OB99W

I suspect they have several reasons, not the least of them the continuing rise in the cost of gas.

Given NH's motto, I hope someone verified the car was unoccupied.

Is a "Subota" only for weekend drivers? http://en.wiktionary.org/wiki/subota

That doesn't answer the question of whether "conditioner" (stop-leak) has been used, and to what degree. I'd assume the dealer has been adding that stuff to satisfy Subaru's requirement for head gasket leakage "preventive maintenance". Normally it's not an issue, but if someone isn't diligent about getting all the air out of the cooling system, the "conditioner" can harden and clog things internally. I certainly hope coincidence explains your problems, but I'm not convinced of that.

If that's really the issue, see the first two sentences of my post #34 above.

If the issue is mainly ease of availability, and not so much pricing, online dealers sell the OEM spin-on filter for about $23 plus shipping charges. You could probably have one delivered to your door in a few days. The quality of aftermarket filters is variable -- some may be up to the job, some not. If we're talking about an '03 GT, I personally wouldn't risk the trans versus saving $15-20. However, "Ya pays yer money an' ya takes yer chances".

Just in case the compressor/clutch/belt aren't the source of noise and smoke (it shouldn't be difficult to determine that), perhaps the blower is. You could run the heater a bit, and try the blower at several speeds -- if it's bad, it should be possible to detect without A/C.

Hmmm, I thought necromancy would be considered unholy. :-\

That's the general idea, but even if they open at 212 degs F or so, it won't determine correct operation. To do it properly requires a thermometer that covers the opening range, and both the thermometer and thermostat should be suspended in the water, not touching the side or bottom of the container. Heat the water, and note at what temp the stat begins to open, and at what point it's fully open. If working correctly, it should start opening a bit under or at the rated temp, and be fully open by a few degrees above rating. If it's not fully open by a maximum of 10 degs over rating, it should be replaced.

It does seem a bit more than coincidental thermostat failures would explain. I assume this is on the '01 OB. Has it had "conditioner" used on the cooling system? If so, how much and how often?

Scottbaru, apparently, but that post is over three years old.

Just to be certain, are you saying that three OEM Subaru thermostats have failed (or been initially bad) within one year?

Agreed, it's definitely worth looking at the transmission cooler lines. They run from the trans to the radiator and back, so a leak in one of them could certainly put ATF in the area mentioned. The leak might be substantial and easy to spot, or could be a small/slow one that finally made you aware of it when the fluid got to the exhaust. Fluid can travel due to gravity, car motion, etc., so the source may not be obvious at first. Especially if the leak is small, you may have to clean the area in order to more readily see it.

Assuming you're using a high-quality oil filter, 5,000 miles may be overly conservative. It's somewhat dependent on driving conditions -- length of trips, speed, ambient temperatures, etc., all play a part in determining what's a reasonable interval.

It wouldn't surprise me if an O2 sensor was "unhappy" about that. I'll join the consensus, and recommend not crossing deep streams.

If the noise changes when a fuel injector is disconnected, it doesn't necessarily indicate the injector is the source. By stopping combustion in a cylinder, the high pressures associated with it are eliminated. Therefore, a change in noise level or characteristic when an injector is unplugged could also point to piston (slap), connecting rod (either end), etc. There's some pretty good engine noise diagnostic info that might help localize the problem here: http://remanufactured-engines.com/page4.htm

Good point John, although from post #12 in this thread, it appears that ignition pulse pickup was directly used to verify the tach.

The only two simple (non-miraculous ) explanations are that either the tach is non-linear or the speedo is. Since you've verified the tach's accuracy with other instruments, that leaves the speedo. You said "i went by a speed sensor for the cops the other day and i regestered perfectly.....25 in a 25", but that doesn't prove indicated 60 MPH is accurate. If there's a stretch of highway in your area that has mile markers and is relatively free of traffic sometime during a day, you can verify the speedo. Doing a constant 60 MPH, measure the time between mile markers; it should be exactly one minute from marker to marker. If you can hold 60 over several miles, the accuracy will be better. (For safety, this should be done with a second person in the car doing the timing.) Any discrepancy from one minute per mile will indicate speedo error.

Let's assume for the moment that the picture represents actuality. If that's the case, engine RPM at 80 MPH should be 4/3 of that at 60 (in 5th gear at both road speeds), or about 2200 RPM (not "3500-3800 RPM"). If the RPM reading at 60 is accurate, then 3500-3800 at 80 indicates a really badly slipping clutch at the slightly higher road speed. I'd be more inclined to think the instrumentation is wrong.

Sorry, but that just doesn't "compute". There's a pretty linear relationship between road speed and engine speed in any particular gear. Let's assume you're turning 3600 RPM at 80 MPH. Then you would be doing about 60/80 times 3600 at 60 MPH. That reduces to 3/4 of 3600, or 2700 RPM at 60 MPH.

It will totally cut the supply to that cylinder, if the injector isn't leaking at all. I assume you're attempting to determine whether this will result in better fuel economy. As Nipper pointed out, the 4th cylinder will be a load on the engine. Cars that drop cylinders to improve fuel economy address the valve train as well as the fuel, so the unused cylinders aren't pumping air from intake to exhaust. Also, injecting large volumes of air into the exhaust can have "interesting" results on the engine management system, etc. If you're still inclined to experiment, I'd suggest at least wiring a switch that can be operated from the driver's position to the injector. That way, you can try cutting the fuel only above the point that vibration seems to disappear, and be less likely to cause damage.

By accelerating wear, and if severe enough, inducing metal fatigue that might lead to catastrophic failure.

Perhaps the next generation of gas mileage computers will have a means of entering the cost per gallon. That way, they can not only show how low the MPGs are when you punch it, but also indicate how much the thrill just cost you -- that might help lower usage.

I think all the EJs have 4 valves/cylinder, for a total of 16. http://www.phrases.org.uk/meanings/10400.html

I appreciate your concern over the qualifications of some "professionals"; it's one reason I participate in this forum.However, this is a situation where it's not practical to DIY. See: http://www.bacharach-inc.com/refrigerant_recovery_faq.htm http://www.professionalequipment.com/refrigerant-recovery-tools/ While I don't want to be the "refrigerant police", here's what the EPA has to say about it:http://www.epa.gov/Ozone/title6/608/608fact.html