OB99W

Oil does expand with increased temperature. The dipstick upper hole is the mark that indicates full when cold, while the notch above it indicates full when hot.

Probably, but just to be sure, what's the part number?

I think you'll find this thread, and the PDF file links within it, very helpful: http://www.ultimatesubaru.org/forum/showthread.php?t=19805 General 4EAT info is here: http://endwrench.com/pdf/feb2004pdf/4EAT.pdf I suggest downloading the PDFs, and reading the entire thread.

The front O2 sensor is sufficiently critical that we usually suggest it only be replaced with an OEM unit. However, you can usually get an original factory one from an online dealer for around $100, including shipping, so why buy anything questionable for $140?

If it's truly "just dusty", why not carefully use compressed air to clean off the majority of it, ignore the rest, and avoid the whole issue of moisture getting into unwanted places? Dust doesn't typically prevent you from monitoring things like developing leaks, etc.; in fact, it often makes them more evident. However, if you're going to spray the engine down, a minor amount of water entering the intake isn't going to kill it; just don't inject a stream of water in there.

I forgot to ask -- does the hesitation seem at all related to temperature and/or humidity?

That sort of thing shouldn't affect operation directly off idle; it's usually more apparent at higher speeds and loads. Ken/Suby Skier: I've had a very minor stumble right off idle in my '99 OB pretty much since new, but never enough to be worth the trouble to me to diagnose. Just how bad is the hesitation on yours? Is it noticable if you very slowly depress the accelerator? Not that I'm suggesting you do it if you're concerned, but you could scribe the TPS in a few places before trying to adjust it, so that it could be set back to very near the original position. However, putting a meter on the TPS output and verifying that the voltage changes smoothly when the throttle is moved off idle should probably be done before touching the adjustment. If the voltage is at all "jumpy", cleaning or replacement may be the key. More basic than the above, have you checked the ignition wires? Corrosion or loose fit at the plug or coil ends can sometimes cause problems similar to what you're describing. Clean, tighten (carefully, as to not damage the boots), and check that resistance is within spec.

Road-speed-related vibration can be "fun" to diagnose. Here are some links to possibly useful info: http://endwrench.com/images/pdfs/ChaseEWAug06.pdf http://www.aa1car.com/library/vibrations.htm http://www.aa1car.com/library/tires2.htm

The PCV system is designed to deal with "normal" engine blow-by. Since an engine in decent shape doesn't tend to have much blow-by at idle, intake vacuum is used to counteract the spring in the valve and mostly close the valve down. As engine load increases, blow-by will as well; the valve's spring will more readily overcome the lowered intake vacuum at load, opening the valve further and allowing the blow-by to be dealt with. If the PCV valve or hoses are restricted, or the engine is worn and produces more blow-by than the PCV was designed to handle, crankcase pressure can become higher than it should be. That can cause oil to be forced from around seals, etc. The hose and the valve could be checked and cleaned if necessary, although purchasing a new valve isn't usually that significant an expense.

I'm not sure which would take that many trips, or have other stringent requirements. I did find the following, with a list covering several manufacturers. Notice that they're all from 1996 or just a bit later, the early years of OBDII. http://www.njrec.org/Downloads/OBD_Probs2.pdf

There's a problem on the '96 models. On that year's Subaru cars, each time the ignition key is turned off, the readiness monitors are reset. Therefore, if a drive cycle isn't completed during the trip to where the car is tested and the engine left running, the status will be "not ready". Subaru isn't the only one to have that problem with early OBDII firmware, although some companies have reprogrammed replacements to fix it.

An OBDII "not ready" status can be caused by someone disconnecting the battery long enough for the ECU to lose its memory, and the car having not been driven enough to complete the necessary cycles. http://endwrench.com/images/pdfs/OBDInfo.pdf The vehicles listed as exempted from rejection due to not being ready are ones that have "issues"; that is, the drive cycle requirements are sometimes difficult to meet. In that case, a particular state can decide to loosen inspection requirements, or mandate that tailpipe emissions be directly measured (rather than depending on OBDII) to verify that the "not ready" status isn't hiding an emissions problem.

You're describing what is usually a wheel sensor problem. The ABS ECU runs two basic sensor tests each time the car is started. The first is done before the car is moved, and checks for electrical continuity. The second happens after the car is moving for several seconds at sufficient speed, and looks for signal from the wheel sensors. If the ABS ECU sees something that isn't correct, the ABS is disabled and the dash light comes on. Things to check for are proper gap between the wheel sensor tips and the tone wheels (also make sure the gap doesn't vary much as a wheel is rotated), no damage to the tone wheels (severe corrosion or missing teeth), and that there isn't some foreign material packed between the tone wheel teeth.

Since there's no "2x" mark on the dipstick, the actual degree of overfilling is open to interpretation; it's very unlikely to really be 2 times a normal fill. The fact that the only complaint was of an oil smell (no mention of strange noises or other unusual behavior) leads one to believe that the overfill probably isn't severe, and reducing the level to normal (and perhaps checking the PCV, as I mentioned previously) might be all that's needed.

I asked because I know that certain parts of South America are using high percentages of ethanol fuel blend. Ethanol has a higher vaporization heat than gasoline. When the engine is cold (running open loop, with the ECU unable to compensate) ethanol effectively leans the mixture at a time when the engine needs it richer. However, 10% ethanol isn't excessive; if higher, it might have explained all or part of the problem. Since you can't know for sure when the fuel filter was last replaced, and a partially obstructed one could cause the mixture to lean during high demand (cold engine or hard acceleration), changing the filter is a reasonable starting point. Please let us know if that resolves the problem.

Try the diagnostic sequence in the link I gave above. If nothing's happening at the igniter input (ECU output), then do check cam and crank sensors.

Leakdown might be a better choice than compression with a broken t-belt.

Perhaps. Other than not firing up, does the sound of the engine cranking seem different than usual? When was the belt last changed?

The following has a section on no-start diagnosis: http://endwrench.com/images/pdfs/DirectIgnition.pdf (The resistance specs given for the primary and secondary coils are in error, they're reversed.)

Agreed, although of course there's no guarantee that will fix the problem. Here's something I previously posted: Based on the full capacity of the 4EAT and typical drainage from the pan by just removing the plug (with the car level), I roughly calculated the percentage of old ATF replaced with new by multiple drain/fills. The following is in the form #-of-changes -- %-fresh-ATF: 1 -- 40% 2 -- 64% 3 -- 78% 4 -- 87% 5 -- 92%

Yes, or the fuel filter, as has been mentioned, or some other possibilities. What is the octane rating of the fuel? Are you using an ethanol/gasoline blend, and if so, what is the ratio?

Umm, you're considering having these people take apart your engine and inspect things? :-\ They may have overfilled the oil, or they may have neglected to fully drain it before filling. Either way, I'd suggest that they do a proper drain/refill, including filter replacement. Someone should probably check that the PCV system (valve, hoses, etc.) didn't get oil-logged. Make sure you get paperwork indicating why you're bringing the car back. Do check the oil level before you leave the shop's premises. Overfilling the oil could cause some long-term problems, but that's not a "given". As long as the circumstances are documented, I'd be inclined to just drive the car for a while after the situation is resolved, and see what happens. If you're feeling that you want things looked over, based on what you've reported I'd suggest finding another service shop to do that.

For those interested in more info: http://en.wikipedia.org/wiki/E85 http://running_on_alcohol.tripod.com/id36.html

Spark demands go up when cylinder pressures are higher, so that may increase the likelihood that there's an ignition problem. The "wasted spark" circuit puts the companion ("paired") cylinders' plugs and wires in series; spark current for plug #1 also flows through plug #2, and that for #3 flows through #4 as well. (When one cylinder is firing, at the top of the power stroke, the companion is on the exhaust stroke; it only "wastes" a few KV to jump the gap of the plug that isn't under compression.) However, since the paired plugs are connected at opposite ends of a coil in the pack, the spark polarity is reversed for one versus the other. If I remember correctly, cylinders #1 and #4 wind up with negative at the center electrode, and #2 and #3 negative at the ground electrode. Since a spark can jump more readily from a hotter surface (the center electrode), all things otherwise being equal, it's easier on #1 and #4 than with #2 and #3. So, a defect in the coil, etc. might affect the firing of a plug that has "reversed" polarity, but not the companion one getting "correct" polarity. In other words, a P0303 code doesn't necessarily mean that only cylinder #3 plug or wiring could be at fault. If you haven't already done so, check the wire resistance and connections to plug #4 as well, even if there's no P0304 or other evidence. As I and others have already mentioned, looking for stray sparks and spraying down the coil, wiring and boots before digging in further could be worthwhile. Glad you like it.

Anyone besides "Ray L" ever try this?: http://groups.google.com/group/alt.autos.subaru/browse_frm/thread/de2cf0b44beeaf8d/60e45ff877c9abcd%2360e45ff877c9abcd