Vacuum leak? Other possibilities? A/F Learning

I was throwing a lean CEL and A/F learning was maxed out at low MAF. I did a smoke check for leaks and the Turbo Inlet Hose was torn and leaking moderately.

 

I replaced it and here's what the new A/F learning values:

 

0 - <5.60.....5.60 - <10.00.....10.00 - <40.00.....40.00+

.....15 ..................1.6........... 0.6..................... 0

(The FLKCs were all zero, IAM is 1)

 

 

 

To me that appears to be a small leak, but I did the following to check for leaks:

1. When installing the Turbo Inlet Hose I rechecked all the lines (I had the TMIC off).
   
2. I did a smoke test and found no leaks (yes, it was a homemade system, but it's worked pretty well before).
   

Possible next steps?

1. Get a professional smoke test done (not sure it would be that much more effective).
   
2. Replace the O2 sensor? (Not sure if this is applicable to the readings I'm getting.)
   
3. Tuning: Any possible issues with MAF scaling, injector scaling, etc.? The car has ran ok in the past with no A/F Learning issues.
   
4. Injectors: Could there be injector issues?
   

Subscribed....

 

I had torn inlet hose and replaced and the first trim is high as well. Not maxed out but around 10ish usually.

I take it you reset the ECU? If so, you still have a leak, a big one. 15 is the A/F correction maxed right out. MAF sensor o-ring okay? Forget a hose someplace? Tore your new inlet tube?

Good thought but I did reset the ECU. As noted, I did a new smoke test and not a bit of smoke. It's a homemade test, but smoke was coming out significantly on original torn turbo inlet. Now with new one, not a bit of smoke anywhere. I realize that doesnt guarentee no leak, but seems to be an ok indicator.

 

The cold air intake was off during the test though so good idea to check MAF sensor. I'll recheck the throttlebody hose too but it looked good last I checked.

 

I can try the whole carburator spray thing to look for leaks but that method didnt work for me even on the torn turbo inlet hose...might not have sprayed it exacltly on the spot.

gotta be a vacuum leak somewhere. Check all hoses to/from the manifold.

Will recheck. I agree it sure looks like a vac leak. When I put on the AC at idle, the slight bit of extra load causes trim to change...again, that looks like a leak.

 

 

O2 Sensor: Wondering if it's possible for O2 sensor to be working but somewhat "clogged"...I would think this would produce a lean reading. Even thinking it would be worse at low load (not much exhaust gases) and not as bad at higher loads, though not sure this is right thinking (though I do think the result would be lean, just not sure if lean over entire load range or worse at low load). I need to research if this has ever caused someone similar problems (yes, I agree it still looks like a vac leak to me).

 

Speaking of O2 sensors, does anyone know what the front one does (near exh manifold) vs the rear (after CAT). I know delta helps measure CAT efficiency, but is just the one by the exhaust manifold the main one for monitoring and adjusting AF Learning?

 

 

MAF: Last, could MAF be messed up and somehow be producing this?

 

 

Smoke Tests: Do smoke tests sometimes miss leaks? I thought they were pretty reliable but haven't really worked much with them. Local mechanic said he'd do one for me for $100 (I already did my homemade one)...hate to shell it out but if I inspect everything and no hoses appear to be loose, not sure what to do.

Forgot to add above:

 

Injectors: Could injectors be faulty, specifically at lower flow?

UPDATE: Did better smoke test in dark so I could see well with narrow light beam; a little smoke was from Air/Oil Separator connection...since it was under a little pressure, I think it just worked through the crankcase system and out the separator at the hose/separator connection (should it do that? not sure how the whole venting system works/is piped.). Also very minute amount of smoke at the MAF sensor when I put it under some pressure with the smoke machine.

 

- Replaced leaking crosspipe (found exhaust leak where flex section meets solid pipe)

- Installed new O2 sensor (might not even be an issue, but had one so put it in)

 

Haven't driven it very long (about 15 minutes) but AF learning values are better (need to give it more time):

0 - <5.60.....5.60 - <10.00.....10.00 - <40.00.....40.00+

.....5 ..................0 ........... 0 ..................... 0

(The FLKCs were all zero except one...checking logs to see where issue was, IAM is 1)

 

NEXT STEPS:

- I need to see what everything reads after a few trips (sometimes the AF learning gets worse and worse so we'll see). If reading are not good:

- Consider replacing MAF.

- Ensure there are no issues with the crankcase venting system.

- Check compression.

- Check injectors?

Any updates to this? I have the exact same problem right now. I can't find a leak but there must be one.

 

My A/F correction is pegged at 25% at idle and I threw a p0171 (bank 1 too lean).

 

I just finished my stage two install along with:

 

catless sti uppipe

infamous oil feed line

perrin inlet/afta maf tube

 

I will drive some more and post a learning view if necessary.

 

Thanks

Still running about:

0 - <5.60.....5.60 - <10.00.....10.00 - <40.00.....40.00+

.....11 ..................7 ........... 3 ..................... 0

 

I have a couple hoses to tighten up (they have zip ties but going to use hose clamps) and the MAF to make sure it's sealed (it had a very slight leak).

 

If none of that works, I'm going to take it to a shop to get a second leak test done. If that shows up negative, I'm going to check all MAF areas inlcuding MAF tab in Rom Raider again and WB O2 pulls to check AFRs at higher MAF and loads. Then I'll rescale MAF slightly if some things just slightly off. I'd like to check individual injectors too but that's probably a pain.

 

I love my Subie, but it's very hard sometimes not having a real good aftermarket shop nearby (like Cobb, PDX, etc.) who can troubleshoot when I'm running out of ideas or running out of time :-0

Any updates to this? I have the exact same problem right now. I can't find a leak but there must be one.

 

My A/F correction is pegged at 25% at idle and I threw a p0171 (bank 1 too lean).

 

I just finished my stage two install along with:

 

catless sti uppipe

infamous oil feed line

perrin inlet/afta maf tube

 

I will drive some more and post a learning view if necessary.

 

Thanks

 

I found my vacuum leak; the purge control valve under the alternator had a hose pulled off of the middle of the tee. I had to take out the alternator to find the other end. It ended up being under the manifold and part of the bpv hose tee.

hmm,pics ?

I'll get some in the morning and post them.

Here are two pics showing where the hose was disconnected at the purge valve and where the other end of the hose is connected. The third is a piece I found (credits to boostin) highlighting where the purge valve should connect according to the fsm.

I had the same problem.....but looks like you found what I was looking for....are you still having an issue ? sorry for the late post..I was on vacation

I forgot to add one thing....the leak your showing caused my vehicle to not idle at all or very rough.

Yeah, I think I got it all figured out finally. My fuel trims aren't perfect at idle, but I am wondering if it has something to do with an aftermarket inlet.

you didn't get a tune with an intake ?.....ummm ya you needs to tune for intake...like bad

you didn't get a tune with an intake ?.....ummm ya you needs to tune for intake...like bad

 

I know that. It isn't an intake; it is the turbo inlet.

cone filter ? ....... tomatoe tomato ...get a tune

I do not have an aftermarket intake. It is the factory intake with the oem filter. Nothing has changed pre-maf. I have a perrin after maf tube and perrin inlet, both of which do not require a tune because they are after the maf.

Interesting, I used the k&n typhoon.... Never filt the need to Change anything else...I'm running 387hp and 401tq

I do not have an aftermarket intake. It is the factory intake with the oem filter. Nothing has changed pre-maf. I have a perrin after maf tube and perrin inlet, both of which do not require a tune because they are after the maf.

 

 

recheck your vac and pcv lines with the inlet, you probably rustled up some things and now something is disconnected.

I have triple checked all conections and done a vacuum/boost test to no avail. one new interesting piece though.

 

When I turn on my a/c and idle the afr learning 1a goes down immediately. The afr learning 1b won't go up though and stays close to 1 or 0.

 

I think I may increase the injector latency a little and see if it helps even though I have stock injectors.

 

Other ideas? I may rescale my maf too but shouldn't need to.

don't touch the latencies, clearly there's something else wrong. if you have a stock intake box and filter, as well as stock injectors, and NO vac leaks your trims should be within 5-6% + or -. assuming your maf scaling was not altered during a tune, and you were tuned with the stock intake setup. when the a/c is clicked on, the idle usually goes up, which means the airflow goes up as well, which means now the scaling is moving to different cell. it almost sounds like the tuner adjusted your idle trims, but maybe not any other ones..... if you're using an accessport ots tune then disregard that.