OB99W

Cougar brought up a good area to consider; a stuck PCV valve (open or closed) can cause lots of engine problems. Even if the PCV is operating normally, it can create some "interesting" conditions. I don't know what vehicle we're discussing, but if it's losing oil and it isn't leaking out, then it must be being burned, whether that's obvious at the tailpipe or not. With the mileage you mentioned, there could be significant blowby. If so, that could be giving the PCV system lots to contend with, and in addition to exhaust gases aerosolized oil might be being fed to the intake. On many cars, temporarily disconnecting the PCV at the crankcase end (valve cover, wherever) will allow any significant blowby that may exist to be observed. On the other hand, there may actually be a slow external leak that isn't obvious when the engine isn't running or at idle. If the PCV system (valve, hose, etc.) has gotten plugged, crankcase pressure can build up. That can force oil out of places it might not normally leak; it will usually be worse at higher RPMs, meaning that it might lose a drop every so often on the highway, but not show up in your driveway.

Agreed, which is why I linked to the older thread that had a post suggesting this technique. By the way, sometimes it takes a bit of time for the moisture to seep into a problem area, so spray (gently at first) and wait a minute or two before assuming a part is okay.

Sounds like ignition voltage leakage when wet. See this thread for clues: http://www.ultimatesubaru.org/forum/showthread.php?t=45322 Be sure to get OEM (Subaru) ignition parts when replacing.

"OK" is a relative term. You might get away with it, in terms of knock, if the engine management systems have enough leeway. Of course, if the ignition timing gets retarded sufficiently to compensate for the lower octane, the gas mileage may suffer enough so that the effective cost of fuel will be the same as with using premium, or even worse. In general, manufacturers don't recommend premium if regular will do, since customers obviously know that the higher grade gas will cost them more.

Checking the timing isn't the issue. A damaged t-belt could cause lack of synchronization between the crankshaft and camshafts. If the engine is an interference-type design, that can result in piston crowns and valves trying to occupy the same space at the same time. Without getting into relativity theory , this usually leads to bent valves or worse. Compression or leakdown testing is how it's determined if things are still sealing correctly. Since the engine is a '96 2.2l, it shouldn't have an interference problem; even if valves are open with the pistons at TDC, they won't hit. The thing that seems to have gotten this discussion started was the mechanic's claim that there was almost no compression on all but one cylinder. As has already been stated, you can't do a compression test when the valve timing doesn't relate to crank/piston position.

Desperate situations sometimes call for desperate solutions. No, but I play MacGyver on USMB.

I'm not taking the bet; I agree, a non-interference model probably would. Although I do have the proper tools, you'd be surprised what an old spark plug, a short piece of metal tubing, some epoxy, a bicycle pump and a hose clamp can accomplish .

Other than the apparent confusion (Subaru's fault ) over the 16 blinks being in groups of 4 or not, I think we're saying the same thing. The 16 blinks just indicate a fault exists (they're not themselves a code), and you indeed have to do the diagnostic stuff to get the actual code(s). Thanks for the offer of the docs, but I already seem to have what I need. I hope that clears up any confusion, no matter who it is that's confused :-\ . As I'm sure you know, sleep deprivation can do wonders for your mind too, so that's it for me tonight (especially since I just noticed it's now "tomorrow"!).

Yes, I suppose you could just put on the t-belt "for testing purposes". I'd at least first make sure that the tensioner and water pump operate freely and wouldn't damage the new belt. But I may have to take back what I said above; if you're willing to go through all that work twice, maybe you should do a cylinder leakdown test.

That's the normal operation I described above. Yes, that would appear to be significant. I'm not sure who is braver, you for biking or your GF for driving a car with those brakes .

Subaru must have changed the design, because on page 13 of http://endwrench.com/images/pdfs/4EATPh1Win04.pdf they describe the 16 blinks as 4 groups of 4. Other info does indeed seem to show the 16 blinks as a string of 1/4 second on/offs with no additional pauses. This has nothing to do with the fault codes themselves, by the way, which require use of diagnostic connector B82 and ground probe, etc., to retrieve. See my previous post, which has a link to an earlier thread with good diagnostic info, etc.: http://www.ultimatesubaru.org/forum/showpost.php?p=478251

For a '96, the AT temp light should be lit when the key is ON, but the engine hasn't been started. Once the engine is started, the light should remain on for about 2 seconds, then go off and stay off if all is okay. The 16 blinks that indicate a problem follow the initial 2-second on time; they're supposed to be in 4 groups of 4 each (1/4 second long, separated by 1/4 second), with each group of 4 separated by a 2-second off period, according to Subaru.

Assuming that you're driving enough to sufficiently warm the fluid, and that the dipstick is being fully inserted, it seems that you're checking the level correctly. In my experience, you may have to insert, clean, and reinsert the dipstick a few times to eliminate fluid on the dipstick tube wall causing an inaccurate reading. Dexron III/Mercon should be compatible with Dexron II. It isn't just overfilling that can cause air bubbles in ATF. The level may still be too low, and if so, air can be picked up along with the fluid. The aerated fluid will then have a greater volume (although part will be air), and appear to be at the correct level on the dipstick. Since you previously said that lowering the level made the bubbles worse, and that the level now still seems to be a bit low, that would appear to verify that the level is indeed low. I'd add some more fluid, until the level is at the correct mark, drive a bit, and recheck the level and for bubbles. It may take a while for the air to work its way out. Recheck the level tomorrow, and add fluid again if it is at all low on the dipstick.

I came across this thread, which might be of interest: http://www.ultimatesubaru.org/forum/showthread.php?t=47262

Nipper asks a good question. For a link to Subaru info on the proper procedure, and some comments of my own, see http://www.ultimatesubaru.org/forum/showpost.php?p=479136 By the way, what fluid did you use?

Not really, since the timing belt is what synchronizes the valve opening/closing with the piston position for each cylinder. You could do a cylinder leakdown test, but it would require manually turning the crankshaft and camshafts so that each cylinder's piston in turn was at TDC (Top Dead Center) at the end of a compression stroke (or you could say beginning of a power stroke) with both intake and exhaust valves closed. If you're not precisely at TDC, the crank could want to turn from the cylinder pressure, so you might have to prevent that from happening or try several times. By then you'd have done about as much work as changing the t-belt .

That's probably the parking brake taking its time releasing. You might want to verify that the cable isn't hanging up and causing the p-brake to drag.

Agreed! Also, good that it's a '96, because the 97 2.2 wasn't non-interference. See http://www.endwrench.com/pdf/engine/FtEngineOverhaulSp98.pdf for lots of info.

In addition to the link I previously gave, here are a few more that might be of interest. By the way, the '99 4EAT (with external filter) is a Phase 2. http://endwrench.com/pdf/feb2004pdf/4EAT.pdf http://endwrench.com/images/pdfs/4EAatPh2Win04.pdf http://endwrench.com/images/pdfs/4EAT.pdf http://www.endwrench.com/pdf/drivetrain/02Winter4EATTransOp.pdf

There are certainly a number of possible causes of the behavior you're describing. On the simple end of things, my own '99 OB gave me grief when the fluid level was just a little low, so that might be worth considering. Subaru's method for checking the level is included in: http://endwrench.com/images/pdfs/4EATDiagnosisWin02.pdf As with automatics in general, the 4EAT is sensitive to fluid level, and neither too low nor too high is good. I've found that it isn't easy to get an accurate reading of the level. Start with the info in the above link. Because the fluid tends to stick to the dipstick tube, giving a false reading, I found it necessary to insert and wipe the dipstick off more than once. Even then, I look at both sides of the dipstick and use the side that indicates a lower level as being the correct one. Once I got the fluid level correct, my OB's trans has been behaving. Perhaps you could be as fortunate.

That's what it was, one lifter had lost it's prime. Looking at the rockers it was the upper right lifter. Thanks, Rob You're welcome. Excessive lifter leak-down can be caused by several things. Engine oil that's dirty, fuel-contaminated, wrong viscosity, aerated, or just not getting where it needs to can all cause the problem, although those things tend to affect more than one lifter. If the oil hasn't been changed in a while, doing so might be a good idea. Sludge and/or varnish can be a cause, and of course, a damaged/worn lifter is also a possibility. If an engine isn't run for a long enough period, a marginal lifter can leak down while the others hold. Even good lifters leak down after sitting unused for long enough, and which ones will be more prone to do so can depend on the rotational position at which the engine has stopped; pressure on a particular lifter will vary with the point in each cylinder's cycle (percentage of cam lift). If things are okay, they'll pump up again within seconds to maybe a minute or so. EDIT: See the sections concerning HLAs (Hydraulic Lash Assemblies) and how rocker shaft relief valves can affect them in http://www.endwrench.com/pdf/engine/FtEngineOverhaulSp98.pdf .

The brake rotors rust very easily, and if the car is driven infrequently rust can build up a bit between uses, especially under humid conditions. You could have a sticking caliper, but even if the calipers are releasing correctly, there isn't much clearance between the rotor and pads. You can hear the pads scraping rust on the rotor surface, especially when driving down an alley or near a fence, wall, etc. If the brake lining is worn sufficiently, you might hear the wear indicator scraping. Hopefully, you're not hearing a pad's metal backing plate running on a rotor, but with under 30,000 miles that's unlikely unless brake usage has been very hard or a caliper is really stuck.

No problem, you're welcome. Page 10 of http://www.endwrench.com/current/winter02-03/EngineMaint.pdf has Subaru's recommended approach.

As Carl (1 Lucky Texan) suggested, here's a link to one possible approach: http://www.ultimatesubaru.org/forum/showthread.php?t=28193 By the way, I assume that by "vapor lock", you're referring to an air pocket in the cooling system. Usually, "vapor lock" refers to a condition where liquid fuel has gotten so hot that it's vaporized (typically in a fuel line, pump, or carb bowl [where it's called percolation] of an older vehicle) before it should; when that happens, the engine can't run. It's not usually a problem on modern fuel-injected systems.

It's not likely a closed exhaust valve (checking #4 for leakdown on what should be the exhaust stroke for that cylinder would easily determine that), although that would effectively give you twice as many "compression" strokes as usual. Cranking a sufficient number of revolutions when checking compression would even the readings if that's what's happening. EDIT: I believe that the '91 had hydraulic lifters, so one that wasn't pumping up could cause an exhaust valve problem, and explain the noise as well. Perhaps you're doing an unintentional "wet" test. Can you peek/probe into #4's plug hole and see if it's got some fluid in there? By the way, temporarily disabling the air-hold valve on a compression tester can sometimes be revealing. So can using a vacuum gauge, often even at cranking speed; of course, the spark plugs have to be in for that .