OB99W

It appears that the most recent post from turbo92subaru was about 1.5 years ago, so you may not get a response. However, this thread does bring up the "fix" for torque converter drain-down on some of the early '90s 4EATs -- a cooler line wth a check valve was installed.

Google "tesla wireless power transmission".

Both of these links have info on using the black and green connectors to clear codes without doing a full reset of the ECU: http://www.scoobypedia.co.uk/index.php/Knowledge/ECUReset http://www.cycoactive.com/Urabus/urabus_seldom.html Be sure to start with the black and green plugs disconnected, warmed engine, not running. Follow the rest of the procedure as in the links, and when completed, disconnect both sets of plugs.

Good point, Nipper -- a clogged PCV (valve and/or hoses) can allow the crankcase to pressurize and increase the likelihood of the problem.

Hint: page 12 of the previously linked article.

Descending a hill is a condition when the throttle is closed, intake vacuum is high, and combustion chambers cool down. Oil gets past rings, and especially past worn valve stem seals, and doesn't get much of a chance to burn off. When you level off and open the throttle, you're moving a lot more through the engine and the combustion temps go up, burning off the accumulation. Bingo, instant smokescreen.

Prudent self-control? Rev limiter? Sheer good luck?

Good start, probably healthy rings and valves, timing likely okay. Also a good sign. However, what was the vacuum reading when you held at 2000-2500 rpm for several seconds? I wouldn't jump there quite yet. Please answer the question above, and see Cougar and my posts about sensors.

See Cougar's and my posts. As Cougar said, just replacing a sensor doesn't mean the entire circuit is working. Assuming that you reset the ECU and this isn't an "old" code, I would concentrate on this problem; it certainly could be at the root of the gas mileage and emissions troubles. For the TPS you apparently have, it should be closer to 5 volts at idle. Idling at 1100 rpm will waste a lot of gas, among other problems; with a fully warmed engine, it should be several hundred rpm lower -- more like 600-700.

Thanks!

As it says on page 5 of http://endwrench.com/images/pdfs/Fuel.pdf (which I previously provided a link to): ------------------------------------- Some Legacy models were equipped with a TPS with an inverted voltage signal. Voltage was approximately 5 volts at idle and decreased as the throttle was depressed. -------------------------------------- You might want to reread that information, including the part about the idle switch.

Magnetic fields weaken with heat, and clearances can also change. A sensor that's marginal might become balky when warm. Under any circumstance, getting trouble codes read is the first step.

There is much more than "the obvious" to using a vacuum gauge. The following links should provide an idea of just how much info about engine condition you can get:http://www.international-auto.com/index.cfm?fa=ad&aid=47 http://www.classictruckshop.com/clubs/earlyburbs/projects/vac/uum.htm http://www.earlycuda.org/tech/vacuum2.htm The basic concept is that a healthy normally-aspirated (non-turbo) 4-cylinder engine running at normal operating temperature will indicate a relatively steady 18+ inches or so at most engine speeds below wide open throttle (WOT). Opening and closing the throttle, depending on how rapidly that's done, should cause certain changes in vacuum, returning to about 18+ inches once the engine speed is again constant. Make sure the engine is fully warmed beforehand. A steady low reading, periodic drop, or significantly vibrating needle all indicate problems. There are two specific tests for restricted exhaust: 1) The first is to bring the engine to 2000-2500 rpm for several seconds, and then rapidly release the throttle; the vacuum should begin at about 18+ inches, jump to near 22 inches when the throttle is released, and then rapidly and smoothly fall back to about 18+ inches again. If the vacuum falls off when the engine's at 2000-2500 rpm, or if there's a hesitation getting to or returning from the higher vacuum when the throttle is released, it probably indicates an exhaust restriction. 2) Rapidly go to WOT. If the vacuum drops to a very low number (near zero) and then slowly rises as revs increase (be careful not to over-rev!), there's a "breathing" problem; it could be an exhaust restriction, or one at the intake. If any vacuum readings don't seem "healthy", get back to us here with details. Exactly. It's not uncommon for engine blow-by to eventually gum up the PCV valve and tubes, especially after a lot of miles. You said you replaced the valve, so the tubing should be cleaned, or replaced if it's deteriorating. If the PCV clogs up again in a few hundred miles, there's probably excessive blow-by. You're welcome. I hope we help you get to the bottom of this before you have to buy another car.

Certainly dragging brakes could cost mpg, but of course that wouldn't explain the high CO and HC readings at idle that caused the car to fail emission I/M. Usually even if the rotors are a bit rough or rusty, if the caliper is operating freely the pads/pistons get knocked back with driving and there shouldn't be much impact on mpg. Gary, did you rebuild the calipers (or relube the sliding parts) at the time of pad/rotor replacement?

On my own '99 OB, the temp gauge tends to mainly sit (once the engine is fully warmed) at about one "tick" below 1/2 scale. The owner's manual says the entire range of the gauge, except for beyond the next-to-the-last heavy tick, is "normal". Personally, I'd be more than a bit concerned if mine got near those upper marks .

Well, I don't know about "always" , but thanks for the compliment, Cougar.

This is from the '99 Owner's Manual (which is at hand, for perhaps obvious reasons ), but should be applicable to all similar models. ------------------------------------- Flooded engine If the engine does not start, it may be flooded (excessive fuel in the engine). In case of a flooded engine, turn the starter motor for five seconds with the accelerator pedal fully depressed. Repeat this two or three times until the engine starts. Release the ignition switch and accelerator pedal as soon as the engine starts. -------------------------------------- Fortunately, unlike on a carb, there's no accelerator pump.

Since it starts more easily with the pedal depressed, Cougar may well be right. Holding the pedal down to the floor "unloads" it, just like doing that opens both the throttle and choke on most carbureted engines.

See: http://endwrench.com/images/pdfs/Fuel.pdf You're welcome.

Agreed; that's why I suggested certain tests be run, especially use of a vacuum gauge. The $20 (sometimes even less) ones are perfectly adequate for the job, and can tell you a lot about engine-related things if you learn how to interpret them. Mouse nest in the intake?

Did you use original factory (OEM) parts? These are just the sort of things that are critical to operation, and if not closely meeting specs can really affect operation. For example, a coolant temp sensor could be within the "normal" range of resistance that the ECU "expects" (and therefore not throw a code), but inaccurate enough so that the ECU thinks the engine is colder than it actually is; that could lead to constant over-enrichment of the mixture, or even to the system running open-loop all the time.

Okay, I really do know the term "doughnuts" or "donuts", but what fun would there be in that?

Maybe I'm naive, but I've never before heard the term that 91Loyale used. I think I know what crazyhorse001 is writing about, but it doesn't seem to be at all the same thing.

My response was a follow-up to where I thought Sonicfrog was going, and based on something you said in your PM to me (sorry, but for some reason my browser is freezing when I try to return-PM). Usually a plugged converter won't allow good performance in general, but sometimes the problem only shows up when loose material shifts, and going uphill might be the condition needed to cause that. The high CO and HC readings indicate a rich condition. An intermittent misfire could be related, and since there's a code for the cam angle sensor, maybe it's not always triggering proper ignition, and should be replaced; check connections first, though. On the other hand, it may be time to stop throwing parts at the problem and have someone do a real-time scan. Knowing for certain what the engine sensors are doing, and whether the management system is eventually going closed-loop would be very useful information. It's too bad the car doesn't have OBD-II, but even the earlier systems can be diagnosed. A vacuum and/or compression test also might be worth considering, just to check general health of the engine at 170,000 miles.

Another possibility with P0420 and an exhaust "whistle" is partial blockage; a failing cat can be the cause. However, you would likely notice some performance loss when the engine is under heavy load (accelerating at highway speeds or hill climbing), and since you said "It is very loud, but does not seem to affect how the car drives.", this possibility is less likely.