98 Impreza Outback runs crappy sometimes..

[deezul]

some background:

Just bought a 98 Impreza Outback (automatic).. When I bought it the CEL was on and the previous owner said it was because the stock airbox and intake tube had been removed to put a cone filter in its place.. He included the stock airbox and intake tube with the car, so put the intake tube back on.. I was unable to put the airbox back on because the ears on the MAF sensor had been cut off to fit the cone filter onto the maf.. 

what the car is doing:

Sometimes when I start it, it will idle at about 2000rpm and then drop to 1500rpm, and then when I put it in drive it will drop to 1000rpm and i can drive it with no major issues..
But other times it's different. When i start it it will start at 1500rpm then slowly drop to around 800rpm and start misfiring bad. When this happens it's completely undriveable. But then I can shut it off and start it back up and it will run fine again.. 
Occasionally, it will be running fine and then start running crappy out of nowhere..

I had the CEL scanned and it read: knock sensor, iac valve, and misfire on all cylinders.. I replaced the knock sensor today and it runs better when it runs good, but it still does the crappy misfire thing sometimes..

any ideas?

[okamikai]

Hi, the initial warm up is triggered by the ECU even if the car has only been off for an hour or so, it will accelerate and the slowly decelerate until it reaches around 800-850 RPM. The misfiring may be caused by air in the fuel line, bad ignition module, bad spark plugs, or bad TPS (Throttle Position Sensor), although on this last one the misfiring is very sudden and of short duration since it usually happens when the throttle is at a certain position. I hope this information is useful to you.

[deezul]

I bought spark plugs for it. im going to install those today and see if that helps

[idosubaru]

use NGK plugs and OEM Subaru wires or at the very least high quality wires.  those engines are not very forgiving of other ignition components for some reason, older subaru's it didn't matter like that.

 

if the plugs/wires are unknown age/manufacturer then that needs to be checked.

 

clean the MAF - they make MAF cleaner - the oil from an aftermarket filter may have saturated the miniscule hair thin wire in the MAF.  or just replace it.  they never really fail so just get a used one for a few bucks.  i'm sure i've got 3 lying around.

 

IAC valves - on older subarus cleaning them out can help, just spraying them down with carb cleaner if there are passages/valves that can get dirty and stick.  i haven't had issues with newer gen so unsure if that could still help - but either way, same thing as the MAF the IAC's almost never fail so a cheap used one is a fine option.

[deezul]

changed the plugs today and it hasnt misfired again yet, so that may have fixed it.. im still going to clean the iacv, pcv, and maf just to be safe..

[johnceggleston]

your issue is probably air flow related.

maf or just an air leak.

 

the parts you need, what ever they are should be cheap and easy to get.

once you get it back to stock you should be on your way.

 

have you had tyhe codes read?

this might tell you where to look.

[FlyBoy83]

I've got an issue that sound some what related but it still has me confused. I have 1996 Legacy Outback with the 2.5l that I just bought two weeks ago, it had the classic "dead spot" in the tps behavior (nothing changes with the engine behavior with certain pedal positions) easy fix right? it had 5 codes present: TPS fault, MAF fault, EGR solinoid fault, upstream O2 and Downstream O2. I thought the easiest thing to fix was the TPS and it might help resolve the others. new tps and down the road we go... not!!! new tps and it ran pretty good til I cleared the codes to see which coeds would come back...and weird behavior here we come. while tring to drive at a constant 35 mph the engine will rev to about 3000 then cut out entirely and catch again about 1500-2000 and then repeat. it retained the TPS, up and down stream Codes and the EGR. I looked and somebody in the history of the Car had removed the EGR sulinoid and plugged all the vac lines to the EGR system, but I didn't think that the EGR would effect it that badly. I put the old TPS back in and it drove fairly decently, switched it out with the new one again and it went bonkers again. why would it be driveable with a bad tps and not with the new one? is there something else Im not seeing/understanding with the ECU system?

[Fairtax4me]

I've got an issue that sound some what related but it still has me confused. I have 1996 Legacy Outback with the 2.5l that I just bought two weeks ago, it had the classic "dead spot" in the tps behavior (nothing changes with the engine behavior with certain pedal positions) easy fix right? it had 5 codes present: TPS fault, MAF fault, EGR solinoid fault, upstream O2 and Downstream O2. I thought the easiest thing to fix was the TPS and it might help resolve the others. new tps and down the road we go... not!!! new tps and it ran pretty good til I cleared the codes to see which coeds would come back...and weird behavior here we come. while tring to drive at a constant 35 mph the engine will rev to about 3000 then cut out entirely and catch again about 1500-2000 and then repeat. it retained the TPS, up and down stream Codes and the EGR. I looked and somebody in the history of the Car had removed the EGR sulinoid and plugged all the vac lines to the EGR system, but I didn't think that the EGR would effect it that badly. I put the old TPS back in and it drove fairly decently, switched it out with the new one again and it went bonkers again. why would it be driveable with a bad tps and not with the new one? is there something else Im not seeing/understanding with the ECU system?

Yes. You can't just slap a new TPS on and expect it to work. It has to be adjusted so the ECU gets the correct voltage signal at closed throttle.

 

You could also have a reference voltage problem from the ECU.

 

Post up the actual codes (p0XXX) and we can give more specific advice.

[FlyBoy83]

P0170, P0505, P0120, P0403, P0141, P0130.

 

I did some further checking last night and poking around on several Subaru forums about how to adjust the TPS found that there are some with 4 pin and some with three. if it has four it has an idle switch if three then it doesn't and "auto calibrates" on each engine start. Mine only has three. I disconnected the batt. Cleaned the MAF with special cleaner from AZ. and checked with an ohm meter between the middle pin and each of the outer pins. ( I never could find a diagram or labels to indicate which was pin 1. figured that the middle pin had to be #2) When I open the throttle plate the readings from the TPS were 0.75Kohm-3Kohm and the other was 5Kohm-1.5Kohm. I finally just more or less centered the TPS and put it all back together and is running with out the CEL coming on so far.

[okamikai]

My 98 Forester had the dead spot issue, I replaced the TPS, and with the car on, I calibrated the position of the throttle until it was purring like kitty, hasn't failed since. I will take the advice of cleaning the MAF and IAC for better performance, but where is the IAC located at?

[FlyBoy83]

the CEL stayed off for most of yesterday am but this am it was on when I hopped in to go to work, I haven't checked to see what the codes are yet that came back but I can tell I'm getting something like 12mpg... I'll see which ones they are this evening and go from there...

[Fairtax4me]

P0120 TPS A circuit

P0130 O2 circuit B1S1

P0141 O2 Heater circuit B1S2

P0170 Fuel Trim B1

P0403 EGR Control circuit

P0505 Idle Air Control System

 

The thing to do is see which codes came back.

 

The O2 heater circuit code usually pops up if the sensor is unplugged. Check the wiring and make sure its plugged in and actually there. That code is for the rear O2 sensor.

 

The other O2 sensor code is for the front sensor and that one combined with the fuel trim code could mean the sensor is bad. This could account for the bad fuel mileage. But if you're trying to figure mileage by what the gauge says it will be way off. These fuel gauges are notoriously inaccurate.

 

IAC code will probably go away if the TPS is calibrated properly. Same with the TPS code.

 

The EGR stuff isn't that complicated to put back on. Grab what's missing from a junkyard and that code should go away.

[FlyBoy83]

the ones that came back are as follows:

p0106, p0170, p141, p0403

 

I'm calculating my mpg at each fill up. I have figured out that even though the Owners manual says it is a 15.2gal tank when the gauge says E I usually put in about 10 gal. I've been getting 12-15mpg here the last week. the crazy part is that when we drove it to E Tenn for thanksgiving we were getting between 25-26mpg with the bad TPS and all the other stuff whacky too. I tried to figure it out while out there ended up not really changing anything and coming back we averaged between 19-20mpg (it is roughly 800mi each way).

 

while I was working on the MAF I found that a number of the VAC lines are hard as rock and/or oil saturated. the MAF had what sounds like a loose bee bee rattling inside the box. my Chilton's manual wasn't really helpful is telling me much about resistance readings on the pins of the MAF to check if it was OK or not. a new one from AZ is like $230 really hoping that it isn't the MAF.

[ivans imports]

I put a HP fuel pump in mine kicked it in the butt was suffering from low fuel volume also check ends of plugwires for tracking

[Fairtax4me]

P0106 is the MAP sensor. That's going to be caused by one of those vacuum hoses that goes to the passenger strut tower. You might have pulled one loose and not realized it messing with the MAF sensor. The line to the MAP sensor usually has a T to the fuel pressure regulator. If the FPR vacuum line is disconnected the FPR will close off and make the fuel pressure in the rails too high. This will cause the injectors to spray too much fuel. The FPR can also split inside and leak fuel into the vacuum line which will get sucked up into the intake, then it runs really rich all the time. Pull the hose off and see if liquid gas pours out. If there is liquid in the vacuum line the FPR is bad and should be replaced.

Either of these could be the cause of the fuel trim code, and give you very poor fuel mileage.

 

 

Leave the MAF alone for now. They don't usually go bad, and if they do they cause major driveability problems like stalling and really rough inconsistent idle. But you do need to make sure all of the hoses that connect to the air tube are tight and have no splits or breaks. Check the hoses that run down to the valve covers too. If any of those have splits or breaks it can cause problems. Check the tube itself for cracks and splits, and make sure the tube is clamped properly at both ends.

Edited December 11, 2013 by Fairtax4me

[FlyBoy83]

it looks as thoug I need to replace all the Vac lines and pcv hoses. they are hard as rocks and cracked and looks like sombody's jimmy rigged mess ontop of that... does any one have a pic of what the vac lines on the EJ2.5 DOHC. also the EGR system so I can recognize all the parts that go with it cause the only piese of the EGR that I have is the valve on the intake manifold all the egr orofices are plugged with pipe plugs. looking at the vac diagram it looks like there is a over-presure sensor of some kind that is also missing not sure how it connects in or what it looks like either...

[FlyBoy83]

I replaced several vac lins that were hard and the upper pcv hose tightend and checked the hose clemps etc on the intake "snorkle" i thought i would trace the rear o2 sensor and see what I could figure out on that one... turn out really obvious it looks like the wires snagged on something and it completely removed the plug from the sensor. there are just three frazzled wires poking out of the o2 sensor. any way I drove the car to work this am and so far the CEL has not returned. it does have a slight hicup every now and then or delayed throttle responce.

[ivans imports]

On my 98 outback sport the cat broke apart and pluged up exhaust got yery quite and stoped whisling changed cats and power came back and the whisle works spanky on mine the rear o2 was broken So changed that to runs way better now

[FlyBoy83]

the TPS Code came back last night. I can't seem to find any good info on how to adjust it... the guy at the import parts store said ignore the rear o2 sensor (fix it last). he said on the 96 it is merely tattle for the cat sys but the fuel trim is not based off rear sensor. im not sure how much the EGR system would effect the TPS or not.

[FlyBoy83]

found the EGR Parts at a Salvage from a 1999 model... we'll see if it makes any difference or not...

[FlyBoy83]

I got the EGR Parts installed. it runs a little smoother but the TBS and fuel trim codes are back... is there any better info on how to adjust that tps correctly. the ohm readings i get don't make any sense with what I'm reading. the tps was fine for like two weeks...

[Fairtax4me]

Thought I posted that, apparently it didnt make it.

 

TPS should read 0.5v on the signal pin (middle) with the key on and throttle closed. Have to back probe the connector to check voltage, TPS has to be plugged in while adjusting.

[ivans imports]

The rear 02 is very important it checks how mutch o2 is left in ex after the fisrt catt has done its job and ecu bases it fuel trim off that no rear 02 no fuel trim. Tps I never see problems with these sensers ever not on bad one yet BUT many bad wireing harness conectors and not fully pluged in conectors and broken engine wire harness that cause tps codes continuity check from main engine conector to plug on tps had a outback with broken harness between main conector and senser was a real pain to find and senser had been changed 3 times

[FlyBoy83]

With the ignition on the plug unplugged I get:

Pin A: 46.2 MV

Pin B: 0.4 MV

Pin C: 4.1 V

 

When I plug it in and back prob pin B I get 58.2-18.4 mV

[FlyBoy83]

Ignore the previous comment. Used a different setting on the volt meter and got it set and the CEL is staying off for the moment. We'll see if only the O2 code comes back.