OB99W

There's a pretty good chance that it's ignition-related. The type of noise heard on the radio is often from a spark discharge. It could be from plug flashover, leakage through insulation of plug wires, a gap that shouldn't exist (internal break in a plug cable or resistor plug, or even just a poorly seated connection at a plug or the coil), insulation breakdown inside the coil, etc. One thing you can try is waiting until dark and looking under the hood for sparks. You shouldn't see any. You might try spraying down the plug wires and coil with some water, and see if the problem appears or worsens. An ignition system in good condition will tolerate the moisture without problem.

There's some truth to ok4550's satement that "There is no way on God's green earth that a faulty half-shaft is going to cause a rough idle while the vehicle is stationary and not moving.". However, the discussion wasn't about idle roughness, but vibration; some roughness at idle (especially when in gear) isn't unusual, but feeling it to a large degree is.

Interesting reading. "What you should do is invite any of those morons on the Subaru board who claim what you state to be true to this board and let them post that half-baked theory here." is an excellent example of why I post on USMB and not elsewhere. As much as the relationship between vibration at idle and half-shafts may seem "half-baked" to "ok4450", I tend to put more faith in numerous anecdotal reports of cause-and-effect than the best of theories.

These older links may be useful, but they won't pull up articles from the last few years: http://www.endwrench.com/archive/sysdrive.html http://www.endwrench.com/pages/archive.html

Why bolts? Grab a turkey baster or similar device, and suck the tank dry as possible; add a piece of vacuum hose or fuel line to extend the tip if necessary. Don't be concerned if a small amount remains. You could also siphon it, but don't start the siphon by mouth! If you do use the turkey baster, buy a new one before the next bird.

See http://endwrench.com/images/pdfs/DirectIgnition.pdf There are errors in the PDF -- in the "Ignition Coil Testing" section, the terminal numbers listed are reversed between the primary and secondary; however, the diagram is labeled correctly.

The laser mechanism and other drive parts can have moisture condense on them (as can the CD itself), causing erratic operation. Even if the player doesn't have an automatic lockout, it's probably best to wait for the car's interior to warm up a bit. However, I doubt that's the OP's problem, since loosing the station presets isn't protective in any way I can think of.

Yep; SOA430Q121 seems to be for 1997 2.5

According to my available info: 87022AA190 is for '91-'94 non-turbo auto (4EAT) 87022AA200 is for '91-'94 non-turbo manual (5MT) 87022AA310 is for '91-'94 turbo auto (4EAT) 87022AA320 is for '91-'94 turbo manual (5MT)

Revs how high (RPM)? Does it still do it when the engine is fully warmed (based on the temperature gauge)?

Exactly. Example: Diodes in an alternator have failed. They get replaced, but in a poor rebuild worn brushes and bearings with lots of miles might not. Yes, the key is "remanufactured". Under the best circumstances, that means several of the particular "cores" are gotten together. They are then fully (or very near that) disassembled, and major components that are in good condition are cleaned for reuse. Other components that are worn (or even might be) are replaced. Where needed, proper lubricants are used. After reassembly, the part is tested under conditions that ensure it meets or beats original factory specs. If done properly, the end result is a part that should last just as long as the original.

Since you've got a scanner, would you mind posting the fuel trims and other relevant data after the front O2 sensor is replaced and has been run a few weeks? It would be a useful reference to compare with current readings.

Just to be clear, although it's worth taking a look, that isn't going to resolve the P0420. My point was that changing the oxygen sensor might not either, and you may still need a cat in the long run. Cars that set P0420 and benefit most from a new O2 sensor typically don't have 200k miles on the cat, and often show the sensor as troublesome by a high percentage LTFT.

The first thing to do is fix exhaust leaks; if any are near the rear O2 sensor, that can cause a false indication of cat inefficiency without affecting other operation. It may be time to replace the front sensor, but there's a chance in your case it won't affect the P0420, or at least not for long. The freeze frame data you posted indicates the LTFT at under 5%, and positive; that means the ECU thinks the mixture was somewhat lean, and is slightly enriching to compensate. That relatively small amount of compensation usually doesn't point to a particularly bad front O2 sensor; in fact, it could be explained by a small intake leak. (Ideally both STFT and LTFT should be near 0%, but indicating a few percent is typical.) At 200k miles, the cat itself may be the culprit. However, assuming that the O2 sensors haven't been changed in all those miles, replacement would be good anyway; the front one will help protect a new cat, if it turns out it's needed.

No matter which way the water spirals (Coriolis not having an effect), I agree with Skip that if the contents could possibly go in the direction opposite of down the drain, I wouldn't want to be nearby.

Okay; I went into the discussion of backfire and afterfire because many people, including pros, use the term "backfire" incorrectly, and because it's unusual on injected engines. If it's truly backfiring, then the likelihood of timing issues is increased. Verifying that the timing belt hasn't jumped becomes even more important, because if it's done that once, it's likely to do it again, leading to possible internal engine damage (which may have already happened). Also, as I mentioned before, explosions in the intake don't do anything good. I'd suggest running the engine as little as possible until the timing is inspected, even if that means not getting the codes read right away. If the timing is good, you can relax while driving to Advance, AutoZone, etc.

I've never done it, and don't know anyone who has. Gary's suggestion to check at NASIOC is a good one. However, in general, "modern" engines are made to much closer tolerance/fit than years ago, since manufacturers are trying to eek out the best performance and gas mileage they can. Therefore, the gains to be had are likely to be much less than possible in the past. As Gary also pointed out, it's just as easy to make things worse as make them better. In addition, the ECU will attempt to compensate for changes, to some degree possibly nullifying them.

By the way, GM's latest is Dexron-VI. Although GM seems to indicate backwards compatibility with Dexron-III, apparently in non-GM vehicles at least one licensee (Citgo) says otherwise; see http://www.docs.citgo.com/msds_pi/10165.pdf . This may not be an issue in PS applications, but in an auto trans, it might.

MD-3 is what several suppliers are calling what used to be Mercon/Dexron-III. http://www.havoline.com/products/na/trans_md3.html

http://www.drive.subaru.com/Fall03_WhatsInside.htm

Agreed, get the diagnostic code(s) read (there may be more than one), and get back to us with any and all of them, not the code definitions. Example: code P0302, not "Cylinder 2 Misfire Detected". By the way, the "backfire" mentioned is more likely "afterfire". As you might expect, backfiring is when the mixture gets ignited in the intake; afterfire is when there's sufficient unburned fuel in the exhaust to be ignited there. True backfire doesn't tend to happen with fuel-injected systems. If it's really backfiring (intake), that can lead to damage of things (such as sensors) there, which could rapidly make the symptoms worse. An overly-rich mixture could be the cause of afterfire, but a lean mixture can be the reason as well since it can lead to misfire. Ignition problems and other causes of misfire can also lead to afterfire. Since the problem came on suddenly and progressed rapidly, I'm not inclined to solely suspect plug wires -- they don't tend to fail like that. They could be a contributing factor, so it's not a bad idea to replace them, especially since the maintenance history isn't known. I was somewhat inclined to suggest suspecting bad fuel, and a partially restricted fuel filter, but smooth running above 3k ... Hmmm, does the "not well documented" maintenance history mention oxygen sensor? As already has been mentioned, don't ignore the timing belt.

It's no wonder the car is "cherry" -- it's caught in a time warp. As you drive it, the odometer will show lower and lower mileage as the car's condition improves. Eventually you'll have to bring it back to the original dealer and relinquish ownership of the car. The up side is that the dealer will pay you the original selling price. (Unfortunately, you yourself won't get younger in the process.)

Are you noticing any problems (at what would be equal to about 25,000 miles)? I suppose you could swap the plugs around, but I'm not sure that would be "fun". If it were me, I'd just run them until they gave signs of needing replacement, or you had some reasonable mileage ("kilometerage"? ) on them, and then put in doubles. If you do decide to swap them, the wasted spark "pairs" are the ones that share a coil in the coil pack (1&2, 3&4), so exchanging #1 with #2, and #3 with #4, would accomplish what you suggested.

I doubt you'll find Dexron II available -- in fact, I'm surprised that you found Dexron III, since GM stopped licensing production of it and II a while back. Dexron III, or something labeled as equivalent/compatible, should be okay in a '98 4EAT.

Take a look at http://www.ngk.com/results_app.asp?AAIA=1432742 (This is for the H4.) If you get platinum, get the double (OEM type) -- the wasted spark ignition arrangement has opposite polarity on two of the four plugs, making single platinum a poor choice. I agree that if things are running well, you can hold off on changing plugs.