LTFT vs MPG?

[friendly_jacek]

I think I noticed a correlation here.

My 2000 Subaru was a gas pig, way below EPA and peer group. LTFT -5% to -10%. I spend a lot of time and money swapping parts and couldn't figure the problem.
My 2003 Corolla was better, but very difficult to get EPA numbers and below peer group. LTFT about -5%.
My 2006 RAV4 on mark compared to peer group, just below EPA. LTFT 2-4%.
My 2010 Prius at and exceeds EPA and peer group. LTFT about 5%.

It sort of makes sense as with negative LTFT, gas has to be removed to reach stoch in closed loop and there is too much gas in open loop. The opposite with positive LTFT.

I do remember Subary exhaust stunk a lot on cold engine and I had UOA with up to 5% gas when idling. Corolla had stinky exhaust to a lesser degree, but it would [censored] H2S with lesser quality gas when new (we had no low sulfur gas back then). No such problems in RAV4 or Prius.

 

[gathermewool]

I guess it depends on your driving style. If you spend most of your time in CL, as most do, and your LTFT is within the ECU's capacity, then it shouldn't matter. Whatever is causing the high negative fuel trim needs to be figured out. It could be an over-reading MAF sensor (I.e., it senses too much air, so too much fuel is injected for the actual amount of air, and the O2 sensor is used by the ECU to get the AFR back down to stoich.)

So, if the MAF is wonky, and the ECU is always chasing stoich, then there might be a waste of fuel before stoich is actually achieved.

What does your AFR look like at high idle, fully warmed up?

On the other hand, an intake leak may cause intermittent lean conditions, possibly resulting in high fuel economy, even though the ECU will try its best to add fuel to maintain stoich, so you do have a point... Hmmm

I'm also not sure how it is with NA fuelling tables, by in my STI, OL uses the last cells of the CL LTFT, so if for whatever reason I had a +10% LTFT in CL, just before going into OL, but my OL fuelling would have been -5% off at that load, then I'd be adding 5% too much fuel. This is great for some, since too much fuel is better than too littler, were the LTFT negative vice positive in the above scenario.

So, I think within 5% doesn't matter much and won't affect fuel economy for most drivers, and anything higher should be investigated.

 

[[email protected]]

So a negative fuel trim ADDS fuel? I always thought thats what a positive fuel trim accomplished?

 

[gathermewool]

positive fuel trims indicate that the ECU is adding fuel to counter a leaner than anticipated condition.

 

[gathermewool]

Also, to clarify for those who may be lost in the acronyms (I hate that, but do it all the time...)

LTFT: Long-term fuel trim, meaning how much fuel your ECU is adding or subtracting from the mix to reach stoichiometric air:fuel ratios (14.7:1,) or whatever the ECU calls for at that load/RPM range (not always 14.7:1)

CL: Closed-loop fuelling is where the amount of fuel initially injected is based on the MAF sensor (for a MAF-based system,) but adjusted by the ECU based on how much oxygen is sensed in the exhaust by the O2 sensor. This feedback creates a closed loop that instantly adjusts fuel trims (short-term fuel trim,) as well as adds or subtracts an average amount, as averaged over a longer period of time (see LTFT.)

OL: The O2 sensor is taken out of the loop and fuel is injected based on the what the ECU looks up in its fueling tables for that load and RPM. I believe that some more-sophisticated ECU's will also apply the last known LTFT to the OL fueling scheme.

So, to sum things up:

(+)LTFT: O2 sensor senses more air in the exhaust than called for by the ECU, so more fuel is injected to reach the desired AFR. If you're cruising down the highway, the ECU will want to maintain a stoichiometric AFR of 14.7:1. If it senses 15.5:1, then STFT will immediately go positive to add fuel. After a certain number of cycles, LTFT will tick up positively and STFT will start over. LTFT will creap up over a period of time until the desired AFR is reached. At this point, the LTFT should remain relatively steady for each specific load/RPM cell, and STFT will account for any variances in temperature, sensor error, relative humidity, or anything else that might affect fuel trims in a minor way.

(-)LTFT: 02 snesor senses too little air in the exhaust than called for by the ECU, so less fuel is injected to reach the desired AFR.