Help! My subaru is sick

[Dekonx]

So i have a 1999 subaru legacy outback limited 2.5l 5 speed.

It has recently developed an abnormal problem. The car will start and idle and even drive till about 3k rpm then fall flat on its face losing all power what so ever. The further you press the pedal down the lower the rpms drop. Eventually sounding/feeling like it's running on 2 cylinders while mashing the pedal to the floor. When you let off the pedal it abruptly stalls out and will not restart until you cycle the key several times.

Upon restarting the car it will run for a short period of time till it completely bogs down and dying again.

I changed the fuel filter a few weeks ago when this first happened.

Last night on the way home from work i was driving down the freeway doing 60 hit an off ramp to change routes and it bogged down stalling out loosing all power if i took my foot off the pedal it would decelerate slowly however if i hit the pedal it really dropped rpms in a hurry.

So i was wondering if anyone here had suggestions.

I think that it might be the following.

Tps
Iac
Maf
Fuel injectors
Fuel pump relay
Fuel pump

Coil packs

Now for the twist.

The car is only throwing code p0141 which is the o2 sensor. To which i know that the sensor is broken

So what could cause it to misfire like crazy bog out completely and not throw a cel code. I have cleared the cel codes 2x the only one that comes back is the p0141 so i was thinking that the ecu wasn't shot.

[Dekonx]

I have also added the cam position sensor to list of possible problems the car has 255k on the clock.

[prodriver1]

the sensors are real easy to change out, most are rite on top of the engine and require basic tools and knowledge. get some new sensors all over, thats a lot of miles, change everything, sensors, full tuneup, the works. the p0141 code means the O2 sensor is bad, or the cat is bad, super common prob, when it happened to me the inside honeycomb structure of the cat collapses on itself and the O2 sensor isnt getting the rite temp due to that so it throws the code. you can take it off and put a 2.2 ohm resistor in the O2 plug that will make the ECU get an "all good" signal, but you should get rid of the cat all together unless you have checks in your state, then you need a new one $$$$.

[Dekonx]

So far my plans are

Replacing all vacuum lines
Tps
Cam position sensor

[Garrison]

I think it's more than likely -not- a sensor. I'd check the fuel pump/spark first. Check your coolant overflow for oily substance - it's possible your head gasket is gone.

[Wiscon_Mark]

Sounds like spark plugs/wires to me. Coilpacks don't usually suddenly go out. If a spark plug or wire was having trouble conducting well enough to create a spark, I can see a higher rpm/more fuel & air choking it out worse. Something similar happened on my old Outback and changing plugs & wires fixed it 100%.

It's possible the 02 sensor could be making performance worse, but it's unlikely that it is choking your car as it's more of a "final reading" type sensor that affects how the ECU learns rather than telling the ECU directly what to do.

I would also make sure that your throttle is actually opening smoothly with the movement of the pedal - take off the intake and pump the gas pedal (car off of course) and make sure it's moving well. Also, clean the MAF.

[Dekonx]

I unplugged the fuel pump then cranked the car figuring the car would start with the pressure in the fuel system. Upon doing so it didnt even try to start. Ordered a new fuel pump. Going to start there. Since the fuel pump is 21$ @ rockauto.com I think with shipping and everything the total came to 40$ with the fuel sock and a new sending unit gasket.

The car got new plugs and wires at 240k miles. AFAIK they should last longer than 15k miles.

The oils changed pretty regularly well the filter anyways. Im typically adding .5qt every month or so. I drive 72 miles every day for work.

The butterfly for the throttle body opens up just fine its not stuck shut

The coil packs are probably og to the car. So its likely they have 255k miles on them im not trying to rule them out. Im just saying that the wires I know are good and they are a pain in the dick to change anyways. When I got the car it had a terrible miss if you splashed in a puddle.

http://s669.photobucket.com/user/zarake ... f.jpg.html (http://s669.photobucket.com/user/zarakenpa/media/36489_10150942311644112_146035410_n_zpsf70c7def.jpg.html)

These are the plugs that came outta it. They were all sorts of fouled out. It had 3 different types of plugs in it. The one on the far right was on the back pass side where the plug wire was literally melted onto the plug. When I tried to pull the wire off it fell apart. It also left the rubber thing to seal the end of the wire on the plug in the tube so I couldnt get a plug socket in the hole after it. Ended up getting a pair of long nose angled needle nose to reach in there to get it out.

That was about 15k miles ago that I did that.

Once I get it back to running shape im quite sure itll need a new set of plugs anyways since Im betting its detonating like a mofo.

BTW you know the intake isnt the plastics that run between the airbox and the throttle body right? The intake is actually bolted to the engine itself and what the throttlebody is connected to. Not trying to be a smart ass just saying....

But yes, the throttle body butterfly is moving as intended both by hand and when you pedal it.

The o2 sensor has been a problem since I dropped the exhaust off the car to weld it back together and broke the sensor wire. Im not concerned with it. Its been like that for well over a year. Its not spontaneously going to make the car not run anymore.

Im willing to bet that the fuel pump is probably shot and not pushing enough fuel to actually open the injectors enough to get the fuel into the engine.

While it may have been getting enough pressure to open the injectors enough for idle / very low rpms. What im assuming here is that as I pressed on the pedal more the lack of pressure combined with the excess air was not conducive for a good combustion. Hence the reason why when I would let off the pedal it would return to idle fine. Upon pressing the pedal down it would sputter spit and throw a fit cause the a/f ratio was completely off and the combustion was being smothered with too much air.

Ill update in a few days once the parts come in and I get this back together.

Im still not ruling out the TPS or the CPS.

The CPS I suspect would be the culprit behind improper spark.
TPS I suspect would be the culprit behind too much air and the computer not realizing the tb butterfly is open where it is. So it doesnt step up the fuel map accordingly causing the engine to get too little fuel and way to much air.

___________________________________________________________________________________
Why I started at the fuel pump______________________________________________________

I started at the fuel pump because when I went to start the car up to start testing other sensors and what not. IT refused to start after a good 3-5 minutes of trying. I could hear the pump prime. I know I have 1\4th tank of gas. While it would sputter and try to start it would abruptly die upon letting off the gas pedal.

Again as I said. I unplugged the pump, Tried to start it figuring w\e pressure in the lines would atleast let it run for a bit. However it didnt even try. When I took the lines off the fuel sending unit there was very little fuel that ran out of the lines.

[Garrison]

Sounds good, keep us updated.

[Dekonx]

Update, replaced fuel pump. The car ran awesome for about 10miles accelerated great, even ran up to 90mph just fine. Stopped in a town nearby turned it off then back on. Ran fine for a bit on the way back symptoms returned. Car wouldn't rev over 3k rpm if you mash the pedal it instantly craps out.

Will run if you turn the key off then back on until you hit 3k.

Dropped the exhaust cause people said could be the cats are plugged ran way worse.

Replaced cam and crank pos sensors.

No improvement.

Car will free rev all day bouncing off the rev limiter.

You put load on it if you drive like a senior citizen the car runs alright for a bit at about 40mph it starts throwing a tantrum. Cycling the key of then on fixes it for a bit.

The plugs are pretty bad but not totally wrecked. Did a compression test all cylinders are 210ish +/- 5 psi

Running out of ideas faster than reality tv show makers.

Halp!

[GTDrifter]

My old lgt would bog when you pressed the gas pedal. Cel was thrown for the fuel pump though so I would say that's not your problem

[Wiscon_Mark]

How about the fuel filter?

Does the coil pack have 3 wires or 4 running into it? If it has 4, PM me.

And yes, I do know that the intake is more than just the plastic piping, I could have been more precise I guess, but you knew what I meant.

[Dekonx]

changed the fuel fileter a few weeks ago.

Coil packs are 3wire.

Sorry this problem has been obnoxious and really fustrating since its just eating money and nothing really seems to be the problem. I ordered a OBD2 scanner that does graphing for all monitors. Hopefully it will be able to help me to trace down the problem.

[Asphalt Art]

Sounds like a case of vapor lock, or a bad/clogged cat. The O2 sensor controls fuel in closed-loop mode, so if it is fragged it is going to use a generic single reference reading.

What do the plugs look like currently? The plug that was missing the center electrode has suffered damage from excessive detonation. That doesn't randomly occur.

Timing issue perhaps as well.

[Dekonx]

Those plugs in the picture are what I changed into the car probably a year or so ago, no clue how long they were in the car before I changed them. I know there was 3 different types of plugs in the motor. Only 2 of them were paired.

That was also because the plug wires were shot, and its quite possible that the plug missing the electrode had 20k+ miles on it since it was the only plug that was beat to hell and back and was also on the back pass side where the plug wire literally was melted onto the plug and the probably was NEVER EVER CHANGED.

The new plugs could be changed but thats not the issue im sure of that.

Im pretty sure I mentioned that it was a very old picture anyways.

Compression is fine, The exhaust isnt connected to the motor it runs even worse than with the exhaust hooked up. It barely idles with the exhaust off of it.

The 2ndary o2 (after the cats) isnt hooked up. Hasnt been hooked up since about a month after I got the car 2 years ago.

Fuel pressure is fine. Have spark.

For w\e reason its being thrown into limp mode. However it doesnt even limp its in like death bed mode.

I have already eliminated the clogged cat by removing the exhaust. The car runs even worse with the exhaust not on the car what so ever. Unbolted it right from the heads.

The throttle body looks good. The gasket doesnt seem to be leaking.

It has plenty of vacuum hooked up the vacuum tested and it is fine at idle. however while revving up it will almost act like its building boost. I seen it spike as high as +3 psi of positive pressure. however when it let off it would jump all the way to the other side of the spectrum having retarded amounts of vacuum when you let off.

I dont even need to drive the car to get it to the point of running like crap. Once it starts getting warm it auto jumps to craptastic run mode.

Could the Intake air temp sensor, or coolant temp sensor be causing problems at this point?

[Thomistopheles]

I spent most of my night diagnosing a similar issue with a friend's 99 L. Once the car warmed up it would have almost no power from 1500-3000 rpm. After logging his timing advance I saw that it was dipping to -10° when the hesitation/bogging/bucking happened.

I searched and found that Subaru put out a TSB for the 99 model year, hesitation at low RPM. Search "Subaru TSB 11-58-99" on google and it should come up.

Your ECU might see ignition as knock, causing it to pull timing and then store that value in memory. The TSB says to use a subaru select monitor and view knock signal. If timing dips to -10° when the hesitation/bucking/bogging occurs, it says the ECU needs to be replaced with a new one. For a 99 OB 2.5L it lists 22611AD68B as the replacement part number. If your knock sensor is over tightened, it can give false readings and give you the same issues... so make sure it's torqued to spec.

[Dekonx]

Well that's no good i hope it's not the ecu. That's one of the most expensive parts. I ordered an obd2 scan tool that does live data monitoringand graphing for 10$ less than the 79.99* any near by shops want for diagnostics and that is only IF they do the work to fix the problem. Otherwise it's closer to 400 for diagnostics.

Im always working on cars might as well invest in better scan tool other than my 7yr old actron code reader. Lol

Anyways thanks for letting me know ill probably be dormant till my scanner gets here and i can run it log all the monitors the scanner can log 160data streams at once so im sure I'll see the problem when i hook it up and drive it around the block a few times.

[Wiscon_Mark]

Luckily ECUs almost never fail, and they're very common and cheap, used.

Knock sensor was a good suggestion, they'll kick a car into limp mode pretty quick. I'd check the knock sensor first before you jump to the ECU.

[Asphalt Art]

You need to have some form of exhaust on it for it to run well, because there is absolutely no backpressure by unbolting it from the heads. You have destroyed the capability of the engine to scavenge exhaust gasses, and your car is leaving a very bad mix in the cylinder for the next combustion cycle.

Recommendation at this point is to fix what you know needs replacing first, instead of looking for more things that are probably stemming from the issues you already know about.

[Dekonx]

Problem is I have no idea what needs replacing atm.

Other than putting a new 2ndary o2 sensor on the car thats been broken for over a year+ now that im pretty positive isnt the problem. The car isnt throwing any codes for the issue it just runs like complete ASS.

I can simply run down a list of stuff I have replaced.
Fuel filter
Fuel pump
Crank pos sensor
Cam pos sensor
Throttle pos sensor

I get spark, I get fuel. I dont know how to test injectors to see if they are dying or crapping out for some reason at higher rpms. Not to mention I have never heard of anyone saying that their fuel injectors died. I have seen where people use too small of injectors on boosted engines but thats not the case here.

The issues I already know about are nothing that would cause the car to act like this.

Can a IAT sensor cause the car to run like this?
Can the CTS cause the car to run like this?

Possible stuck injectors? After warm up/ closed loop ecu mode? Also why would they only get stuck after warm up? I can pin the throttle and it bounce off the rev limiter. Then falls off after it gets warm?!

Right now everything that I thought wrong to fix it has been changed.

Im outta thought wrongs, and stuck with possible wrongs but they are expensive to just go. Eh, imma try this....

So if you would please, someone. Make a list of all possible issues im sure one or more might make sense.


PS: I have replaced the knock sensor right before changing the plugs since the car was throwing a code for it. Its one of the things that innitially got the spark plugs changed in the first place. I tightened it up with a thumb ratchet so that I wouldnt over tighten it and crack the sensor. (Yes I made sure it was tight and not loose. Also put lock tight on the bolt threads to make sure it wouldnt work itself loose since I didnt crank it down to my normal tightness.)

[Asphalt Art]

P0141 is the heater circuit malfunction for the rear O2. In layman's terms that means it is taking too long to heat up per the ECU, and it doesn't like that. If you broke the wires, then that is the cause. Emissions-only device used to verify catalytic converter operation.

You say you have replaced the Fuel Pump, FP Relay, and the Fuel filter. You also say you have enough fuel pressure, but was this verified with an actual gauge or guesswork? A symptom of fuel pressure issues is exactly what you are describing, as the FP cannot support flow under load. I see no mention of the Fuel Pressure Regulator, and if the spring is weak it will bypass a lot of your much needed fuel pressure. Are you 100% positive that you reinstalled the lines properly onto the fuel filter? It only flows well in one direction.

Now the big obvious thing to me. If it is running like a dodge when it warms up only: No questions asked, that is a problem with CLOSED LOOP OPERATION. Now, your ECU uses your front O2 for fuel trim. Reading plugs is the best thing in this case. Is it running rich (plugs wet and dark) or lean (plugs dry and light)? Your CTS and IAT are used primarily for open loop operation until the front O2 has a chance to reach operating temperature. Odd anomalies are that under heavy acceleration or WOT sometimes ECUs will go open loop for fuel trim, ignoring the O2.

On the same note, if your IACV is dirty as hell on the sealing surface, it will allow excess unmetered air into the engine and lean it out. A good sign of this would be a previous cycling of the the idle constantly as the sensor attempts to remove all excess air by moving it to close the gap, then resets to full out after it cannot travel far enough to do its job.

Crank/Cam/TP sensors have nothing to do with open/closed loop operation. If they are bad it will be evident cold or operating temperature.

Injectors being stuck is a mechanical failure and also has nothing to do with Loop operations. What it can do is lean out your engine from having clogged orifices, and basically can't supply enough fuel under load. I wouldn't focus on this just yet, as usually one will go bad before and cause a misfire on just one cylinder. On the other hand of this, if you have one really wet plug after a few failed starting attempts, then the injector is probably stuck open by debris.

[Dekonx]

Actual pressure test to verify im getting enough fuel pressure.

The plugs look rather normal atm as they were just changed trying to fix the problem the old ones have carbon buildup on them some are whitish others are darker. But 15k miles + i drove on them almost 20miles with this problem trying to make it home.

This weekend ill get my friend to go for a ride with me and record what its doing and post the video.

Id do it myself but don't want a bunch of people telling me how unsafe it is to do that on YouTube.

[Dekonx]

The fuel filter was done about 2-3 weeks before it really crapped out. It did this once more, changed the fuel filter. Ran great 350 miles a week. For 2-3 weeks then it really crapped the bed.

I dont see the need to spend another 26$ to replace the fuel filter again since I have the old one still laying on the floor boards of the car.

[Thomistopheles]

PS: I have replaced the knock sensor right before changing the plugs since the car was throwing a code for it. Its one of the things that innitially got the spark plugs changed in the first place. I tightened it up with a thumb ratchet so that I wouldnt over tighten it and crack the sensor. (Yes I made sure it was tight and not loose. Also put lock tight on the bolt threads to make sure it wouldnt work itself loose since I didnt crank it down to my normal tightness.)

Spec is 17.3 ft-lbs, anywhere from 16-17 will be fine and anything above it will result in a hypersensitive knock sensor that gives false knock readings. Use a torque wrench, this is one of those parts where you can't guess. The sensor is also supposed to be positioned in a certain way.

https://sl-i.net/FORUM/images/imported/2013/11/knocksensor_zpsd416ecb6-1.jpg

[Dekonx]

Not arguing, just going to point out that the sensor was replaced a REALLY REALLY REALLY REALLY REALLY REALLY long time ago. No reason for it to work just fine and now out of the blue go MEH NOT WORKING ANYMORE AND DEFINITELY NOT GOING TO THROW A CODE.

It was changed because it was throwing a code and acting up. Since being changed it hasnt thrown a code, or seemed to be a problem what so ever.

The knock sensor is, like the rear o2 sensor, is not something im even going to look at for being a problem until it presents itself AS a problem.

[86bratman]

Don't discount that little knock sensor just because it's not throwing a cel. My old 98 L acted all kinds of weird, bucking under load, no power at random, random hesitation. No cel at all, knock sensor replaced problem went away.

Oh, and mine did just out of the blue say screw you I'm done.


You can pick up a knock sensor on eBay for under $10 shipped btw.

[Asphalt Art]

Don't discount that little knock sensor just because it's not throwing a cel. My old 98 L acted all kinds of weird, bucking under load, no power at random, random hesitation. No cel at all, knock sensor replaced problem went away.

Oh, and mine did just out of the blue say screw you I'm done.

Yep, it will only throw a code if the resistance value the ECU looks for goes out of whack. There is a ramp time until that happens when performance degrades though.

They detect knock by Piezoelectricity, which means pressure and/or vibration creates voltage. They can also be "fooled" by other vibrations around them (like bad valve racket). They can also be contaminated by oil from leaky engines, or water from pressure washers. Corrosion at the mounting surface will also distort the readings, as they are extremely sensitive and work within a fairly harsh environment.

[Dekonx]

Well, as soon as my new scanner comes in. Im sure ill have a much better handle on whats going on looking at the live sensor monitor data streams.