Persistent blinking cruise and/or P0103 under high load

My Legacy has a problem. 2005 2.5i NON turbo. 114,000mi

 

When going up a long, steep grade, (30 seconds to a minute of steep hill) regardless of gear, at the top of the hill the cruise light will start blinking. It does not go off until the car is turned off.

 

If the hill is long enough, the check engine light will eventually illuminate in conjunction with the blinking cruise light. The code is P0103 - Mass Air Flow (MAF) Circuit High Input.

 

When it starts blinking, the car stinks of burning catalytic converter.

 

Replaced cat. No change

 

Cleaned MAF, no change. Replaced MAF, new Subaru OEM air filter, no change. Resealed the exhaust with new gaskets, no change, stink still persists somehow.

 

Tried 93 octane, no change.

 

Is it possible one of my O2 sensors are on the way out, but aren't throwing a code? The wiring to the MAF is in perfect shape, and the code only gets thrown on hills or if the car is under high load. (WOT triggers it too if I'm floored for long enough)

 

Since the O2 sensors are the only thing I haven't changed, that's what I'm leaning towards, but I hate throwing parts at the problem.

 

If I were to datalog with my laptop, what parameters should I look for to pinpoint the problem? I have the VAG-COM cable and necessary software.

 

Thank you for any help!!

Have you done anything to the intake in front of the MAF? Is it all there/normal filter/etc?

Brand new Subaru OEM Air filter, air box is intact with resonator box and factory air duct. Forgot to mention that.

Is it possible this is a fuel delivery problem triggering a code that is otherwise unrelated?

 

Does a lean or rich condition cause a cat to smell like burning?

 

I feel that if it was a fuel mixture issue, it would be throwing a different code, so I'm stumped.

 

Fuel pump? FPR? Fuel filter?

burning cat usually means rich.

 

the maf sensor is 3 wires ground 5 volt and signal.

the code you describe is when the signal is becoming greater than what ever magic number of volts that subaru put in there. could be like some thing over 4.8 volts or what ever.

 

the normal voltage output from the maf during driving is typically lower than that. between .8 and 4

 

is the maf installed backwards ? i don't think its bolt holes allow that though.

 

but of coarse do some data logging if you can. and make the problem happen.

 

maf volts,

 

front o2 sensor,

 

rear o2 sesnor,

 

rpm,

 

coolant,

 

air intake temperature

 

throttle angle,

if you have vag com ? get ecu flash and just disable the troubble code ! and the cruse light will stop flashing too.

 

check the grounds by the left side of the engine. there is a aluminum thing that pops put with a few grounds screwed to it .. take it apart and clean it.

Grounding problems for the ECU or between engine and chassis.

Will go out and check the grounds now, thanks guys! If that doesn't work, I'll try to datalog the parameters you outlined frank.

 

The MAF only goes in one way I believe.

 

I should note that the car runs 100% fine regardless of this issue. Very odd. Just stinks!!

if you have vag com ? get ecu flash and just disable the troubble code ! and the cruse light will stop flashing too.

 

check the grounds by the left side of the engine. there is a aluminum thing that pops put with a few grounds screwed to it .. take it apart and clean it.

 

I would disable it however the stinky cat leads me to believe the engine isn't running as it should when the problem arises. If there were no symptoms BUT the trouble code, I'd disable it.

 

I can't find an aluminum ground bar or anything like that, just two ground straps that come from the bottom of my cylinder heads and connect to the frame rails. Is that a GT only thing?

 

I just logged my car twice, but as luck would have it, it didn't start flashing the cruise by the time I got to the top of the test hill. It still stunk though. I can replicate it another day. For now, maybe someone can interpret this and discover a problem.

 

Anyone want to read these?

log 1 no cruise light.csv

Log 2.csv

i dont see a higher throttle angle higher than 60% in the first one and 51% in the second one.

 

ideally you would get a log and make the cel come on. and then i can look for a suspect maf value.

 

as for the grounds there on the drivers side( left side) , on the back of the cylinder head.

there are also some strap grounds on each head on the bottom of the engine . they tie the heads to the frame rails. check them . but on my car i ripped one off and never reconnected the other and i never had a problem.

 

the rear o2 sensor is reading steady .7 volts about the same as my 2.5i

the smell your getting . you sure you didn't run over a plastic bag and its stuck to the exhaust and every time you take the hill it gets hot enough to burn some more ?

get a gallon of methal hydrate , find a small vacuume line and suck it through over about 10 min. and see what that does.

 

also you say the cat is been changed ? doesn't it have two ? both ok ?

Thanks for checking it out

 

That's interesting, I believe I floored the accelerator at least once in the first log. The second log I don't remember doing that, I kept a steady pedal around 50%. The TPS readout says I get full throttle plate movement when I'm standing still with the car off.

 

The smell is most definitely exhaust, it's been doing this for years. When I bought the car it was doing it for the previous owner as well.

 

Any other parameters I should log or does that pretty much cover it? It was only 72 degrees out, so maybe I need a hotter day to replicate the CEL. It almost always does it on very hot days.

Yes it was the whole cat pipe section that was changed, with both cats. Made absolutely no difference unfortunately. Stock cats were probably fine.

the airfuel ratios are fine also so that doesn't show any mehanical problems.

 

could be the throttle plate doesn't open fully untill you reach higher rpms. it follows a map of throttle pedal vs rpm.

 

the smell can you decribe it ? sulpher or rotten egg ? i wonder if the mufflers are full of some thing. oil or some residue for what ever happened in its past. maybe the car was or the exasut was submersed in some thing.

 

can you also data log iam ignition advance multiplyer and knock count you could also be getting false knock or knock that retards the timing. and that would make realy hot exhaust temps.

Interesting.

 

The "hot exhaust temps" thought makes sense. The smell is like a hot burning sulfur smell. It smells hot.

 

The catback has been replaced with another stock unit (old one had the resonator removed so I got one that was untouched) with no change. The entire exhaust has been changed, actually. I got a replacement header off of a 2.5RS because my original was warped.

 

So, mechanically, the parts are fine.

 

I'll get you another log either later tonight or tomorrow. Thanks for your help so much.

Got a log for you. At the last 10-15 seconds of it, the cruise light started flashing and it stunk like hell.

 

The IAM Ignition advance multiplier was not an available parameter, so I logged as much as I thought I could for knock related events.

 

When I got in my driveway, I didn't turn the car off and checked for pending codes. The pending code is

 

P1153 Manufacturer Contrl Fuel Air Metering

 

This eventually turns into the P0103 code once the CEL comes on.

romraiderlog_20110810_172114.csv

That rear o2 reading doesn't look good, nor does the front o2 actually. Seems like you're running rich as hell. The MAF voltage looks normal (to my untrained eye), but 0.98 on the rear o2 I think is as high as it reads? Not sure about the 20.67 on the AFR from the front sensor.

i find it curious the last about 1/3 of the data log the rear o2 sensor is displaying .98 or fairly high readings.

 

this would indicate very rich. the very rich would make the exhaust system hot ..

 

the a/f sensor #1 is that the maf ? or the front o2 sensor ? i think its the front o2 sensor. and it seams to be rock steady in its display voltages.

 

i would tend to think that the front o2 sensor is bad . its seams to be sluggish in movements, and causing the system to go rich in its learning.

 

has the sensor been changed already ?

Front O2 reading is in whatever unit the AFR is in, the rear O2 reading is in Volts. Thought I'd clear that up

 

When I'm at 100% throttle, the front O2 reading of ~14 is correct then, right?

 

Yes A/F Sensor #1 should be front O2 sensor. There was nothing listed in romraider as "Front O2" so I guessed.

Sensors are all original, 114k miles.

well at this point i would swap out the front o2 sensor. i don't see anything else.

 

there expensive so shop around a bit.

 

theoretically the rear should always be hovering around .49 volts.

 

and its always on the high side.

 

humm interesting so looks like you did a pull full throttle to 42 mph, then maintained speed at about 40 mph, ath this part throttle area the front 02 sensor is dispalying 20.56 airfuel ratio .. and stayed there to the end of the data log. and you ahd 15% throttle and 0 % throttle and about 5% again. looks like this is when the cel came on.

 

so looks liek the front o2 sensor got hot and stayed there.

and since the front 02 sensor is displaying with in limits its not going to throw a code.

the o2 sensors typically only throw codes related to the heating circuit anyways.

 

and the rear o2 sensor was displaying wicked rich and then lean when the over run kicked in ( injectors stop injecting during deceleration) so the rear appears to be functioning correctly but the font is displaying funky.

14:1 at WOT seems way lean, my guess is it's overfueling because the front O2 is reading bad. The rear o2 is probably reading correctly. Probably a good way to kill the cats, which might be why yours were replaced, but sounds like maybe they missed the actual issue (or they busted the front o2 sensor replacing the cat pipe)

I love logging my little non turbo car.

 

I thought so too, it should be near 12:1ish, correct? Even on my N/A.

 

Thanks guys, I'll track down a front O2 and see what happens. You would think the car would throw a code for this...

 

 

Also, "Knock Signal 1" and 2 in the readouts are both 0, is this correct? Shouldn't the car show knock at some point?

No, should never have any knock. Good luck on that o2 sensor, my N/A has gone through 4 of the buggers. I've never actually logged A/F from the normal o2 sensor, so I actually have no idea what it should read at WOT.

Grab a multimeter and look at check the voltage at your ECU for your MAF, the vacation pix should give you pinouts and specs. My guess is that you have a wire in the harness intermittently shorting to ground.