IW STi Forum banner

1 - 2 of 2 Posts

2 Posts
Discussion Starter #1
This is my first ever post, but I have been scouring the internet and forums for months. Long post ahead and ill be as detailed as I can. Ill structure as background, issues, tests completed(what i can think of at the time of writing this), then the call for help.


Thanks in advance for any all help. I'm a noob to Subaru's, but am a technician by trade and have history with DSM's. Im factory certified for Honda, Suzuki, Yamaha, Kawasaki, Stihl, and about 10 John Deere certifications from lawnmowers/UTV's to Combines/Sprayers.

I'm not looking to get talked down to, as engines aren't new to me. I just need some Subaru specific help, and I know there are some phenomenal techs on these forums.

I have a 2012 STI(30,000miles) purchased 8 months ago. Already had a Cobb access port v3, invidia catback exhaust, atmospheric BOV, AEM intake, and turbo inlet installed. Stock turbo, injectors, pump, and boost controller.

When test driving, the car pulled hard, no misfires, no codes. Decided to purchased. Had them install the catless down pipe that came with the car, they said it was off because they weren't allowed to sell with it on due to emissions.(should have been my red flag, I don't know if that is true.) Following day I was able to really test drive the car and during 4th gear, full boost, extended pulls the car would smoke screen when you let off the throttle. Well damn... Called the dealership, an even though they sold it to me with a full factory warranty to 60k miles, they wouldn't touch it and said i would have to go to Subaru. O yeah, purchased through a GMC dealership. I knew damn well if I wanted to go to Subaru id have to pull the exhaust(didn't have the stock one), pull the AEM intake(no stock one), and pull the Cobb V3 after flashing in the stock map. I didn't want to take it there and get the car tagged on the "books" as being modded.

Right then, I should have purchased all of the stock components and went to the dealership for a new engine. Read along, and you'll see why.....


Started some investigating. First thought was the turbo. My logic was that it was torquing the turbine and allowing it to leak oil around the seals. After pulling the turbo it did seem to have a concerning amount of lateral play, so I decided to upgrade the turbo(basically the same price as purchased a new factory unit) because I was going after more power eventually anyway.

Upon installing the turbo I decided I needed to get a tune to support the quicker and more robust boost of the Blouch 18g. The tuner recommended AEM drop in 320 pump, 1000cc ID injectors, Cobb FPR w/ vacuum source relocation, and 3port ebcs. Installed all of the components and the tuner provided a base map(map from another 2012 sti he tuned w/ same components) without without any restrictions(fuel cut, rpm limits, etc.)

While I had no plans of doing full boost runs without a true tune for my EXACT car, I quickly noticed that it would hit a heavy amount of stumble starting around 5 psi of boost. I decided to inquire about this with the tuner, and was extremely adamant that the tune would not cause this in any way.

I proceeded to go in for the $500 dollar tune, already having paid my $100 deposit. Once I arrived, the informed me that they require a PDI(pre dyno inspection) before doing the actually tuning. This would be an additional $165(first time this addt. cost was mentioned), but seemed like a logical idea to check the health of the engine.

Not 15min into the full inspection, they brought me into the shop for some bad news. Told me that cylinder #1 had low compression, even though they were testing Cyl. #2(guess they misspoke?). So they shut every thing down, told me I would need a new short block due to ring land failure. Wanted $10,000 if they put it in. Obviously I wasn't going that route.

I purchased all of the gaskets, drop in wiseco forged pistons, new timing belt kit, stage 3 clutch, gem head gaskets, etc. I rebuilt the engine. Babied it for 500 miles, then changed the oil and was ready for testing. Confirmed with the tuner that the map was fully functioning and would be a "good" tune, while not "perfect", and he said it should be safe and run as intended.

*****important added note*****

occasionally, if i did a 1 through 2nd gear WOT pull, the car would log ZERO misfire until I got midway through the RPM in 3rd gear.


The misfire was still there after the rebuild and testing. Primarily cylinder #1, then #3, then #2, and finally #4. When you read roughness it would be, for example, #1-19 counts, #2- 8 counts, #3- 2 counts, #4-0 counts misfire. (somewhere along the line i was think fuel starvation due to series fueling)

I swapped the coils from cylinder #1 to #2. Problem didn't follow.

I swapped the injectors from cylinder #1 to #4. Problem didn't follow.

I checked voltage and continuity from injectors, coils, cam/crank position sensors, knock sensor, OCV sensors, and TGV sensors plugs all the way back to the ECU. Zero issues.

I built my own boost leak tester($15 from Home Depot), while it would only hold 5psi, it was leaking on the turbo inlet only, so I ruled that out...

I replaced the spark plugs with NGK one step colder. Didn't change a thing.

Finally used the OBD-II to check codes. While the Cobb wasn't showing anything except misfire codes. The OBD pulled P2620(throttle pos), P151A(intake runner performance-TGV), P04C1(excessive airflow- EGR), and Mass airflow low voltage/IAT High voltage(Don't remember the actual code number, Pxxx).

The Mass Air flow code would pop back up as soon as I started the car(following ECU reset through the AP, and pulling the NEG from the battery), 3 times in a row. So I concluded it was time for a new MAF sensor. Once it came in, I did one last OBD check, and Ill be damned if it hadn't thrown the code again. Decided it wouldn't hurt to have a back-up, so I installed the new MAF. No Change.

After giving the car enough time after a relearn phase, it started pulling back fuel in the AF Learning #1 AP parameter. After 3 days of driving, it would be -10 to -15% at idle and all the way up to -15 to -25% while cruising the interstate 75mph. Really got me thinking more about the 02 sensor.

Then started thinking that all of the oil running through the exhaust from the blown ring land could have caused the front O2 sensor to be oil coated and not reading properly. It's a consumable product, justified the purchase, ordered and installed..... No change.

At this point, I'm thinking. " ..l., it " , Sorry figured Id get that middle finger in here somewhere. lol

So...... Finally.... As a last ditch effort, I decided to unhook the reference line going to Fuel Pressure Regulator on the new Cobb Regulator line upgrade, and plugging the reference line with a bolt on the side of the road.

Instantaneously the car ran phenomenal strong with ZERO misfire, NO reduction in DAM, NO Knock.........

....I agree, long ago, I should have already test the fuel pressure. I had already installed a second AEM pump, thinking the first was bad. I even doubled up the Oring on the pump to ensure no pressure leakage in the pump housing. Also, I was over spending money by this point and the Cobb Fuel Sensor Adapter isn't cheap, an I didn't want to splice in a test port on a $300 Cobb pressure regulator kit....


Obviously a FPR allows more fuel to flow back to the tank under vacuum, thus having a lower fuel pressure. Also, under boost it pushes the diaphragm and increases pressure by limiting return flow to the tank, BUT... what does this mean for me?

I know my car thought it was OVERLY rich before, due to pulling back fuel in Learning #1, but now runs better on the lack of increased pressure from boost pushing on the FPR diaphragm.

If the FPR wasn't working at all, the car should be rich at idle(vacuum) and lean at WOT(boost), correct?

Why would unhooking the vacuum reference make my cars driving performance flawless(except too rich at idle, still pulls fuel there)? Doesn't seem to make sense that it would run perfectly with no misfires while not increasing or decreasing fuel pressure.

Is this a bad tune or a bad FPR?

If it is a bad FPR, does this mean, that the whole time I have been actually have a lean misfire in WOT? All the while, I have been assuming a rich misfire since it pulls fuel everywhere that is monitored and backfires sometimes. My air fuel ratio on my logs, all the way, to and through the misfire, has always read 11.03(assuming this is the lowest it will read). If, in fact, I was going lean under 5-10 psi boost, then why would the AFR still read rich?

As mentioned, a little ways up, occasionally I could have a "roughness" misfire free 1st-2nd WOT pull. Is it even possible to have a somewhat malfunctioning FPR?


Observations while logging without the Vacuum reference hooked to the FPR:

--14.5-14.7 at idle and cruising (obviously the 02 sensor is making the necessary adjustments since the learning show -10%)

--100% access pos and 100% Throttle position(before i couldn't get to 100% ever)
-- No learning or corrections at all
--DAM is at 1 across the board
--Knock is 0 across the board
-- Zero roughness on cylinders 1,2,3,&4
--Left and Right EX and IN cams within 1 degree of each other.

Odd changes:
--Target boost:14 vs. Actual Boost:18.2
--CL fuel:13.14 vs. AFR: 11.03

ANY INSIGHT WILL BE GREATLY APPRECIATED! This has been a huge stress and I need this distraction to finally be over. I have a business, girlfriend, 2 girls of my own, and 4 "step" daughters that also need this distraction gone.

Thanks again guys for any assistance,ideas, or insights provided!
1 - 2 of 2 Posts