Subaru Outback Forums banner

1 - 20 of 23 Posts

·
Registered
Joined
·
25 Posts
Discussion Starter #1
Hello guys and gals!

After figuring out my cabin heat gremlin with the help of this fine community here, now it's time to tackle another issue that has been driving me mad lately. My 2005 2.5i A/T Outback has been exhibiting unusually high Long Term Fuel Trims (LTFT) at cruise, +12 to +15 at 110-130km/h (70-80mph). It doesn't go high enough to trigger a lean code, but stays consistently high regardless of gas, load or temperature. It always does it, and the trim is always the same at a given speed (i.e +12 at 110kmh) At idle it's pretty normal at +/- 1-2

I have tested and taken care of pretty much all of the usual suspects - the O2 sensor is new OEM, three separate mechanics tested for air leaks, tried two separate known good MAFs, reset the ECU, ran the "idle relearn" voodoo. Valve clearances were done recently, oil and filters changed on time. Nothing makes a difference. The fueling system can also be ruled out, as the car has an LPG system and the trims stay pretty much the same on either fuel.

Any ideas?
 

·
Registered
Joined
·
18,159 Posts
. . . My 2005 2.5i A/T Outback has been exhibiting unusually high Long Term Fuel Trims (LTFT) at cruise, +12 to +15 at 110-130km/h (70-80mph). It doesn't go high enough to trigger a lean code, but stays consistently high regardless of gas, load or temperature. It always does it, and the trim is always the same at a given speed (i.e +12 at 110kmh) At idle it's pretty normal at +/- 1-2

I have tested and taken care of pretty much all of the usual suspects - the O2 sensor is new OEM, three separate mechanics tested for air leaks, tried two separate known good MAFs, reset the ECU, ran the "idle relearn" voodoo. Valve clearances were done recently, oil and filters changed on time. Nothing makes a difference. The fueling system can also be ruled out, as the car has an LPG system and the trims stay pretty much the same on either fuel.

Any ideas?

Not sure, but perhaps my experience might be related.

LTFT is set for ranges of air intake based on the MAF sensor (mass air flow sensor) readings, rather than vehicle speed.

I don't know about your EDM 2005/6, but for my 07 there's four ranges (g/s), 0-4.5, 4.5-10, 10-22, and 22+. The ECU determines the LTFT level based on STFT and as such, the LTFT isn't fixed for each range. This post has two tables that show the MAF g/s and Learn % (LTFT) at the time the table data was recorded, and below the two tables, there's a graph showing how the LTFT in each range had varied over time. Note that the span of LTFTs remains much the same, i.e., the difference between the high level and the low level of LTFT is in the 15% area. Seems interesting that your's is similar.
@matt88-8 had the same characteristics with another 07. See https://www.subaruoutback.org/forums/109-gen-3-2005-2009/446009-btssm-detected-knock.html
 

·
Registered
Joined
·
411 Posts
What are you STFT like? LTFT is pretty much an average of STFT and other stuff. Are you getting any other stuff on the scanner that looks abnormal?

Sent from my SM-G955U using Tapatalk
 

·
Registered
Joined
·
25 Posts
Discussion Starter #4
@plain OM It does seem you have been chasing a similar issue. I haven't found much information about MAF scales on my model but it probably uses a similar logic. I browsed through your thread, and there are certainly some things that I might try. Did you manage to find out what causes it in the end?
@mprosco91 STFT wiggles about +/- 5 when changing throttle or speed, but quickly stabilizes at or around 0 at cruise. No other causes for concern in the parameters that I can monitor with TorquePro. The car drives normal, no hesitation or stutter.
 

·
Registered
03 H6 OBW & 06 WRX Sportwagon
Joined
·
17,619 Posts
could you get high readings while parked? maybe have someone rev the engine until you get high readings, then introduce some starting fluid around vacuum lines and around the intake manifold to help find a leak?
 

·
Registered
Joined
·
18,159 Posts
@plain OM It does seem you have been chasing a similar issue. I haven't found much information about MAF scales on my model but it probably uses a similar logic. I browsed through your thread, and there are certainly some things that I might try. Did you manage to find out what causes it in the end?. . .
No. Also, as I note in the thread, the characteristics haven't changed since it was first noticed back in 2011.

As mentioned, matt88-8 has the same ECU/CALID, with very similar operating characteristics. Although one comparison is hardly enough to confirm, the ECU programming remains a possibility, in other words, that's the way it's designed to work.
 

·
Registered
Joined
·
411 Posts
No. Also, as I note in the thread, the characteristics haven't changed since it was first noticed back in 2011.

Also, as mentioned, matt88-8 has the same ECU/CALID, with very similar operating characteristics. Although one comparison is hardly enough to confirm, the ECU programming remains a possibility, in other words, that's the way it's designed to work.
I was thinking Calid too. There may be an update at the dealer.

Sent from my SM-G955U using Tapatalk
 

·
Registered
Joined
·
25 Posts
Discussion Starter #9
@mprosco91 I had my mechanic check for an ECU update, apparently there was one to correct for uneven idle with AC on but it was already installed on my car. He did install one for the AT though, which improved smoothness in low-load gear changes. I don't know if a full reflash of the ECU might help, it does with my work equipment sometimes (cine cameras & related gear).
 

·
Registered
Joined
·
411 Posts
@mprosco91 I had my mechanic check for an ECU update, apparently there was one to correct for uneven idle with AC on but it was already installed on my car. He did install one for the AT though, which improved smoothness in low-load gear changes. I don't know if a full reflash of the ECU might help, it does with my work equipment sometimes (cine cameras & related gear).
Did you recheck everything after the update? Sometimes updates do other things that's not listed in the description. Anything change?

Sent from my SM-G955U using Tapatalk
 

·
Registered
Joined
·
25 Posts
Discussion Starter #11
@mprosco91 The ECU was not updated as it already had the update installed by the previous owner. They only updated the A/T, it did not make a change in the trims.
 

·
Registered
Joined
·
25 Posts
Discussion Starter #12
@1 Lucky Texan I tried the revving while parked bit and I did get the high readings again, but did not manage to find any leaks. It does, however, at least suggest that it is not something in the drivetrain resulting in increased load on the engine. It is something else I considered as a potential source of the issue.
 

·
Registered
Joined
·
25 Posts
Discussion Starter #13 (Edited)
So today I stumbled upon a couple of threads about grounding problems on older models leading to various sensor and drivability issues, and people had good luck running a couple of extra ground wires. I figured improving ground can't hurt anything, so I ran three wires from the battery negative terminal to the intake manifold and two body ground points. I also sprayed contact cleaner in all the connectors I could get to and visually went over the wiring to check for any fraying or other damage. Unfortunately this does not seem to be the issue as there was no change in the way the car drives or the trim values, just something to cross off the list I guess.
 

·
Registered
Joined
·
25 Posts
Discussion Starter #14
At the moment I am thinking this might be related to a pre O2 sensor exhaust leak I suspect I might have. Although it doesn't seem to make sense for the effect of the leak to be more prominent at high RPM, I can't rule it out. Perhaps it has to do with the exact location of the leak, air pressure at speed and other variables. Anyone dealt with an exhaust leak that has affected O2 data and therefore trims?
 

·
Registered
Joined
·
411 Posts
Do you hear any exhaust leaks? If small enough you may not. However my experience any leak before the cat or sensors is usually pretty loud.

If all else fails the sensor(s) themselves may just be bad. I don't remember if you said you replaced them or not. I've had lean codes before without having it tell me a sensor was bad. But the sensor was the cause of it.
At the moment I am thinking this might be related to a pre O2 sensor exhaust leak I suspect I might have. Although it doesn't seem to make sense for the effect of the leak to be more prominent at high RPM, I can't rule it out. Perhaps it has to do with the exact location of the leak, air pressure at speed and other variables. Anyone dealt with an exhaust leak that has affected O2 data and therefore trims?
Sent from my SM-G955U using Tapatalk
 

·
Registered
Joined
·
25 Posts
Discussion Starter #16
@mprosco91 I can definitely hear it, but I don't know where it is. I will hopefully find out on Saturday when I have it in for AT oil flush.

The front O2 sensor was replaced with a new OEM one about a 8 months ago. I did not monitor the trims before the change, only started doing it afterwards and they have always been high. The rear O2 (post-cat) has not been replaced but diagnosticians that saw it seem to think it behaves as it should. There is always a possibility of the new O2 being faulty, that's why I have one of the local garages on lookout for a used known good one to test with, but it could be a while until they have one in.
 

·
Registered
Joined
·
411 Posts
Was the original sensor throwing codes? If ya still have it swap it back and see what happens.

We had a few cars that would mess up fuel trim at the dealer and the rear o2 was bad. But those were turbo cars. So I'm sure the algorithm in the computer is different.

Keep us posted on where that leak is when ya find it.
@mprosco91 I can definitely hear it, but I don't know where it is. I will hopefully find out on Saturday when I have it in for AT oil flush.

The front O2 sensor was replaced with a new OEM one about a 8 months ago. I did not monitor the trims before the change, only started doing it afterwards and they have always been high. The rear O2 (post-cat) has not been replaced but diagnosticians that saw it seem to think it behaves as it should. There is always a possibility of the new O2 being faulty, that's why I have one of the local garages on lookout for a used known good one to test with, but it could be a while until they have one in.
Sent from my SM-G955U using Tapatalk
 

·
Registered
Joined
·
25 Posts
Discussion Starter #18
@mprosco91 Nope, no O2 sensor codes but it was messing up the fueling, we discovered it once the LPG system was installed. It was going from lean to rich in wide margins at around 2000 RPM i.e. if you tried to hold the engine at 2000 it would pulse from 1900 to 2100 every couple of seconds all on its own. It was barely noticeable on petrol but much worse on LPG.

We swapped the MAF at first, and the problem was gone for a little while. When it came back I went back to the shop, we returned the old MAF and swapped in a new O2. That fueling issue has not returned since, but it made me start monitoring the OBD data through which I discovered the trims abnormality. I don't know if the trims were the same with the old O2 and unfortunately I did not keep it, it would've certainly helped the diagnostics.
 

·
Registered
Joined
·
411 Posts
@mprosco91 Nope, no O2 sensor codes but it was messing up the fueling, we discovered it once the LPG system was installed. It was going from lean to rich in wide margins at around 2000 RPM i.e. if you tried to hold the engine at 2000 it would pulse from 1900 to 2100 every couple of seconds all on its own. It was barely noticeable on petrol but much worse on LPG.

We swapped the MAF at first, and the problem was gone for a little while. When it came back I went back to the shop, we returned the old MAF and swapped in a new O2. That fueling issue has not returned since, but it made me start monitoring the OBD data through which I discovered the trims abnormality. I don't know if the trims were the same with the old O2 and unfortunately I did not keep it, it would've certainly helped the diagnostics.
Have you tried going back to petrol? I'm not familiar with what LPG readings should be compared to what you're getting.

Sent from my SM-G955U using Tapatalk
 

·
Registered
Joined
·
25 Posts
Discussion Starter #20
All the readings and tests are being done on petrol only. I thought running on LPG messes up with the learning procedure after an ECU reset, so I have also tried resetting the ECU and then running the car on petrol only for prolonged periods, but the trims quickly return to the same high levels.

The trims are virtually identical on both fuels, however at the moment I am not concerned about the LPG system. It mimics the petrol fueling with corrections for the different burning characteristics of LPG, so once the petrol trim issue is resolved it is just a matter of recalibrating the LPG computer.
 
1 - 20 of 23 Posts
Top