Engine running rich -- AF learning D -(12%) to -(15%)!

I need help troubleshooting my latest saga, slowly been fixing leaks, both intake and exhaust... and finally believe I'm good, then I notice my LTFT seems to remain negative all the time, and my STFT swings a lot (always has), but it too is more negative. I've got logs but wanted to be sure I'm "logging" the right parameters to help with the troubleshooting.

 

My AF learning is almost all negative (B is the best looking one, but I believe that's only due to the "learning" not being complete.) I can reset the computer and my AF Learning D will be ~ -(9%) after only 20min of driving! I'm worried this could damage my engine due to my daily commute. I (luckily some would say) have a wide open, fast, daily commute - often at or even above 3500RPM in 5th. During which the ECM is in open loop... and if I understand right, pulling the AF Learning D trims. So the engine runs like this for 20-30mins twice each day and I've noted the AF Learning D actually creeps higher as high as -(16%.)

 

Trying some self help, I disconnected each injector in turn with some surprising results. Each test run for about 20 minutes allowing the ECM to settle into a state where the STFT cycled a little ~ .73% between positive and negative.

 

AF Learning D only --

Prior to starting the test the reading was -(7%)

Cyl #1 9.38%

--> between test (all 4 reconnected) 0%

Cyl #2 8.6%

--> between test 0%

Cyl #3 17.19%

--> between test 2.34%

Cyl #4 15.6%

 

After the test the ECM settled into -(1.56%)... but that was short lived, soon as I started driving it went back up.

 

So I would expect problems correlating to respective banks, but this is "front" and "back" of the engine. I was trying to see if my injectors were leaking or just borked -- I don't believe the results are conclusive (unless TWO are leaking.) What should I do to protect my engine.

Are you sure you don't have a boost leak? did you check the throttle body hose? could be bulged if it is stock. Happened to me before.

 

see http://legacygt.com/forums/showthread.php/pegged-negative-af-learning-all-ranges-no-knock-iam-1-225843.html

Thanks for the tip, the throttle hose is a Mishimoto about 3 yrs old. I doubt it could "bulge", but I did check the clamp tightness, which I tightened a little by hand (8mm screw/bit driver.) It wasn't loose but I was able to give it a couple cranks.

 

Without reseting the ECM I drove in this morning just watching the trims. My AF Learning D (AF/L-D) started at -(11.4%), by the time I got to work it was -(12.5%)

 

I reset the ECM this morning and will observe the drive home. Also I believe the fuel level may be a factor. Overall I'd say the car runs worse when approaching empty and much better just after fueling (although the initial start up runs so rich the STFT pegs @ -(25%), and it stumbles to almost dying.)

 

Also plan to log the following parameters for troubleshooting, hope they're right:

 

A/F Correction 1 (%)

A/F Learning 1 (%)

A/F Sens 1 Ratio

Boost

Calculated Load (g/rev)

Dynamic Advance Multiplier (DAM)

Dynamic Advance

Feedback Knock

Fine Knock Learn

Ignition Timing

Injector Duty Cycle (%)

MAF (g/s)

MAF Voltage

RPM

Throttle Position (%)

Wastegate Duty (%)

Yeah, I doubt the mishimoto is leaking. How about your intercooler? still stock? did you 'bulletproof' it?

I would do a leakdown test..... I was chasing the dragon for a year until I finally had to accept that it was internal engine stuff. Is your car burning any oil? If it is then it could be a valve.

If you think you have a boost leak, easiest test is to plug the intake with something, take the bpv return hose and blow into to pressurize and listen for leaks. 90% of the time you'll hear em with that test if you have one.

 

Also, you should considering getting an etune =)

 

-Brian

Yeah, I doubt the mishimoto is leaking. How about your intercooler? still stock? did you 'bulletproof' it?

 

Yes I did, chased that demon last year... OE TMIC was clearly leaking, dealership "cleaned" the residue off of it and said they couldn't reproduce the "leak" (warranty.) How can you respond to that ... bought the Mishimoto TMIC-

 

I would do a leakdown test..... I was chasing the dragon for a year until I finally had to accept that it was internal engine stuff. Is your car burning any oil? If it is then it could be a valve.

 

This engine has ~ 20k mi, factory short block with machined heads... car has 109k. It doesn't burn a drop of oil (checked on every fillup ~ 4 days!) Run Rotella T6 with oil analysis ~ 5k mi (last report had slight increase in nickle, but no reported fuel in oil.)

 

If you think you have a boost leak, easiest test is to plug the intake with something, take the bpv return hose and blow into to pressurize and listen for leaks. 90% of the time you'll hear em with that test if you have one.

 

Also, you should considering getting an etune =)

 

-Brian

 

Thanks for the troubleshooting tip Brian, wasn't sure that'd be telling unless I clocked the engine so all the valves were closed (if that's even possible.)

 

And a tune is certainly on the table still. I've got mixed feelings about my OTS... I'm not convinced it isn't part of the problem right now. Overall my biggest disappointment in this car is it's unpredictability - sometimes it will go like scat, others it feels like I'm tied to a anchor -- sometimes during the same drive!

Have you tested, and or cleaned the maf sensor or tried replacing? O2 sensor is another suspect.

 

Check this to make sure you haven't missed anything.

 

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

 

And check the pcv system on valve covers, and emission purge valve.

 

If your sure all the other usual suspects are good in terms of vacuum/boost leaks, then I'd start testing your sensors, like the maf, map, and o2 sensors. Test/Inspect your fuel pressure, ocv's, banjo bolts and fuel pressure regulator etc.

Have you tested, and or cleaned the maf sensor or tried replacing? O2 sensor is another suspect.

 

Yeah. How old is your MAF? Could the culprit too.

 

FWIW, last year I was chasing a rich condition just like you forever with my 06. You don't even know how many times I took out that TMIC off/on to check for leaks and stuff. It was driving me nuts .

Then one day, I had the engine running at idle, took my phone with me while starring at the engine bay. I had the btssm app running on the phone with the Air Fuel correct and Air fuel learn variables showing on the screen.

I then decided to 'play' with the MAF wiring harness. So I held it, bent it, kept it bent and lo and behold, I began to see my A/F correction going way positive and correcting my negative A/F learn values!!!!!!!

I did that several times and came to the conclusion that the issue lied in the MAF wiring harness. Can't believe it.

Have you tested, and or cleaned the maf sensor or tried replacing? O2 sensor is another suspect.

 

...

And check the pcv system on valve covers, and emission purge valve.

 

Great list (unfortunately I know this because I've done most of it...) So yes I have cleaned the MAF (carefully), and that made a marked difference... but mostly in the idle. Replaced entire PVC assembly last year, and even updated the O-ring to viton vs. buna. Not betting on the valvecover "breather" tubes, those dang things are hard as a rock and a b*^% to take off. Wouldn't surprise me a bit if they were cracked / damaged in places I can't see or access. Lastly the O2 sensors (both) were just replaced after this condition came up (and they are OE sensor, no knock offs.) <-- that didn't change anything .

 

Yeah. How old is your MAF? Could the culprit too.

 

FWIW, last year I was chasing a rich condition just like you forever with my 06. You don't even know how many times I took out that TMIC off/on to check for leaks and stuff. It was driving me nuts .

Then one day, I had the engine running at idle, took my phone with me while starring at the engine bay. I had the btssm app running on the phone with the Air Fuel correct and Air fuel learn variables showing on the screen.

I then decided to 'play' with the MAF wiring harness. So I held it, bent it, kept it bent and lo and behold, I began to see my A/F correction going way positive and correcting my negative A/F learn values!!!!!!!

I did that several times and came to the conclusion that the issue lied in the MAF wiring harness. Can't believe it.

 

Wow, I'm not the only one who's attended the school of hard knocks! I wasn't expecting the last bit -- a freak'n short! I'd a been ready to play in traffic. So I use Torque and it's STFT/LTFT the same way. Last time I tried vacuum leak detection with starter fluid (well controlled sprays) but came up empty. My MAF looks to be original... do these things just crap out over time?

 

Looks like I'll be doing a wiggle test and more research on the emission valves since I suspect the car runs different when full vs. empty.

 

I wish I could find some folks who could help me hook my scope up to the (fill in the blank), injectors, MAF, throttle... I know it's the tool that would show me the real deal, but I can't get it dialed in to show me the waveform. Hopefully I'll have time to upload the log from yesterday and today, after the ECU reset it quickly leveled out to -(5.4%) LTFT, but this morning it jumped to -(11.4%)... and I mean jumped -- like between a couple red light lengths, and it's responding like a Geo Metro at speed on the hwy... very disconcerting.

Yeah. How old is your MAF? Could the culprit too.

 

FWIW, last year I was chasing a rich condition just like you forever with my 06. You don't even know how many times I took out that TMIC off/on to check for leaks and stuff. It was driving me nuts .

Then one day, I had the engine running at idle, took my phone with me while starring at the engine bay. I had the btssm app running on the phone with the Air Fuel correct and Air fuel learn variables showing on the screen.

I then decided to 'play' with the MAF wiring harness. So I held it, bent it, kept it bent and lo and behold, I began to see my A/F correction going way positive and correcting my negative A/F learn values!!!!!!!

I did that several times and came to the conclusion that the issue lied in the MAF wiring harness. Can't believe it.

 

 

Its true, when i first got the car, my first mods were uppipe and downpipe, and I didn't use the almighty loctite on the bung hole on uppipe..... needless to say, it came out or rather shot out and exhaust fumes quickly burnt my maf wire causing it to short the ecu fuse. I repaired the wire and missed the blown fuse, causing me to spend 180 bucks for dealership to figure that out.

If you have no cels, no leaks , not burning oil, fuel pressure is good, spark is good. Basically everything is doing what it should be then it looks like a valve, but does a bad valve cause a rich situation.... . it would have to be an intake valve, because you aren't seeing any smoke or burning any oil.

 

When was the last time you had your valves lashed?

I think if there were a valve issue in a given cylinder, you would see a noticeable roughness count at idle for that cylinder. OP, have you logged roughness count per cylinder at idle or when driving? The value should be zero at all times. A value higher than 30-40 will trigger a misfire cel.

I'd also replace the valve cover pcv hoses if they are brittled...... that system is directly connected to ur manifold pressure, which effects your fuel trims big time!!.... the fact you said it means u should change it lol!!! If it "wouldn't surprise you" then you should change them!!!!

So the MAF wiring problem doesn't seem as odd knowing about the exhaust leak... makes this amazing or folks have trouble on the horizon for their Gen 4's.

 

Got home and couldn't help myself... wire wiggle'n is easy. Here was the result;

 

 

I haven't seen a positive LTFT in over 2 months...STFT only went positive on throttle blips/pulloff's. Generally the STFT -(5%) with LTFT -(12 to 15%). Until tonight-- WIN!, post this thread and a couple days later the nail is hit on the head! I'll snap a pic of what the harness looks like-- doesn't look defective to me. Those trims changed, along with the idle, as soon as I wiggled/straightened the first wire. I owe you a beer sir This site rocks-

That's awesome. Even in my case, the wiring harness looks just fine. Honestly, I have not changed it, I just 'bent' it and got it stuck in a way where my trims have been pretty stable. It's been 7 months now. It could be a poor ground or something.

so basically, you wiggle the wire and the values change?

It's not really wiggling. It's more like you bend the wiring harness to a certain angle for instance, and hold it at that position while looking at your live A/F correction data with the engine running.

Yes, I said wiggle, but they don't really wiggle. Moving them made a huge difference. I can't be certain but it may have been caused by the IAT... pin out as connected was one of its wires (from the passenger side.).

 

Crazy as it is, I'm now running positive LTFT as much as I was negative before 10.4%. At least the AFR D is sane, swings between 0 and -(2.4%).

Having the exact same issues, I'll have to try this. Thank you; you're all gentlemen and scholars!