OB99W

For reference:http://endwrench.com/images/pdfs/ExhaustInfo.pdf

Yes, if by "left" you mean while facing the engine from the front of the car, and by "distributor", you're referring to the ignition coil pack. The following link should help for many normally aspirated (non-turbo) models: http://www.northursalia.com/modifications/drivetrain/pcv/pcv.html

You may already know this, but since you didn't say so... the "warm up" isn't just of the engine, it's the ATF that has to be at full operating temperature. In other words, idling the engine isn't sufficient -- the car has to be driven. Code P0778 is for an electrical problem, so if that's what you're referring to, flow shouldn't be an issue.

Ooops, I do (usually!) know which cylinders are paired in the wasted spark arrangement; #3 would of course be the mate to #4. That's what I get for posting at 2:19 AM when my brain is apparently half asleep. That's precisely why NGKs are preferred -- they seem to still work reliably long after other brands fail, often even when pushed well beyond the recommended replacement interval.

If you can clean it to the point of the plunger moving freely and properly sealing as required, that's almost all that's needed. The remaining consideration is whether the spring has fatigued through usage, and if the valve therefore still operates as when new (although that doesn't seem to be a common problem). Beside the spring fatigue issue, mainly the advantage is in profit to those selling PCV valves.

From the compression readings, it certainly seems like the engine is basically healthy. Not that an NGK can't be bad (or just too well-used), but I've seen more problems with Champions. It's interesting that even the "nasty" NGK on #4 worked better than its Champion wasted-spark-mate on #2. By the way, could you ID the Champion plug model? It wouldn't surprise me if the NGK was an original, and whoever put in the Champions found replacing #4 to be too much of a challenge. Or perhaps a Champion was installed in #4, failed early, and a used NGK was put back in.

Yes, and the fuse should be as close to the power source (fuse box) as possible. Putting it in the line elsewhere risks the possible consequence of a "short circuit somewhere underneath the carpets". Also, be sure to use the appropriate gauge of wire for the current being drawn, because (besides affecting the audio) if significantly too small, it can overheat and burn, whether fused or not.

Insulation, or the lack of it, doesn't affect inductive cross-coupling, which is what you're describing. All other things being equal, it's dependent on the distance between the two conductors, the energy level, and the relative impedance of the circuits. However, there could have been more noise induced, as you suggested. An increase in wiring resistance due to cut conductors, in addition to slowing injector pulse rise time, could also make cross-coupling to the injector more likely. Whatever the mechanism, you made a good point that damaged wiring could lead to misfire.

I mostly agree. The only way a bad coil pack could cause just #2 (and not #4) to misfire would be if that secondary coil was leaking to ground near the #2 end. I can't see it being the igniter either, based on wasted spark. A bad plug, on the other hand, could affect just one cylinder. Yes, and if compression is okay, you can cross "burnt valve" off the list; if it isn't, you can stop wasting time and money on other possibilities. You may not always like the outcome, but diagnosis is usually better than throwing parts at a problem.

I should also mention that a PCV valve stuck open often causes poor idle quality.

It's certainly possible. Injectors have relatively low impedance, and need to turn on rapidly. Excessive resistance due to some cut conductors could affect the response.

If the PCV valve is stuck closed (or hoses are obstructed), blowby pressurizes the crankcase and can force oil past seals. If the valve is stuck open, the engine can consume (burn) more oil than it normally would, via the intake.

No, the definition of P0183 is "Fuel temperature sensor A circuit high input". That's not the same as saying the sensor has failed. It's just as likely (perhaps more so) that there's a bad connection at the fuel pump connector or elswhere in the path between the sensor and the ECU. I suggest you look over the wiring/connectors before assuming the sensor is bad.

Congrats. What's that old saying, "haste makes waste" ? You're welcome, although you did all the work!

Several things can cause misfire. Running the A/C adds a load to the engine, which can aggrevate problems. Ignition issues are certainly one possibility. So are fuel delivery (injector) problems. Then there's valves... Since you seem reluctant to replace plugs and wires, at least verify that the wires are making good contact at the coil pack. If you do pull plugs, be sure to check compression while they're out. You could try swapping injectors between cylinders, and see if the problem moves. Here's some info on removing and reinstalling them: http://endwrench.com/images/pdfs/ProperInsideEWFall05.pdf Do you have access to a scan tool that can read "live" data (specifically short and long term fuel trims)? If so, it's possible to check injector balance without pulling them.

Welcome to the forum. An engine needs "fire, fuel and air", but the "fire" has to be properly timed, and it also needs compression. Even if the t-belt isn't worn, it's possible that inertia from a very hard hit may have caused some slack that allowed the belt to jump. You could measure compression, but since loss would probably point to mistiming, I'd pull the crank pulley and cover, and inspect the belt and other components while checking the timing. Fortunately, a '92 has a non-interference engine, so unless something really unforseen happened, you should be able to get things back to normal without too much grief. Let us know what you find once the remaining cover is removed.

Various injector/fuel system cleaners use different formulations, but in general they have a base of something like kerosene or Stoddard solvent (mineral spirits), with naptha, benzene, amines and/or some proprietary ingredient(s). Depending on what's ailing a particular engine, one formulation may work better than another. Following are links to Material Safety Data Sheets (MSDS) for some of the brands, which will provide an idea of what they contain. Red Line SI-1: http://www.redlineoil.com/msds/49.pdf Techron Concentrate: http://www.jeadauto.com/msds/chevron/techron-concentrate.pdf Shell V-Power Fuel System Cleaner: http://msds.walmartstores.com/cache/49135_1.pdf STP Complete Fuel System Cleaner: http://www.thecloroxcompany.com/products/msds/armorallstp/stpcompletefuelsystemcleaner4-03.pdf

Since it wasn't mentioned in this thread yet, I will -- the newer style tensioner should not be compressed in a typical vise, as it's not supposed to be done in the horizontal position. Subaru recommends using a hydraulic press, etc., or replacing the tensioner; see pages 16-17 of the EndWrench article mentioned in post #3.

Overfilled PS level is easily lowered using a turkey baster. Just don't return it to kitchen service -- buy a new baster before Thanksgiving.

MSDS for DEX-COOL: http://www.havoline.com/images/products/pdfs/anti_exlife.pdf

http://www.sancarlosradiator.com/dex-cool.htm

Gary makes a good point. If necessary, temporarily remove the accessory drive belt -- that way, it should become obvious if something's wrong externally, or there's a timing belt, etc., related problem.

You hadn't previously mentioned gas mileage issues in this thread. That's a good reason to verify the state of tune as others have suggested, but are you sure that the brake work was top quality? Even slight drag can sometimes cause shift problems. If you haven't done so, you might want to read the thread I linked to above.

By the way, sometimes even dragging brakes can cause shift problems. For a situation causing 3-4 shift problems, read through this thread: http://www.ultimatesubaru.org/forum/showthread.php?t=85717

Yes, as I mentioned previously, the ECU and TCU share sensor data. There's no question that the TPS data is used in determining shift characteristics, and bad data can cause problems. Sometimes when it's only off a small amount, or is intermittent, a trouble code won't be generated by the ECU or TCU even if the shift is affected.