frag

Maybe I'm missing something but a CEL without a fault code seems pretty fishy to me. I would try replacing the temp sensor (cheap) before putting all that dough on a new fuel pump.

Your symptoms are exactly what mine were when I had a coolant temp sensor failure.

I dont know the answer to your second question. On my car which is OBDII, (your's is OBDI), i replaced the sensors and every thing was good immediately without any resetting. Dont know about your car.

Wait, I'm not so sure anymore... Just remembered that starting in 94 and on, these two parts dropped from the prices you quoted in your other post (and i dont know of a lower price than that but maybe others here will chime in) to 17$ each. One could easily infer that the sensors also changed dramatically. That's the parts (the 17$ ones) i replaced on my car (96 Legacy) without having to touch anything but the sensors. Maybe the older model requires one to do all that Chilton's says. Would be a good occasion for you to ask a dealer or your part provider what exactly is the différence between the pre 94 and the post 94 sensors.

I had spotted that error already. Chilton is absolutely wrong on that. Nothing easier than replacing both those sensors. Disconnect and remove one bolt. You just pull out and push in the replacement. Keep the old ones and mark them to identify. They are very similar parts and could easily be mismatched when reinstalling. I'm not absolutely sure one would fit in the place of the other but it looks very much like it could.

Glad to help but don't get too excited, maybe it wont work for you. The description of what happened after I did the O ring replacement is fact. The explanation of why is just educated speculation on my part. Hope it works also for you. Nothing space age about this repair. You remove the two bolts hodiing the fill tube with a long extension (10mm head bolts and no nuts). You then pull on the fill tube (it's made of plastic by the way) and you'll see the O ring at the small end of it. Replace, refasten, cross your fingers.

120$ to replace both radiator hose!!! And all the rest... Call the police!

990bw, at what pressure reading does one conclude that the cat is plugged? Or what is the normal pressure for a exhaust system?

I have had for a long time an oil leak at the oil fill tube where there's a flange just on top of the left valve cover. Nothing to have oil dripping on the ground but enoug to mess up the valve cover and adjacent parts. A few days ago I replaced the O ring there to stop the leak and here's what happened. The idle speed rose slightly to normal and got a little smoother (it was slighty low compared to what it was when I got the car), about + 65 rpm. OIl pressure at idle was thus a little higher by a few psi. Also it got rid of an annoying intermettent «miss» I had at idle one or two times a day: suddently the engine would act like it was firing on three cylinder's but only a few seccond at a tiime. This is gone. All I can think of is that this leak at the fill tube created a small vacuum leak that slightly lowered the idle rpm, a condition that maybe got worse a few seconds at a time each time the PCV valve opened fully. In case someone has similar symptoms, it's a 2$ fix.

Exotic... I think you're right in thinking the knocK sensor is a possible culprit, but I beg to differ about ... In saying that, you are factualy right. Indeed the check engine light is a result of a sensor not performing correctly. But it MATTERS much in my opinion. The last time I ignored a CEL (When i was driving a Loyale and was not aware of what it meant: my previous car did not have an ECU) i was left stranded, mercifully just BEFORE I entered a 2kilometer long bridge. It was for the coolant temp sensor. Modern engines are electronically controlled and the ECU controls them relying on info gathered from all those sensors. Most sensor's failure can at a point or other prevent the car from starting or have it stop functionning at the worst place or moment. I would never take a CEL lightly (pun intended)

It's a long shot and probably not the case, but i had once a temporary CEL after spilling some coolant on the cam position sensor. Since it's just under the oil fill tube, maybe they spilled oil on it. Like I said a long shot, but if you ever pull the code and it's for the cam sensor, it will be because of that.

I use only Subaru OEM (organic) pads and have no squeal and the disks are still A1 after 200,000 kilometers. HTHelps.

I'll describe it the best I can and hope you understand what I mean. 777 will prob do the same and in a better english. I slowly insert the screwdriver's blade (greased or oiled) between the seal's lip and the shaft, (the blade flat touching the shaft). I then go in just enough so that by pushing downward on the handle the blade part of the driver will exert from inside a pressure on the seal's rim (crown ?) At that moment the part of the driver closest to the handle is pivoting on the block just under the seal. What I recommend at that stage is to also put some lateral pressure on the screwdriver away from the shaft to lower the possibility of scratching it and making sure the blade tip is not gouging the seal's seat. Like 777 said, it would take a real hit or a heavy pressure of the driver's stem on the shaft to gouge it enoug to pose a problem for the seal's lip. Hope that's clear. Good luck!

About the seals, I know what you mean 777. I lost a lot of time trying a method i read about where you drill two small holes on each side of the seal, put metal screws in and pull on the screws with a couple of small vise grips. Did'nt work out for me. Before i was finished, i had to go to the dealer to get a part and just went into the working area and flatly asked a mechanic how he did to remove those seals. Exactly the srewdriver method you outlined in your post. One has to be careful though not to scratch the shaft while doing this. You have to pry upward and AWAY from the shaft. Only a mirror finish will prevent the seal from leaking again. Like yourself, I did'nt replace the idler pulleys nor the tensionner. They seemed to be OK. Hope I wont regret it. You did'nt say anything about the water pump. Did you replace it?

The probability of the oil pump's bolts (pump to engine block) being stuck enough to break upon removal are in my opinion very low. The environment is oily and protected by the timing belt covers. Same thing for the backing plate screws. Moreover, the torque value of these fasteners is relatively low. When I did mine I replaced the Oring and resealed the mating surfaces but did not touch the back cover's screws. They were already tight enoug. Having had a look gave me peace of mind.

It's a cheap part, easy to replace and gives you the opportunity to check the backing plate and reseal the pump to engine mating surface. Why take a chance when you'ere already in there? IMHO

I just read an article in the paper talking about how newer cars are less and less maintainable and repairable by independent mechanics and DIYselfers. I dont remember the exact figures but the last Volvo has around 15 different computers monitoring hundreds of parameters. And more and more manufacturers refuse to make public the coordinates of these computers. My question is: how is it with Subaru's 04 and 05 models? Do we still have only one Ecu for the engine's function monitoring and one for the AT or has Subaru gone the way of Volvo, GM and even Mercedes on this count?

Do you mean you DON'T live in a pineapple under the sea?!!

JaapH, if you have some tools for the Legacy that you dont need I might be interested in buying them. Cheers!

«Chamer?» Thanks in advance.

I switche at 140,000 kilometers and have no issue. I'm now at over 200,000 kilos. Not from personnal experience but for being a regular reader of what's discussed here, I think you should have at least one oil test made to know for sure if you're not extending the oil change intervals too much. Contaminants getting into the oil are the same be it regular or synth oil.

Cookie wrote I'll keep that in mind. Seems like a very good idea everywhere it's possible (crankshaft and camshaft seals would be good candidates on older cars). The problem with the Legacy's front inner grease seals is that they have two lips. One sticking outwards and making contact lateraly with the side of the c.v. joint housing and one, more regular, making contact around the axle's butt shaft itself (this is a place where my unperfect command of the english language does not help me...) You could'nt move that seal without crushing this outer lip against the c.v. joint housing. Since you seem to have experience with wheel bearings and seals, maybe you could shed some light on this question. I'm amazed at how little grease can be made to stick to these seals and I'm wondering how in heaven they can last so long. Do you usually pack the space behind them (between seal and bearing) full of grease or do you just smear some on the seal itself?

Cookie, I was not specificaly thinking of the Forester bearing problem. In that case the fact that ball bearings were used instead of rollers seems reason enough for them to fail even if the seal is not at fault. I was thinking of the Loyale and Legacy roller bearings. In their case I think the inner grease seals are often the cause of failure. They wear over time and replacing them at every axle change seems to be a worthwhile precaution. Just my opinion based on a couple of personnal observations.

Commuter wrote I found the same when I had the right front bearing replaced on the Loyale. I think this tells us what is responsible for bearing failure most of the time: a bad axle seal letting water enter the hub and bearing area. That's why I think replacing those seals when replacing the axles is good preventive maintenance.

22mm and normal thread.