Fairtax4me

First thing is to is get rid of those damn platinum plugs. They are a complete waste of money in that car. The $2.49 NGK standard copper plugs work the best. Gap at .039" Second get some new plug wires. The old 2.2s aren't too picky about brand as long as they aren't the cheapest of the cheap. Ive used Standard Ignition, and Carquest house brand in both of my cars and have no trouble with either brand. 100k miles is about the lifetime of a set of plug wires, so even if they don't cure the problem they're still a good investment for reliability.

Probably the alternator. Charge the battery and have the charging system tested (most parts stores can do both). That alternator can be bought remanufactured from Subaru for about $75.

Front O2 sensor should have a quickly changing voltage. Should also be between 0.1 and 1.0 volt. Or in the case of some sensors it should be closer to the 3 volt range. Are you sure you checked the sensor output and not the heater circuit? Little or no change in voltage means the sensor is bad or the fuel system is not cycling between lean/rich as it should. Usually its a bad sensor. Since the rear sensor is reading a quick voltage change, I think its safe to say the front sensor is bad. Have to have a lean and rich cycling for the cat to work properly.

Fuel level senders are wired in series. Generally the lower the resistance the higher the gauge is. You can pull the signal wire off of one of the senders and ground it to see what the gauge does. If the gauge spikes to the top (full)high reading = low resistance. If it spikes to the bottom (empty) low reading = low resistance. Figure out how the gauge works and then you know if you need to look for a short to ground between the senders and guage or a bad sender ground.

I've had my fair share of fun with worn out transverse arm bushings. Yes they will clunk and knock when they go bad. If you can move it AT ALL by hand, don't you think 3,000 lbs of car is going to have a LOT more push and shove when going around a corner or braking/accelerating hard? They are fairly easy to replace though. They just cost about $70 a piece from the dealer. Poly is a cheaper alternative, but they will start to squeak after a few months of heavy use on dirt roads. You may have to take them apart to clean and regrease them once in a while to prevent wear and noise.

Possible the old knock sensor has gotten "slow". It's also very easy to over tighten the sensor and crush the piezo element that creates the signal. This could make the sensor numb and not respond to a knock unless it is severe. Very easy to inspect and replace. Search for a thread called "Knock sensor 101" that covers removal, inspection, and replacement.

Kinda sounds like it jumped timing. Double check all of the vacuum hoses and breather hoses that go to the intake tube. Turn the Idle control solenoid back to its original position. The Idle control driver is not meant to be adjusted, that's a big no-no. The ECU will readjust idle on its own regardless of the position of the solenoid, except when its all the way one one direction or the other. With it all the way to one side the valve doesn't have the range that the ECU expects and it will start to set codes for idle control.

Could just be that the fluid level in the system was slightly low. Lots of places in the engine for air pockets to hide out, and it takes good circulation to budge them and push them out to the overflow. When the system cools it draws in coolant to replace what was pushed out. Put a little extra in the overflow and keep an eye on it for another couple weeks before getting too excited about bad head gaskets. Subaru recommends the use of their Coolant Conditioner in that engine (helps to prevent the small seepages from the head gaskets). Might want to pick up a bottle and add it to the system.

Oil and coolant down mix. If coolant drips into the oil it will settle to the bottom if its not agitated and you will get a puddle of slightly off color coolant and all of the oil floating on top of it. The green tinting in the coolant will not fade into the oil. Pour some of your oil into a bottle and pour a bit of coolant on top then let them settle out for a few hours and you'll see what the two do. If the mixture is agitated, the two blend together into a frothy disgusting mess that looks like bad chocolate milk, but it will still not be green. Some oils have UV dye in them to assist with finding leaks. Shine a black light or UV light on it, or look at it in direct sunlight and it may have a fluorescent/neon orange or green hue. Open a bottle of fresh oil (the same kind that's in the car now) and see if it has the same greenish tint.

Nitrous! I see now! Exactly how far does it need to retard? 9 degrees is quite a lot when the most advance at WOT is only what, 18-20 degrees?

Last time I checked prices for one of these Subaru wanted close to $500 for a reman. The after market places wanted $275-400. Cheap fix: grab a two wire plug and from a 95-98 and splice that in. The wire for the light is the same, the other two just end up in the fuse box (they may have even joined before getting there IIRC). Then you can buy the $75 Subaru Reman alternator for the 95-98 Legacy. Bolts right in, works great. Did this on a friends 99 foz a few years ago and its still chugging along just peachy. (And she was thrilled that I saved her almost $400)

CEL can be reset by disconnecting the battery for a few minutes, but this resets everything else as well. Or you can go to an auto parts store and have them scan the code and clear it. (I've heard that some will no longer erase codes due to liability reasons). Or you can get your own code scanner off eBay for about $20 and read and erase codes to your hearts content. You could spend a bit more and get one that reads freeze frame and live data for better diagnostic capabilities.

I never use lines on the belt. I always go by the marks on the engine. If they aren't there (covers missing) I use a triangle and a 3 foot level to set the timing mark straight up and down.

The crankcase is not sealed airtight, it has breather hoses and a vent (to the PCV valve) so pressure would not be able to build and prevent coolant from leaking in during a pressure test. How long did you let the tester sit? Sometimes a slow leak will not show a drop in pressure right away. Leave the pressure tester on overnight and see if there is a drop.

How bout this... What is the reason for wanting to retard the timing? If driveability isn't an issue you can just unplug the knock sensor and it should go into "limp mode".

You're SURE it's coolant? It's the same (or similar) color to what's in the radiator? And the coolant level in the radiator has dropped sufficiently to indicate a leak somewhere? Head gaskets seal multiple areas at one time. They seal coolant away from oil passages, oil away from the combustion chambers, and combustion chambers away from the coolant jackets that surround the cylinders. Any of these can have a leak between them without affecting another section. You can have coolant leaking from a lower jacket into the oil return ports at the bottom of the block (where oil flows from the heads back to the oil pan). With this type of failure the fire ring around the combustion chamber is not breached, so compression and exhaust gasses are not forced into the cooling aystem. This is the most likely scenario in this particular case. You can also have oil under pressure from the oil gallery being forced into the upper cooling jacket, which would yield a large amount of engine oil in the cooling system, but not much coolant in the oil, and again with no breach of the fire ring, there would be no exhaust gasses in the cooling system. The most common on the older DOHC 2.5 is the failure of the gasket between the cooling jacket and combustion chamber, which is what creates the bubbles in the cooling system. Other less common areas that coolant and oil can mix are in the turbo housing, and the O rings between the block halves.

Did you use the dash marks? Arrows only indicate TDC and can't be used for setting the timing belt because of the way the cams are loaded at TDC. The drivers side cam is loaded when at the proper timing position, but it will sit in one place on its own if positioned properly. If it slips either way, the valves that are open (valve spring pressure loads the cam lobes) will just close. More than likely the timing belt is off a tooth on one side or the other. Remove the side covers and compare marks side to side.

Higher compression. More power comes from higher compression, makes a more efficient engine. To do this they have to sacrifice space between the piston and the top of the combustion chamber. Usually by making the top of the piston flatter. They also tend to make the valves open a bit more (allows more air in, more exhaust out) which puts the edge of the valve further below the deck of the block, in an area that the piston occupies when it's at the top of its stroke. Not all of them are piston to valve interference. The DOHC engines can have valve to valve interference because of the proximity and angle of the intake and exhaust valves.

Timing belt interval is 105k or 10 years. However, these belts almost never break due to age or mileage. They break because one of the idler bearings or the water pump locks up and shred the belt. Probably a good idea to change those soon as preventative maintenance so you don't end up on the side of the road with bent valves. Auto trans, keep clean fluid in it and it should be fine for 300k. Tire rotation and proper inflation are a must to keep the AWD transfer clutches happy. Mismatched sizes, brands, models, or tread wear will cause excessive wear of the AWD and lead to problems. Any head gasket leaks are likely to be external on that year. Generally starting with small seepage on the lower rear corners. Some people had seen leaks there as soon as 35k miles, but it seems like they usually make it to 75k before they start to seep, and even then its not a huge issue as long as you keep an eye on the fluid levels and keep them topped off. If yours are seeping and you don't want to worry about them, the best time for replacement is when you have it apart to do the timing belt. Pull the engine, head gaskets, reseal the separator plate, valve cover gaskets, replace a few oil seals and o-rings, then you're leak free and no major maintenance for 105k miles.

Was the coolant just dark colored or sorta grey-ish? Oil will collect into drops and little pools and float on top of the coolant. It does not mix unless there is some serious agitation going on, but then it will be a nasty brown frothy layer (kinda like a wendys frosty) on top of the coolant. It will separate if left for a long enough period.

Heartless, your point was made by me several posts ago. I'm sure the risks are understood at this point. Cut the guy some slack. A decision was made, it has been done, and there is no changing it now. Before this thread gets any further off topic I'll suggest that everyone step back and take a breather before the Mods have to intervene. Capice?

Either of those could cause a no-start, but would not affect operation of the fuel pump. Best way to test those is with an O-scope to see if the signal waveform is correct, and by probing both sensors at the same time you can check to see if they are in sync. But most people dont have access to a multi-channel digital scope. (The only reason they would be out of sync is the cam and crank are out of time) So... you can check their resistance (spec should be in the FSM) to see if the coils inside are shorted. But, if the sensors have the wrong resistance, the ECU will probably set a code for that sensor. You can also check with a voltmeter, the AC voltage output to see if the sensors are producing a signal at all.

Tested working does not mean its currently getting power. Double check that you have good power and ground at the pump. If that checks out pull the pump and re-test it. Is the gas in the tank really old? Does it still smell like gas? Check to see if it burns. Water is heavier than gas so any water in the tank sinks to the bottom. Possibly the pump is just pumping water and burned out?

Sounds like it. Limp mode pulls back the ignition timing to a certain preset and just keeps it there. Feels like the engine has about half as much power as it should.

How did you find that?! Someone has to be the guinea pig... It has the same shape and reinforcement bars. Even has the arrow! Maybe its made from cheaper steel? Maybe its not zinc plated?