Car is running lean on light throttle - any ideas?

[Davezj]

Interesting stuff chris, but when you recalibrated the gauge was it warm weather or a cold damp morning.

[SEAN-NZ]

a dodgy engine temperature sensor could give a similar symptom, couldn't it? ie when engine is cold it actually thinks its hot and doesn't put as much fuel in

dodgy temp sensor should make the ecu think engines cold, because when sensor is cold it has a much higher resistance, when it warms up the resistance drops, in my experience when its stuffed it maintains a high resistance, does drop as it wamrs up, but doesnt get as low as it should have, in my case it caused the car to drink petrol, didnt matter how i drove, also cut when i put foot down a bit to far, was worse when cold than when warm

[fassi1]

Interesting stuff chris, but when you recalibrated the gauge was it warm weather or a cold damp morning.

It was just before Christmas so it was rather cold and damp morning.

[Davezj]

something must have gone wrong with my calibration last time, i am just clutching at straws to find a reason.

but on the plus side rchecckdit thhis morning an it does the rich to normal gradual trasition.
Sorry tim i know this ddoes not help you.

just out on interest tim have you recalibrated you gaauge recently, it might be worth doing.

[TAR]

not calibrated for a while and I will do it however, regardless of whether is reading wrong, its the sudden change in gauge reading which is confusing me. And my plugs are showing signs of running lean as well.

I will re calibrate but I don't think it will make any difference (may just move the readings up or down the scale)

[Davezj]

just one thing to cross off the list of suspect.

[lateshow]

Hey guys this is rather normal and is due to the fact that factory has scaled the injectors so that LTFT maxxes out -> meaning that the scaling IS too big and if you set the car to open loop it doesnt get enought gasoline. I've been messing about with injector scaling. The original value was 400 with my new method and now I'm down to 350 or something and long fuel trim sails between 1-3,5.... And i can say: No more lean spots but car feels lazier. Yes I get about the same AFR i want in open loop but acceleration enrichment gives me a dip to the rich side when i even press my foot a tiny amount down. There has to be some kind of compromise for that but since I will be tuning for etnahol soon (actually flexifuel) I will buy bigger 670cc injectors.

By the way, a middle trim has been discovered! Or a way to use it, I currently am running with middle trim and low trim just as in Evo5. Ceddy/Ralliart have this switched off. The only problem is that no matter i do with the hi -> mid limits, trim wont go off in full throttle.

[TAR]

Lauri - thanks for your advice. I've read it twice now and think I get what you are saying

[fassi1]

Hmmm so we r back to the fuel trims
VR4 has so little adjustment range (-5%, +5%) in comparison to V6 N/A
and truth stays in closed loop only at low load (vacuum) and as soon as is in boost goes in to open loop
so mid load and high load fuel trims are disabled.

[Humpty's Revenge]

Also a quick note

If you are still running your cat the readings from the afr gauge will read richer due to build up of gasses not escaping fast enough

[TAR]

I've currently still got the cat fitted as I forgot to remove it after the last MOT.

I'm seeing the car lean out whilst in vacuum (so long as my boost guage is accurate) and it gets worse as it comes onto boost. A harder press of the throttle will cause it to richen up a bit but as soon as I ift off again it shoots back up to 16. something.
I've also now noticed that occasionally when I start the car if its warm, I'm straight into 17's and it appears to missfire, which must be the timing getting pulled. This only happens occasionally and must be closed loop. I am going to use the narrowband emulation from the wideband to check if my lambda sensor is going faulty.

[Davezj]

tim do you have two O2 sensors fitted?
the wideband and the original factory narrow band.
if you only have the one wideband sensor fittted the narrow band output of the AFR gauge needs to go in to the ECU.

or were you going to check the narrow band output with evoscan to see it bounce up and down from 0-1V.

just ramblinng

[TAR]

Dave - I have 2 lambda sensors, the OEM narrowband one and the wideband one.

I suspect that the OEM narrowband sensor is playing up, so I'm going to use the AFR gauge narrowband emulation to input to the car's ecu.

[Davezj]

aahhh, very clever.

that should give you a result one way or the other.

[mike74]

So when the engine's pulling vacuum you're seeing the AFR shoot up & when you boot it, it goes rich? To me, that sounds like you've got an air leak somewhere

[fassi1]

Having 17's ECU is definitely in open loop.
I had cat on for while after MOT and didn't notice any difference.

[Nick Mann]

I'd thoroughly recommend narrowband emulation anyway. Both the VR4's I have done that to have improved slightly on fuel consumption.

[fassi1]

Which injector scaling does ECU use?
I have:
EVO (Untested) 375cc/min
DSM (Untested) 409cc/min
What does untested mean???

[fassi1]

I was playing with my occasionl lean idle. Checked all the pipes and connections from MAF to the tubes, everything looked fine to me.
Decided to play with injector scaling.
Having DSM scaling 409 changed that to 380, didn't make any difference to STFT and LTFT.
Changed 380 to 359 and still no change at all.
LTFT 4.6875
STFT about 14
why is there no change to STFT and LTFT?
Am I beeing thick?

[Davezj]

i think injector scaling relies on something else in the ecu other than the figure you input.

i can't remeber what it is, but it rings a bell,
how ever i don't know if that bell is reliable or not as i have had quite a lot of guinees tonight.
i am in training for a 2 week holiday in august to southern ireland.