I hold pretty strong consistent numbers on target boost. Is there any trick to telling a faulty sensor from a true read?

typically a lean code means vacuum leak, like a loose hose, throttle body coupler or any hose attaching to the intake manifold, or improperly seated intake box. bad intake manifold gaskets can throw this code especially if cold starts are accompanied by misfires. you can log roughness on each cyl, maybe doing so during cold starts can confirm the manifold gaskets being an issue.

I thought that too, Ripe, however, I checked cylinder roughness after I installed the new IM gaskets about 2 weeks ago and everything was clear. When I clear the code, my idle is really rough for about 40-50 miles of driving and it wants to stall out when it's in neutral, (much like an IM leak) and then it balances out. This is the second time I've gotten this code but it takes a few weeks to even pop up. It's also consistent with the time I switched to the 91 Map

you probably have a bad vac leak somewhere, or your inlet is torn. the cobb ots 91 and 93 map does not affect fueling since they are both based off a stock intake stock injectors and pump.

I'll check those first though I didn't expect it to hold boost so efficiently with a torn inlet. Since it's a stock inlet, you may be right. Thanks big guy! I'll let you know how I do.

torn inlets can be funky, sometimes a tear can cause lean and overboost. some tears cause lean at idle and throughout boost but not overboost

I'm guessing the former then. And I know what comes next if I don't fix this...goodbye turbo!

Ok so....here's the update. I cleaned my MAF and at idle I was reading 14.6-14.9 AFR and it was bouncing around in that range. I then brought out the carb cleaner and sprayed around the IM-TGV joint and there was no change in idle. Then I sprayed around the turbo inlet-turbo interchange and when I sprayed, I got a noticeable drop in idle and AFR (12.3).

 

Sounds like there is a tear in the inlet tube. Kinda sucks but I'm glad to have found the problem. Another excuse to upgrade! I've read that people really like the Perrin for durability and that most aftermarket inlets are little improvement compared to the OEM. I'm also planning on heading to a larger turbo down the road so I'm prepping for that as well.

That's really the only reason to replace the OEM inlet tube, is if you rip the damn thing. They're pricey as shit for the gains you get from it. (It's the only reason I replaced mine, as it ripped when I installed my first BNR turbo.)