LTFT Negative at Idle, Positive with Revs. MAF?

[ShaunTheSheep]

I have a Subaru 2.5 Turbo in an '06 Forester and a very basic OBD scanner. The engine's misfiring slightly at idle once warmed up. I've never noticed hesitation when driving and it's smooth with WOT.

This doesn't sound like a spark or timing issue so i'm thinking a fuelling or air issue.

I plotted long term fuel trim, at idle LTFT is down about -18%. At 2500rpm with no load, LTFT goes to +5%. The MAF is reading 15g/s at idle which i believe is high, should be around 3g/s? It only gets up to around 160g/s at WOT which i understand is low, should be at least 200g/s. AIUI this points to a dirty MAF sensor and supposedly this is a well known issue on these engines. There's no aftermarket air filter or anything like that.

I'm going to unscrew the MAF from the air ducting and will give it a clean (assuming it is a 2 resistor type) and see how that changes the numbers.

What else should i be checking?

[ChrisPackit]

Not entirely sure on a 06 Subaru, but I used to have loads of issues with coil packs on Imprezas, though I know later cars had coil and plug wires. If it has the former, then I'd look at that.

[ShaunTheSheep]

Looks like COPS on this engine so fingers crossed they're ok. There's no missing under load, it's really smooth throughout the rev range.

Have given the maf a clean. Before:



And after some brake cleaner and a delicate brushing:



Will give it a bit to be sure it's fully dry then see how that works.

[ShaunTheSheep]

After 30 miles tonight its running better at warm idle.

Misfire is not totally cured yet and LTFT is -12%. Better but not fixed.

I was hoping with the amount of crud removed from the resistors thatd be the cure!

[ShaunTheSheep]

Next up, an italian tune up. Hope its not a partially blocked cat or something like that stymying airflow, but id expect more symptoms in that case.

[Pumaracing]

For the current fueling concerns why are you not looking at the STFT rather than the LTFT?

[ch427]

previous research tells me id expect to see maf at idle between 4 and 7 g/s and at wot works out at 0.8 times your engine horsepower so roughly 176 g/s if your around 220 hp.

[Pumaracing]

Air consumption at idle is primarily proportional to engine size and hence internal friction to be overcome and works out at about 1.6 g/s per litre of engine size so circa 4 g/s for your engine as the previous poster indicated.

I also agree with the previous poster that air consumption at WOT in g/s at an A/F ratio of about 13 is circa 80% of the power output in bhp.

Rather well done previous poster. A welcome addition to the signal to noise ratio in here.

[ShaunTheSheep]

Pumaracing said:

For the current fueling concerns why are you not looking at the STFT rather than the LTFT?

It was hovering around +/- 5% before the MAF clean. Immediately after the MAF clean it was crazy positive, like 30+ but since i didn't clear the long term memory i'm assuming that was just the ECU responding to the MAF's new readings.

Since the 30 mile run it's sitting at +5% when warm giving a total fuelling change of approx -7 (-12 LTFT / +5 STFT).

AIUI fuel trim is a function of MAF & Intake Temp, O2, Coolant Temp. That's all i've been able to find so far. I can't find a definitive answer for this ECU's FT calculations either.

Intake temp looks ball park correct, i'm assuming the few degrees above ambient are because of engine heat or something like that.

I can't graph the o2 signal on my OBD and i don't have an oscope to be able to check its switching speed.

My book says i should be taking a look at the MAF signal wire when momentarily flooring the throttle, there's a particular trace i should be seeing, a sharp peak followed by a 45 degree rise to 4.2v approx in the diagram (may be different on my car i guess).

Coolant temp is reading 91 degrees C according to OBD, i don't know the thermostat temp on this engine but that sounds ballpark correct to me.

ch427 said:

previous research tells me id expect to see maf at idle between 4 and 7 g/s and at wot works out at 0.8 times your engine horsepower so roughly 176 g/s if your around 220 hp.

That's handy to know

Appreciate the help.

[ch427]

may be worth hooking up a voltmeter to the sensor and watching the voltage rise, cant see how snapping the throttle open will help as its bound to spike initially.
If the base voltage is too high then maybe that is your culprit, is the sensor a 3 wire?

[ShaunTheSheep]

Seems to be ok now. Maybe the 30 miles last night wasn't enough or maybe tonight's exercise dusted off some cobwebs.

Either way, cant provoke misfiring now. LTFT is still slightly rich at -5. STFT fluctuates between +/-5

Thanks for the inputs folks.

[ch427]

sounds about right, i think +/- 10% is acceptable.