teahouse

Ah, interesting - I guess they did away with the thermometer on this model, for some reason. So today I drove about 20 miles, highway and around town... with no stalls! I think the problem has been fixed.

Made my way through a few junkyards today, and eventually found a 93 Legacy with the same MAF sensor... unbolted it, and cut wiring to the harness about 8 inches down in case I needed that too. They charged about $52... which is still less than a quarter of new price, so I was OK with it. Put it in tonight, went for 2 test drives, and it ran great... no stalling. My harness only has 3 wires, and the one on the Legacy had 5 running to the connector, so I wasn't sure if the wiring in the receiving connector would be compatible... but, luckily, it seems to be running great... It seems at this point, then, that the problem was in the connecting wiring in the housing for the MAF sensor... which would explain why it would die when I would wiggle the connector... it wasn't the connector it's self, then, but some intermittent connection wiring behind the connector that I was wiggling... I think. Of course, it could have been a problem with the sensor it's self too, but the sensor wires looked good, and I would think that a problem with the sensor would most likely be a completely on or completely off situation. So, I'll report back again to verify that everything is working as it should. Again, much appreciation for the help... sincerely.

Oh, I figured a MAF would include only the cylindrical sensor... but along with it comes the entire black housing on top?

Oh right, I'll check the wires under the loom... I wiggled them quite a bit, and didn't find correlation to the stalling, but it wouldn't hurt to take a look. Before I cut the MAF sensor housing, let me try just replacing the connector and MAF first... it's the black housing on top that you cut though, right? I.e., where the female part of the connector harness is attached to? I just found the optimal fuel pressure for a 1997 Legacy... 26-30 psi... so, considering that people say these engines are very similar to the Impreza, I'm guessing my PSI is within optimal range.

Ok, well... second test drive today was the complete opposite of the first one - total failure... I went about 100 yards, and it died... and I could barely get back - had to start it about 20 times before I was back to the driveway. Ha, I didn't knock on enough wood I guess. So, I do still think that the problem is in the wiring of the MAF sensor, or perhaps the sensor it's self - it's just my initial attempt at a solution wasn't adequate... I mean, who would have thought that rubber bands and tape wouldn't work - whaaa? lol. GD - you were right again, I'm heading to a junk yard tomorrow to clip a connector, if I can find one. Do most subaru's use the same clip? I.e., can I take one from a 2000 Forester, for example? Fairtax - sounds like the same symptoms I have. Good idea, checking the wiring under the loom... I didn't even think you could get under there - are there screws that I didn't notice? I'll go down and check soon. Also, I just had the thought to grad a MAF at the junk yard too - should be considerably cheaper than the $220 for a new one. Alright, I'll report back soon. **edit** - also, does any know what the fuel pressure is supposed to be on a 97 impreza? Mine was steady at 26-27... which, since it was so stable I figured that was what it was supposed to be at... and I can't find the optimal pressure online anywhere... nor is it listed in the manual.

Ya, you called it - nice one. The thing that threw me is that the wires and contacts looked perfect, when I checked them. I actually had to really wiggle that thing around while the engine was running to realize that it was a bad connection. I just triple cleaned the contacts with electrical contact cleaner, and since the plug had a little bit of play, I secured it in place with two rubber bands and tightly wrapped electrical tape... sounds like a hob-job fix, but it seemed pretty ideal for securing the clip tighter. I just got back from an initial test drive, and it ran beautiful - better than it has in a long time since I've gone through the engine and replaced sparks, cleaned the MAF, etc... I'll go for another drive tonight and we'll see if indeed this was the fix that was needed, and it keeps running consistently, fingers crossed. If not, good idea to go to a junk yard and just pick up a new clip -- that's actually what the guy at O'really mentioned this morning when I returned the fuel pressure gauge. Thanks so much for the help everyone, and I'll post back once more to let you know if everything keeps running smooth... and probably many times more if it doesn't (-.-).

Ok, borrowed a fuel pressure gauge from local O'reilly... hooked it up last night. Didn't realize until I read the manual that I could install it right before the fuel filter, just using one extra 2" piece of hose, instead of bypassing it completely (duh). Pressure was rock solid - first time I cranked it to ignition on, engine off, it primed to 15 psi... second time kicked it up to about 35. When idling, it consistently stayed around 26 to 27 psi... and when the engine died, it hadn't changed... in fact, it would instantly kick up to 35 psi when the engine would cut out. So, thank you GD - you saved me some work there, and possibly money as I may not have been able to return the fuel pump if I installed it and used it, even if just briefly. Knowing that my fuel pressure was good, I thought that it almost had to be a faulty sensor... so I started wiggling plugs all over and eventually found a correlation to the engine dying if I moved the MAF connecting harness in a certain direction. Ah! It's a faulty plug - just slightly too loose, and it is either shorting out or losing contact. So, hopefully today I will find out if there is a way to replace just the plug, since the MAF sensor it's self seems to work fine... This may very likely be the issue. I'll report back to update with results.

Hmm, your right... I'll go buy a pressure gauge. It's OK to bypass the fuel filter temporarily, then, I take it? Also, what is ideal pressure, and how much tolerance is there? I read 35psi somewhere... I should be able to just get a refund for the pump, if unused.

It's true, but the thing is, I don't have a full raster of necessary tools -- so without an in-line pressure gauge and a way to mend the hoses that I would need to cut, presumably, I'm left in the dark. Yes, I could buy an in-line gauge... which, maybe I should have... but I've already ordered the fuel pump, so I may as well try it I guess. Hopefully the OBDII scanner I bought will give me a clue.

It does stay on, when I first start the engine after a stall... then when I shut it down and start it again, it will usually be off... It really does sound like a dying fuel pump, the more I read about what symptoms that would usually exhibit... who knows though. Would a dying pump somehow trigger a CEL? Maybe by somehow triggering a crank sensor, as an after effect, for example?

It's weird - the first time I unplugged it, it actually was relatively stable, just fluctuating a little... then when I tried it again this morning it was all over the place. I just cleaned the MAF with MAF cleaner, and it actually seems like the engine is running smoother, having more of a smooth 'purr'. But... the problem persists. I drove back and forth in the driveway just now, and noticed that the CEL is coming on after it dies... so having the code reader is going to be good, when it gets here (it's in shipping). I also noticed that it runs rough when I start it right after it dies... it will sort of struggle to keep stable RPMs... then if I shut the engine off and restart it, it usually runs well again...

Well I gave it a shot just now (disconnecting the MAF) and it was pretty rough - RPMs were all over the place, and the only way to keep the engine from dying was to keep on the gas. It does show me how big of an effect the MAF has... I'll try cleaning the contacts really well and see what that does.

Really? Interesting - I'll try it. I actually did unplug it a couple days ago to see what would happen, but only let it idle for a few minutes. The change I noticed was that the idle RPM kept fluctuating... but it didn't die in that time. So the effect of the MAF is such so that it could cause the engine to die completely, instantly? I'll try it tomorrow morning and post results.

I looked at both sensors today. I'm not sure what else to look for besides obvious abrasions in the wire and dirt on the contacts... of which there were none. So the CEL wouldn't signal a bad crank, cam, or MAF sensor? I want to specify that the car does not need gas to idle... it will idle perfectly, for any given period of time. Then, usually without even stuttering, it just shuts off instantly. It WILL stutter when I'm driving just before it dies, though... which will happen either 1 or 15 miles down the road. So I guess a fuel pump was a good one to try next, as we throw the dice.

Swam through all the grease just now to find and unplug the crank sensor. While dirty on the outside, the contacts looked perfect, and the wires look good. MAF sensor looked good too. IAC -- there's a new idea. I'll give it a look. If I were to take the MAF sensor out, is there a way to check it using a voltmeter?