2013 WRX AF Learning 1 at +25% at idle

[muertos]

Background Information
Within the past week, I noticed my car's power would drop noticeably once I hit around 5,6k RPM, typically in third gear pulls. There are no check engine lights.

Next, I noticed the AF Learning 1 value was at +25% while at idle. It seemed to always stay at this value. So this tells me the car's A/F ratio is running rich.

I removed and cleaned the MAF and MAP sensors. I did originally clean the MAF with isopropyl, so it might be busted now, but did buy the MAF cleaner and used that. I reset the ECU, ran the car, and AF Learning 1 starts at 0 then slowly creeps up to 25%.

I replaced the upstream A/F ratio sensor and the downstream O2 sensor. The downstream O2 sensor was black with carbon. The upstream one did not look terrible, but was a bitch to remove. Using OEM Denso parts for both.

Throughout this, the ECU was reset via the Accessport several times.

Current Issue
With these changes performed, the car seems to idle okay and has come up to operating temperature. I drove it around the block and it did not seem to like that. I'd come to a stop and it felt as if it would stall and throttle response seems delayed a second or so. Next, I turned it on again, and now it idles pretty rough.

Datalogs:
I logged idle data for ~36 seconds this morning: https://docs.google.com/spreadsheets...it?usp=sharing
Another log, mostly at idle, but also moved the car forward and back: https://docs.google.com/spreadsheets...it?usp=sharing

This is my first post. I tried following the advice in similar posts, but feel a bit stuck. I'm not sure if there's a good resource for troubleshooting the fuel systems for 2013 WRX EJ25 motors. I've found loose information on the matter and have been piecing things together.

[Router Rob]

Check for vac/bost leaks.

[enduroxt]

When the A/F corrections are positive, that means the engine is running lean, so the ECU compensates by sending more fuel (here >25%). You must have some major vacuum leak somewhere between the MAF and the turbo. Carefully check for this first and don't push the car.

[muertos]

Thanks for the insight! It seems the first thing I should have done is test for boost/vacuum leaks.

I rigged up a boost leak tester. I clamped off 3 vacuum lines I could find, in the same location as the MAP sensor. I think there's a couple more I need to clamp off on this EJ25 but I can't quite tell. I traced one thicker one leading from the intake manifold that splits to both sides of the engine block. I would post pictures here, but the NASIOC image website won't send me my registration email...

[muertos]

Circled in yellow are what I clamped off, question mark is what splits to both sides of the engine
https://imgur.com/a/WtDpjUF

[thewraith420]

Originally Posted by enduroxt

When the A/F corrections are positive, that means the engine is running lean, so the ECU compensates by sending more fuel (here >25%). You must have some major vacuum leak somewhere between the MAF and the turbo. Carefully check for this first and don't push the car.

blow some smoke into a vacuum line I used the line for the bypass valve and also cover up your intake filter make sure nothing comes out of it

[muertos]

Originally Posted by thewraith420

blow some smoke into a vacuum line I used the line for the bypass valve and also cover up your intake filter make sure nothing comes out of it

Just performed a quick test with the smoke machine. I blew it into the bypass valve vacuum line. Blew smoke for about 2,3 minutes. Did not see any obvious smoke coming out of anywhere (prolly gonna test at night too). The pressure gauge reads 0 psi the whole time, so the system is not holding pressure still.. which seems like I have a huge leak somewhere if that's the case. I'm just afraid to push the smoke machine too far with pressure

[muertos]

For what it's worth, I asked ChatGPT, "How do I fix a boost or vacuum leak in my 2013 Subaru WRX?" and it replied with:

"There are a few potential causes of a boost or vacuum leak in a 2013 Subaru WRX, including a damaged or disconnected hose, a faulty or stuck wastegate or blow-off valve, or a leaking turbocharger. To fix the issue, you will need to diagnose the specific cause of the leak. Here are some steps you can take to try to diagnose and fix a boost or vacuum leak:

Check all of the hoses and connections in the engine bay to make sure that they are securely attached and not damaged. Replace any hoses or connections that appear to be damaged or disconnected.

Inspect the wastegate and blow-off valve to make sure that they are functioning properly and are not stuck open.

Check the turbocharger for signs of leakage, such as oil or coolant stains. If the turbocharger is leaking, it will need to be repaired or replaced.

If your WRX is equipped with an intercooler, check the intercooler hoses for any signs of leakage.

If you are unable to locate the source of the leak or are unsure how to properly diagnose and repair it, it is recommended that you take your vehicle to a qualified mechanic.

It is always recommended to consult your car's manual or consult a professional mechanic to make sure you are doing the right thing for your specific vehicle."

[muertos]

For what it's worth, I asked ChatGPT, "How do I diagnose a vacuum leak in my 2013 Subaru WRX with compressed air?" ->

Locate the vacuum lines: Locate all the vacuum lines that run through your vehicle, including those in the intake system, PCV valve, and brake booster.

Use compressed air: Using compressed air, blow air into each vacuum line one at a time. If you hear a whistling sound coming from the line, it's an indication of a leak.

Inspect the vacuum lines: Carefully inspect the vacuum line for any cracks or breaks. Replace or repair any damaged lines.

Check the PCV valve: The PCV valve is responsible for maintaining the proper pressure in the crankcase. If it is malfunctioning, it can cause a vacuum leak.

Check the intake manifold gaskets: The intake manifold gaskets seal the area between the intake manifold and the engine block. If they are damaged or leaking, it can cause a vacuum leak.

Check the throttle body gasket: The throttle body gasket seals the area between the throttle body and the intake manifold. If it is damaged or leaking, it can cause a vacuum leak.

Check the brake booster: The brake booster uses engine vacuum to assist with braking. If it is malfunctioning, it can cause a vacuum leak.

Take it to a professional: If you are unable to locate the source of the vacuum leak, it is best to take your vehicle to a professional mechanic for further diagnosis and repair.

[muertos]

For the full series of questions I asked about boost/vacuum leaks and how to diagnose and fix them: https://imgur.com/gallery/JglgRc5

[enduroxt]

So, did you finish your boost leak test? Technically, it is not a boost leak test you need to do since it appears you are not suffering from a boost leak. Boost leaks results in negative fuel correction, not positive. Additionally, boost leaks occur post IC. Vaccum leaks are pre IC.
Since you have a 2013 WRX (similar engine as mine), there is an easy test that can help identify obvious vaccum leaks: it is in the first post of this thread.

[muertos]

I did, the primary issue I had with that was I missed blocking off the crankcase breather hose. Found a leak at about 5 psi. I had the oil cap off. I found out about the crankcase breather hose through https://www.wrxforums.com/threads/bo...ak-test.19570/

Once I blocked that off, I used the soap trick and found bubbles at the BPV hose connection. It seems like this could be the source of the issue from my understanding, because that hose is connected to the intake, which should be under vacuum, and any extra holes would cause the vacuum to be off. It was mounted on there improperly so I fixed that. It's been raining since so will likely put the car back together tomorrow.

[muertos]

Hey, thanks for sharing that link! That's probably the most informative one I've come across. Just got into using the Accessport to datalog. It was super simple to get going. It's nice that post shares the important values to check and provide in the datalogs. Once again, thanks!

[enduroxt]

So, did you finish your boost leak test? Technically, it is not a boost leak test you need to do since it appears you are not suffering from a boost leak. Boost leaks results in negative fuel correction, not positive. Additionally, boost leaks occur post IC. Vaccum leaks are pre IC.
Since you have a 2013 WRX (similar engine as mine), there is an easy test that can help identify obvious vaccum leaks: it is in the first post of this thread.

[enduroxt]

Sure. So yeah, keep an eye on your fuel trims and see if they go back to zero (or very close).

[muertos]

I put everything back together and the car runs the same at idle. I let the car warm up to operating temperature then made a datalog. Here's that datalog recording the values mentioned in the LegacyGT.com thread -> https://docs.google.com/spreadsheets...it?usp=sharing

There's 2 tabs to this sheet. Chart1 is a graph of some of measured values.

[enduroxt]

I'll check it. But to be clear, did you perform the simple vacuum leak test suggested in that thread? (i.e. block the intake hose with a spray paint bottle cap, and use your lungs to blow through the bpv hose and listen for leaks).

[enduroxt]

I just looked at your log. did you just reset the ECU? Because your AF learning idle trim is zero. Now I saw that the fuel correct went to +25% but that was when you gently pressed the throttle? The RPM was not at idle (should be around 750 rpm or so). Also, what was the coolant temp? Was the engine fully warmed up?

Essentially, you want to check the AF learning at idle (without pressing the throttle pedal), and at operating temp.

[enduroxt]

Deleted. double post..

[muertos]

edit.. double post!

[muertos]

I reset the ECU through the Accessport. I did not press the throttle at all during the previous data log run. And my bad, the car was not up to running temperature last time.

I performed another datalog. This time, I started recording once I hit 174F coolant temperature. I did not press the throttle. I added a couple new tabs to the sheet: datalog-idle and datalog-idle-charts -- I'm curious what you think. Thanks for looking at the logs

[muertos]

(removed double post) Forum moderators/Webmaster: Anything you guys can do about double posts? Seems it's not just me that's having this issue. I clicked the button to reply once, I swear lol

[muertos]

Added another tab to the sheet: datalog-1st-gear-small-pull and another tab for its charts

This time, I drove the car in first gear about a block down the road. The car feels so bad to drive lol.

[enduroxt]

OK. I saw it. The learning is still around zero but I am guessing it will soon catch up to +25%. How were your other learned fuel trims (B,C,D) before the reset?

And did you ever do that simple vacuum test mentioned above?

[enduroxt]

To me, there has to be a fairly obvious vacuum leak somewhere.