CEL, symptoms...please help me decide what to do next

[skizix]

Hi folks. Not a new member, but haven't posted for a few years, so I might as well be. Yes, I've done much searches/reading here, but gotten no firm ideas, and hoping someone can pick up on the nuances of my situation.

 

Warning: this is long, but please bear with me (EFI diagnosis is subtle -- trying to give details)...

 

I recently bought a 2002 OBW, 2.5/4cyl, 5MT, 100,000 miles. Mostly quite happy with the car, but something is not quite right.

 

Symptoms: 1) I've got a hesitation on takeoff. Like I give it some juice, it bogs down for just about 1 second, then takes off like it should. 2) I've got an "anti-hesitation" when I let up on the accelerator, transitioning to engine-braking -- like I let up on the accel., it engine-brakes properly for a sec, then surges (minor, but definitely there) slightly for just a second, then settles down to normal. 3) I've got an intermittent jerkiness on low-speed accel. changes. Like, coasting along at low speeds in 1st or 2nd, a careful, gradual push on the accel. pedal will result in a sudden **grab**, and the car will lurch forward; at other times it is a buttery smooth transition. The opposite happens too: going from pedal down slightly at low speeds, gradually to engine-braking mode, the car will jerk abruptly into deceleration, no matter how carefully I feather the accel-pedal. Other times it's buttery smooth. Note that I'm not talking about any involvement of the clutch here -- fully engaged in these #3 scenarios (in scenario #1, I sometimes have to feather the clutch and give more than optimal accel-pedal to compensate for the hesitation). 4) When I'm cruising at a constant speed on a flat road (i.e. not accelerating or decelarating, just holding a very steady throttle), the car pushes and pulls just slightly, as if I were up and down on the pedal, no matter how much I concentrate on holding it steady. 5) Steady acceleration sometimes feels slightly "notchy", i.e. not smooth, all the way through the power band -- not talking hesitation now, but just sorta slightly "grabby". Not a big deal for driveability, but just does not feel quite "right", and seems like a clue.

 

The car drives great on the highway; mileage seems fine (26-27mpg, @ ~66% highway miles). However, stop-and-go traffic is pretty darn annoying at times.

 

Other clues: the problem seems far less prevalent, but not entirely absent, with a cold engine (I do get minor clutch judder then sometimes, but I'm really not so concerned about that -- not my first subaru). Also, when I reset the ECU (done that 3 times now), the problems pretty much go away for a few days, then creep back in, in their intermittent ways. Problems are all amplified with the A/C on (not too surprising, I guess).

 

Also: after having these problems, but with no codes thrown for a couple of months...last week I got a persistent CEL. I was almost psyched, thinking this might tell me what was causing my car to drive crappily, but mechanic (who pulled the code for free, but apparently did not reset ECU) said it was coincidental -- front catalytic conv. code, which according to him should not affect driveability. I asked him if he thought it could be, say, an O2 sensor reading badly but not failing entirely, causing the ECU to infer a bad cat. He says no, it's the cat (but not necessarily trying to sell me -- says I don't need to worry about replacing the cat 'till I want to).

 

Things I've done: today I reset the ECU, and the CEL, went off, and has remained off. Also, previous to that, I've replaced plugs and wires, air filter (dropped in a K&N) and cleaned the IAC (it did not look very gummed up at all, but...). All seemed to help some. I haven't done fuel filter yet (I would, but I can't seem to get the hoses off without damaging them -- may have to pay someone, which is a little embarrassing; any advice on that is appreciated). Plugs and wires definitely helped; I did have an occasional misfire, sans code, I think;felt it a couple of times, and one coil terminal had carbon/corrosion. Plugs looked a bit burnt too.

 

Ok, SO........I'm thinking I've got one of the following: 1) bad/slow, but not dead, O2 sensor(s). 2) Screwed up TPS. Can I clean this thing without wrecking it? Is a new one nasty expensive? I did clean it's elec. connectors, btw -- didn't seem to help. 3) Bad cat(s). 4) OTHER???

 

The cats seem the obvious choice, given the code thrown, and it's what the mechanic recommends. But that's $$$, and somehow I don't quite believe it ('specially since the problems take a brief vacation when I reset the ECU -- gotta be a sensor, not mechanical, right?). I'm thinking of just going ahead and doing new O2 sensor(s) -- starting with front -- but I'll feel like an idiot if that expenditure improves nothing (despite 100k miles, mechanic says he doesn't think that's it, and says no, do not replace O2 sensors pre-emptively; at least if he is wrong, he does seem honest/non-greedy). Same for TPS: bummer to buy a new one if I don't need it (again...CAN I SAFELY CLEAN IT AND IS CLEANING LIKELY TO HELP?).

 

Please let me know if you feel like you know what's going on, and/or what to do first here. My budget is tight (payed cash for this car, and it kinda cleaned me out; my intention was to get the most car I could w/o financing -- need to get my first mortgage later this year and didn't want a loan hanging out there).

 

I am at a loss, apart from blindly replacing sensors.

 

Sorry for such a long post, and...thanks!

[cookie]

first thing I'd do is replace the front ox sensor. I assume that threw code PO 420. My car acted like that too and a complete fix was 10 minutes and $60. I hope you have the same luck.

Front ox sensors are a known weird problem around here.

[sea#3]

Most likely it's the front 02 sensor(front A/F sensor) that is slow or "Lying "to the ECM , It is quite common on the 2000 my and up

 

SEA#3

[skizix]

Ok, so I've spoken to two shop mechanics now, both of whose opinions I trust, and neither believes it is the O2 sensor. They both insist that an O2 sensor that were causing drivability issues would definitely throw a code. Seeing as I've positioned myself to both as ready to purchase one, and they both advised against the purchase, I do believe that they both are confident in that opinion.

 

However, I'm still skeptical, mostly based on evidence on this and other discussion boards. So...has anyone found that a new O2 fixed definite drivability problems, and NOT had an 02 sensor code recorded in their ECU (not necessarily talking about a C.E. light -- some codes are thrown without switching on the light -- but actually no 02S-related code)?

 

Both mechs also believe that my inefficient-cat code/light could not be caused by sickly O2 sensor(s), despite that scenario making sense to me. Anyway, that light has remained off after 3 days and 100+ miles. Would that make sense if my cat were really on it's way out?

 

Anyway, I've decided to hold of on spendy fixes for right now, and try some injector cleaner. Dumped a can of Techron in there yesterday, and will probably follow up with some BG44K in my next tank. Hadn't thought of it before, but it does seem like fouled injectors could be causing all of my symptoms (except the cat CEL).

 

Feedback, anyone?

[cookie]

As I said mine did throw a code, PO 420, usually associated with cats. Several folks have reported driveability problems with Subaru O sensors, but I don't recall if they had a code. Frankly an O2 sensor is supposed to run 100,000 miles and you have done that.

Everybody told me to change my cats but they were several hundred bucks. The O sensor was $60 and I already had over 100,000 miles on it so changing it was just maintenence. If nothing else this has made me happy for a year. My code did not come back for quite a while when I cleared it but third time I changed the sensor.

If the O sensor does not fix everything you have only changed a cheap part you are supposed to change.

[BigMattyD]

I agree with changing the o2 sensor first. It can cause a catalyst failure code. In fact I was recently diagnosing a supposed catalyst problem, and one of the associated faults that could cause that code was a failing o2 sensor.

 

I believe others have reported driveability issues with poor o2 sensors, but I have no expierience of that myself.

 

Just because two mechanics said it could'nt possibly be the o2 sensor doesn't mean it's true. I mean, come on, that's why many of the people on this board work on their own cars...

 

matt

[skizix]

Well, yeah Matt...I can't count the number of times I've found mechanics @ shops to have been wrong (like telling me I need a new clutch when I'd changed it myself a week previous and it was working perfectly...etc.,etc.). That's why I'm here!

 

So it sounds like definitely the front O2S is the one to change first. Let me know if I have this correct: feedback from the front O2S influences the ECU's control of fuel mixture (and timing?); the rear O2S does not influence the ECU's decisions at all, other than to turn on the CEL if it senses the cat not doing it's job. So the front O2S would be the one to change to address drivability, and the rear I assume I would not touch unless the CEL stays on (and I don't suspect a bad cat). Is that about right?

 

BTW, one mechanic suggested a new cat (to fix my CEL/code), but the other suggested an "emissions system flush". So I'm going to go ahead and get a new O2S, and not deal with the cat until further notice, BUT...anyone heard of the cat-flush??? Aparently it's a non-retail product (needs some special applicator tool) made by JB (Justice Brothers). Sounds kind of odd. Mechanic said they have a "75% success rate" fixing bad-cat codes with that kit. Any opinions on that?

 

Thanks folks!

[nipper]

i vote for an o2 sensor. First off we need to know what codes are apparent. Not all codes will turn on the cel if they are of a short duration. They may stay in memory. I dont beleive it is a cat, as they tend to last a very long time. There is a test for the cat. Bang it with your hand. If it rattles its bad (the breakdown of the internal materials will rattle around). Cats are the least likely choice.

When the car is cold it is in an open loop. This means it is running off preprogramed paramiters. Once the car warms up it goes into closed loop, this is where its reading alll the sensors. One of the sensors that are locked out of open loop is the o2 sensor.

K & N filter not a great thing to do. research air filters here, you will see they allow more flow by letting more dirt to go by.

Right now i would drive the car untill the cel goes back on, then pull the codes. If the mechanics never scaned for codes shame on them. If you bought it to them after you cleared the codes they are forgiven.

Also a scanner will read things in real time. you can see exaclty what the o2 sensor, tps and several other sensors are doing in real time. The TPA will have the same sympton at the same point in the throttle, as it is a very simple rheostat.

The o2 sensor tells the car how to adjust the A/F mixture for optimal performance and emissons. Since it is on its way out, it is acting weird. This throws off the AF ratio so you get hesitation, surging and everything else you complained about.

 

nipper

[nipper]

Both mechs also believe that my inefficient-cat code/light could not be caused by sickly O2 sensor(s), despite that scenario making sense to me. Anyway, that light has remained off after 3 days and 100+ miles. Would that make sense if my cat were really on it's way out?

 

?

 

You hit the nail on the head. The second o2 sensor is checking the effeciency of the cat, by comapring to the first one. The first one talks to the ECU. o2 sensors have a life of about 100k miles. Besides even if it is the cats, i still would recomend replaceing the o2 sensors aand the cat at the same time (unless you know they have been replaced once already). So replace the o2 senors. i think the second one can be generic, the first must be OE. Most the time this will fix it, and its much cheaper to replace o2 sensors and have it fix it, then replace the cats and have the problem still be there.

 

nipper

[skizix]

When the car is cold it is in an open loop. This means it is running off preprogramed paramiters. Once the car warms up it goes into closed loop...

Wow, thanks nipper. That is very useful info indeed. The running-ok-while-cold phenom was throwing me off, thinking it might be a bad temp sensor or... Makes sense now. Still wonder why it runs well for some days after resetting the ECU, but I suppose it probably takes a conservative approach to engine management whilst gathering a baseline set of data.

 

The second o2 sensor is checking the effeciency of the cat, by comapring to the first one. The first one talks to the ECU....i still would recomend replaceing the o2 sensors

nipper

Ok, so you've suggested replacing both sensors? But from what else you've said, it sounds like I'll be good to go with just a new front, at least until the CEL lights up again. Correct? That'd be my preference, being a bit short on $$$ right now. Any downside to just doing the front O2S until further notice?

 

Thanks much.

[nipper]

If you are tight for cash, replace front one (the more exp) and cross your fingers, but you have two seperate problems. the running issues (the front o2 sensor) and the cat effeciency code (the rear o2 sensor). It ran fine after you reset the computer cause it took a few days for things to happen again. i bet your sensors arent failed yet, but they are slow/lazy or on the edge.

 

nipper

[BigMattyD]

Nipper: Isn't it possible that the front O2 sensor being out of whack could also cause a catalyst efficiency code??

 

That is, if the ECU is comparing front vs rear O2 readings, and the front is off, but the rear is working normally, it would provide a confusing set of data, triggering the fault..

 

It is still possible that just the front O2 sensor is faulty, in my estimation, but it wouldn't hurt to change both unless the cost of the second sensor would break your budget.

 

Matt

[nipper]

Nipper: Isn't it possible that the front O2 sensor being out of whack could also cause a catalyst efficiency code??

 

That is, if the ECU is comparing front vs rear O2 readings, and the front is off, but the rear is working normally, it would provide a confusing set of data, triggering the fault..

 

It is still possible that just the front O2 sensor is faulty, in my estimation, but it wouldn't hurt to change both unless the cost of the second sensor would break your budget.

 

Matt

in all honesty i'm not sure, it can go either way. My logic is that the cat can handle a wide varity of a/f irregularities and compensate for them without going out of acceptable limits. i should have re worded one of my replies. i think the second is only measuring the cat output, not comparing itself to the first one (i goofed). ild still say replace both, since he isnt geting an o2 sensor failure light, but has all the symptons, but he is getting a failed cat output code.

If i could see the scanner reading the sensor outputs in real time, i would say for sure either way, but i cant.

 

nipper

[skizix]

Hey folks, thanks very much for all the input. I did replace the front O2 sensor, and it has definitely helped. Hesitation on takeoff seems to be gone. Sorry I waited so long to get back, but I wanted to make sure the effect was real, since my driveability issues seemed to go away for a few days after resetting the ECU. The fix seems to be sticking though.

 

So, it definitely helped, and I'm definitely getting close to getting this car dialed in. Still a little funk though...

 

If anyone's still reading: sometimes when I do something abrupt, like suddenly mat the accelerator when I'm coasting along at 2k rpm in 2nd (or if I let up sudenly after hard accel.)...the drivetrain "grabs" a bit, which is fine. What is not so fine is: it grabs, and then bounces along for a few seconds -- not a sharp jerky lurch, so much, as a bouncy lurch. Feels like I'm boinging on the clutch springs, or.... With the A/C on, it is worse, and can go on bouncing for quite a few seconds. Any ideas? I'm thinking maybe tranny mounts. Does that sound likely? How would one assess the condition of tranny mounts? Motor mounts?

[nipper]

You know i dont ever remeber anyone having to repalce a motor mount on these things, and only once a tranny mount.

Is it a fuel type bucking or a mechanical source?

When was the last time the fuel filter was changed?

i'm not really understanding what your trying to describe, but thats just me.

 

 

nipper

[skizix]

nipper, I finally managed to change the fuel filter last weekend -- same wrench session as the O2 sensor (I could not budge the hoses no matter what, last month, then got a great tip: shove a small screwdriver in there to make a gap, and squirt some silicone in there and work it around. Worked like a charm; lubed the new one up too to make it easier next time).

 

Anyway, when I started this thread, I had an intermittent fuel-related-feeling lurch, as described above, but somewhere between the O2 sensor, fuel filter, and some BG44k, that seems gone. That lurch was quite sharp, and only one pulse at a time. This bounce definitely feels mechanical (to me, at least). On sudden accel or decel, it is like the push (or pull) of the drivetrain is set to bouncing -- real springy feel -- and gradually settles down. The direction is back and forth (accel/decel), not up and down. The frequency is about like saying "boing, boing, boing..." casually. If I am smooth on the pedal, it does not do it at all. Happens more easily and takes longer to settle down with A/C on. Does not feel like suspension (e.g. is not induced by hitting a bump; susp. seems nice and damp). The stick (5MT here) does move visibly back and forth while the boing is happening, but not very much. No clunks or anything associated. Just a boingy acceleration when the drivetrain is torqued abruptly.

 

Hope that helps, description-wise. Not so easy to capture in words.

 

So, any thoughts? (and again...THANKS)

[skizix]

Well, crap! Perhaps I re-rang in too soon. I drove the car just now, and experienced one of the symptoms I was having in the first place: the jerk on transition from acceleration to engine-braking and vice-versa. Just a few hours ago I wasn't feeling that, but now it's back. I was cruising down my 25mph street, accelerating up to ~30 after turning onto it, and eased up on the pedal slowly, as I was done accelerating....**lurch** the car decelerates abruptly, chunking me forward in my seat. So I play with it, easing down on the pedal ever so carefully, and then **lurch** it abruptly jerks forward, and then starts accelerating smoothly. I played that little game the rest of the way home, and no matter how light a touch I put into accelerating and decelerating, it jerks hard into go or coast mode.

 

The pisser is: tomorrow when I drive to work, I betcha this will not be the case...then, some time tomorrow or the next day, it'll go back into this mode for a while.

 

This is not the "boing" problem. But it does cause the boing problem. But I don't care so much about the boing problem when this shiity driveability crap is going on. Really sucks in slow-and-go traffic, and is just annoying in general. I really want to like this car.

 

WTF IS GOING ON HERE???

[Cougar]

I would try replacing the TPS. It can cause this kind of thing to happen with dirty contacts. Your total mileage is high enough to also confirm the unit may need to be replaced.

[skizix]

TPS had occurred to me. Only thing is: how to explain that the problem goes away for a few days after an ECU reset? TPS does not seem like something the ECU can substitute a default value for, right? What does the ECU do in response to the TPS signal anyway? Open and close the IAC only? Fuel mixture would mostly be decided by MAP signal, I would think, not TPS. Is that more or less correct?

 

BTW, it does feel like my IAC opens and closes a little late sometimes, but I did clean it, and my idle never drops out, so i'm assuming IAC is functioning ok.

 

Can the TPS be tested? Kind of a spendy part to replace blindly. What, for example, would I look for if I plugged in a probe while driving?

 

Can the cleaned with reasonable ease? I thought I remembered someone saying not to mess with it -- can o' worms. But, being merely a potentiameter (correct?), it should be fairly easy to clean, and it should respond well to cleaning, I'd think.

[Chicobiker]

Hello.

 

I am new to the forum. I posted a question about engine RPM behaviour and shomehow I got looking around and saw this thread.

 

Some of this behaviour is exactly the same as my car! I could not have used a better description for what my car does, particularly #1 and 3. #4 and 5 are happening lately, but it is only a recent thing. I have a 2003 OBS with 110k kms.

 

I always attributed the very brief hesitation to not having enough RPMs for the all wheel drive setup. Never got the problem if I kept the RPMs up high, but it always felt like I shouldn't have to do that. I always assumed the notchy deceleration to be a "feature" of the Subaru. The shifting sure sounds clunky if you don't get it perfect, so I assumed this notchy behaviour to be similarly caused.

 

I'm getting major hesitation now which is annoying me and hopefully the dealer has some luck with it tomorrow.

 

Anyway, I'll keep an eye on this thread to see if you get any revelations. I'm just starting to figure out the mechanics of this car (I'm more of an electrical type!)

 

BTW, what is TPS? Some sort of pressure or throttle sensor? I'll do a search once I post this message.

 

Thanks,

Chicobiker

 

 

Symptoms: 1) I've got a hesitation on takeoff. Like I give it some juice, it bogs down for just about 1 second, then takes off like it should.

 

3) I've got an intermittent jerkiness on low-speed accel. changes. Like, coasting along at low speeds in 1st or 2nd, a careful, gradual push on the accel. pedal will result in a sudden **grab**, and the car will lurch forward; at other times it is a buttery smooth transition. The opposite happens too: going from pedal down slightly at low speeds, gradually to engine-braking mode, the car will jerk abruptly into deceleration, no matter how carefully I feather the accel-pedal. Other times it's buttery smooth. Note that I'm not talking about any involvement of the clutch here -- fully engaged in these #3 scenarios (in scenario #1, I sometimes have to feather the clutch and give more than optimal accel-pedal to compensate for the hesitation).

 

4) When I'm cruising at a constant speed on a flat road (i.e. not accelerating or decelarating, just holding a very steady throttle), the car pushes and pulls just slightly, as if I were up and down on the pedal, no matter how much I concentrate on holding it steady.

 

5) Steady acceleration sometimes feels slightly "notchy", i.e. not smooth, all the way through the power band -- not talking hesitation now, but just sorta slightly "grabby". Not a big deal for driveability, but just does not feel quite "right", and seems like a clue.

 

[nipper]

Hello.

 

I am new to the forum. I posted a question about engine RPM behaviour and shomehow I got looking around and saw this thread.

 

Some of this behaviour is exactly the same as my car! I could not have used a better description for what my car does, particularly #1 and 3. #4 and 5 are happening lately, but it is only a recent thing. I have a 2003 OBS with 110k kms.

 

I always attributed the very brief hesitation to not having enough RPMs for the all wheel drive setup. Never got the problem if I kept the RPMs up high, but it always felt like I shouldn't have to do that. I always assumed the notchy deceleration to be a "feature" of the Subaru. The shifting sure sounds clunky if you don't get it perfect, so I assumed this notchy behaviour to be similarly caused.

 

I'm getting major hesitation now which is annoying me and hopefully the dealer has some luck with it tomorrow.

 

Anyway, I'll keep an eye on this thread to see if you get any revelations. I'm just starting to figure out the mechanics of this car (I'm more of an electrical type!)

 

BTW, what is TPS? Some sort of pressure or throttle sensor? I'll do a search once I post this message.

 

Thanks,

Chicobiker

 

Welcome to subarus and the forum.

 

TPS is a throttle position sensor It is a variable rehostat that tells the car and the transmission throttle position.

 

nipper

[skizix]

Hello.

I'm getting major hesitation now which is annoying me and hopefully the dealer has some luck with it tomorrow.

 

Anyway, I'll keep an eye on this thread to see if you get any revelations...

 

Well, you might think about a new front O2 sensor. It most definitely helped with hesitation on takeoff I was having. That is pretty much gone, and that is a big relief.

 

Anyway...what I've got left is starting to feel like a fuel problem for sure now. New symptom -- probably overlaps with all the others. On mild acceleration at low speeds, it is bucking slightly, like a slight random hesitation. Almost like a misfire, but way to many times not to light the CEL (I'd think anyway). Overall, the car is driving pretty darn well, now, but something is definitely still not quite right.

 

Does this sound like TPS still, or a fuel problem?

 

If the latter...car has a fresh fuel filter, and I've run a can of techron and a can of BG44K , one each in my last two tanks. What else should I be thinking, fuel-delivery wise? All I can think of is fuel pump...? Is that at all likely ('02 OBW, 103k miles), and how would I test it: fuel line pressure alone?

[nipper]

Rarely fuel pump. You need a pressure gauge on the fuel line to see whats going on there, i bet its one of the sensors. Fuel pumps seem to run forever. Knock sensor? This is where my expertise gets a little thin. Have you cleared the puter and reset it? Just because the CEL isnt triggered doesnt mean there arent any stored codes.

 

nipper