OB99W

All true, but if you live in the rust (salted roads in winter) belt, the rotors may indeed look like the ones in porcupine73's pictures or worse. Even with good pads, I've seen rust creep in from both the rotor outer edge and the hub area, leaving only a narrow band of solid metal near the middle of the pad area. Also, the problem with the pad tabs/ears seeming to be tight in the caliper can be due to rust as well; scrape as needed.

Yes, it's a good idea to "crack" open the bleeder rather than risk pushing "crud" into the ABS. In fact, I'd suggest putting a short piece of hose on the bleeder, with the open end going into a jar; besides not getting brake fluid all over the place, it will become obvious why that's important soon. That in itself won't necessarily be a reason to bleed the lines; if you close the bleeder just as the piston has been pushed all the way back, it's unlikely that any air will be introduced. On the other hand, bleeding/flushing isn't a bad thing, especially if the brake fluid collected in the jar looks anything but light in color and free of contaminants.

I've never swapped the hoses, and so can't verify what Paul said, but it should be easy to tell which is which. Usually the front pump sits lower in the tank than the rear (since running out of fluid for the rear isn't as critical, and tells you that the front will be dry soon, too ), and tracing out the front hose isn't a problem. On the other hand, if that doesn't do the trick, are the pumps actually running?

Not good, but better a head problem than in the block. It does seem strange that a valve (or two) could burn that rapidly, but perhaps it was marginal (somewhat leaky, but not that noticable) and you just reached "critical mass". It's likely once you get in there you'll find lots of carbon, which is typically what leads to the valve failure.

Your '92 has OBD-I, which unfortunately isn't nearly as sophisticated as OBD-II. There can easily be sensor problems that won't generate trouble codes. For example, a rich mixture may be due to the ECU running open-loop all the time, caused by a bad coolant temperature sensor (CTS). It's normal for the mixture to be rich during engine warmup, but if the CTS doesn't "inform" the ECU that the engine is warm, the mixture will remain rich. An OBD-II system typically "knows" that it doesn't take forever for an engine to warm up, and generates a code, but the earlier OBD apparently just believes the sensor. Bottom line is I'd start by checking the CTS.

Follow the directions you found at the site; the green connectors are for "test mode", the black ones are for "read memory mode". As already suggested, make sure the sensor tip is clean of foreign material, and that the mounting area is clean; a fraction of an inch of extra clearance between the tip and reluctors can cause the output to be low. Sometimes the coil goes bad, sometimes the magnet weakens; either can result in low (or sometimes no) output from the sensor. Unless you know what the correct amplitude is, a meter test will only tell you if there is output, not if it's strong enough. Also, seeing output at the sensor doesn't mean it's getting to the ECU, so look over wiring/connections. If that all appears okay, consider getting a used sensor, which should be quite a bit less than $130. I seem to remember that this car had a problem with previous work on the timing belt. Did the emissions ever improve after the timing was resolved? If not, is there any possibility that the camshaft sprocket is somehow wrong, mismounted, or has damaged reluctors?

The magnet is in the sensor; reluctors are built into the camshaft sprocket. See: http://endwrench.com/images/pdfs/DirectIgnition.pdf

I don't have much time right now, so this is going to be short. See the last section of this: http://endwrench.com/current/spring04pdfs/InsiderInfo.pdf Don't forget that misfire causes aren't limited to ignition. Good compression probably means healthy rings and valves, but how is the fuel delivery?

Gee, Nipper, thanks for all that good info... ...but I hope you weren't responding to me as if I didn't know the answers to the questions I posed; they were meant to get others to think about alternative theories to the cause of lost compression as measured by the gauge. EDIT: Perhaps I should give up on the subtle approach.

Could be "stop leak". Not usually. Unfortunately, temperature gauge spikes usually indicate bad HGs, but make sure you don't have an air pocket from not filling the system properly.

I'm not going to offer any answers, but I'm going to ask some questions that might be food for thought. How many of you have seen a hole in a piston or a burned exhaust valve be intermittent? How does a compression gauge work? Where does what is being compressed come from? What would happen if an intake valve didn't open?

(I can't tell from what you said just how you intend to use the LED, so perhaps the following information is something you already know.) If an LED is used without current limiting, the ECU lamp-driving circuit can be damaged similarly to with a wrong bulb. If the LED doesn't include a built-in current-limiting resistor (i.e., designed to run at about 12 volts), you'll have to wire one in series with the LED. The resistance value determines the current -- typically to limit the LED current to somewhat over 20 milliamps, a 470 ohm resistor could be used; for 10 milliamps (the LED will be dimmer, and may not be easy to see in strong daylight), use a 1,000 ohm (1 kohm) resistor.

When I've previously suggested shaving to match size, I got responses saying that most places stopped doing it due to liability concerns. It's good to know that it's still being done.

Just to clarify, when I mentioned "disconnecting the injectors", I of course meant electrically. Break-in of "modern" engines doesn't usually require special oil; use high quality oil of the viscosity you intend to run, since a good oil will include a decent additive package. (Some here may recommend their favorite additive, however. ) I would suggest a non-synthetic initially. Although all the ports were "taped" during storage, there might be some (hopefully minor) corrosion. I'd watch the color of the oil carefully for the first few hours; if it darkens rapidly, change it and the filter even if the hours/mileage are very low. Just as important as the oil is the coolant. Be certain that there's no air trapped in the cooling system. It's a topic that comes up on the forum often; do a search. During the break-in, it's a good idea to to be easy, but not too easy. Keep the revs reasonable for the first few hours until you see how things are going. After that, it's important not to hold RPM constant for too long; vary the engine speed and load, and give it short bursts of acceleration that over the first few hundred miles are increased in RPM and duration. Again, monitor the oil and change it if necessary. Best of luck with the project.

If you didn't do so, the filter should be filled with oil so that the engine doesn't crank with it dry for any time.

You might first pull the plugs and put an ounce or so of oil into each cylinder, then turn over by hand to distribute. When cranking, disconnecting the injectors until the oil pressure comes up could help prevent fuel from washing the oil down.

Agreed, a bad enough accident or living where they heavily salt the roads in winter could be reasons to dump the car; normal wear-and-tear can be dealt with as problems develop, as long as parts are available. Now that I can't completely agree with. I'd suggest that you retain the same amount of liability coverage on any car you drive, limiting only the amount of collision/comprehensive. If you get into an accident while driving an old car, and it's your fault, you're just as liable as with a new one.

There are numerous documented cases where it's been shown that companies are aware of dangerous or even life-threatening defects in their products, but they choose to ignore them after doing an analysis of the cost of fixing the problem versus that of potential lawsuits. I don't believe that every lawsuit is brought for legitimate reasons, but some people do suffer real and substantial injuries due to the negligence of others. They're entitled to compensation, and additional punitive damage awards help discourage the wrongdoer from repeating their actions. As to the original poster, my understanding is that the outcome of class action lawsuits is "iffy" at best; they often result in the lawyers taking the majority of any award, with little left for the plaintiffs. Therefore, if the motivation for the suit is to get compensation, I'd forget it and just talk to SoA. If "punishing" SoA is the goal, a class action might work, but why would one expend the time and energy needed to pursue this comparatively small matter? Life's too short, but maybe I'm just getting old.

Yes, if not done recently. Don't forget the fuel filter in particular. Also, if the fuel quality is poor, the ECU may sense knock and retard timing. Is it sluggish in general, or is there a road speed range or other circumstance under which it's especially poor?

You're welcome, for whatever help was provided in the "deglitching". Thanks for getting back to us with the "final" (now, we know that's not the word to use when it comes to car repair ) outcome. Don't be a stranger; come visit even if the OB behaves. Uh-oh -- whenever I know a lot about something, I realize that I've probably spent too much time with it.

You're welcome. Your eyes are probably okay. Start here: http://www.ultimatesubaru.org/forum/forumdisplay.php?f=3 . Then look at the gray bar at the top of the list of threads. On the left it says "Threads in Forum : New Generation of Subarus"; on the right is "Search this Forum" -- click on that and it will open a box for the search. Note that it won't return anything if you don't enter at least four characters; if the term you're looking for has fewer, you need to add a "wildcard". For example, to search for "ATF", enter "ATF*".

I decided to go the simple/dedicated/easily-portable approach, with an Innova Equus 3130. The AutoTap ATU looks nice, and under certain circumstances I could see that some of its features (beyond those of a dedicated scan tool) could be useful. I assume you already have the computer you'd have to interface to.

Yes, #1 is passenger side, front. Some cars get internally oil-fouled plugs due to bad piston rings, and some get them externally "fouled" when the plug wells becoming full of oil from gasket leaks. Here's what Magnecor says about it on some engines that are particularly prone to the problem: http://www.magnecor.com/magnecor1/techbulletins/tech_mrbishi1.htm It should be "safe", since you're taking care of the problem soon. Although the EJ engines seem to like original ignition wires best, if the oil leakage is stopped and the wires can be sufficiently cleaned (and haven't developed carbon-tracking or the like already), you may be able to keep the ones you have. Basically, if you don't get a returning CEL and misfire code once the work is done, you could leave the wires in until they give evidence of a problem.

I used the search function and came up with these, which may help: http://www.ultimatesubaru.org/forum/showthread.php?t=69733 http://www.ultimatesubaru.org/forum/showthread.php?t=65288 http://www.ultimatesubaru.org/forum/showthread.php?t=61106 http://www.ultimatesubaru.org/forum/showthread.php?t=55211

Flashing of the CEL usually indicates a misfire that's severe enough to potentially damage the catalytic converter; don't wait too long, or you may be facing an expensive job. As was already suggested, getting the trouble code(s) read is a good starting point. By the way, why did you need to change the battery? Was the charging system checked? Low/erratic voltage conditions can sometimes cause sensors to malfunction and the ECU to get "confused".