GeneralDisorder

The EA82 feedback is a lost cause. Only fix for it is to throw on a Weber. Instead of giving yourself a headache, take it from someone that spent the time to understand how the feedback systems works and beat his head against that wall for several years on both the EA81 and the EA82 feedback systems. GARBAGE. The EA81 feedback *is* workable. It's relatively simple by comparison to the EA82 system (which isn't saying much, the system is still horrifically complex) and if you can bite the bullet on the relatively inexpensive O2 sensors, and the coolant temp sensors it can be made to work reasonably well. Sadly the EA82 system is far more complex, has more sensors, and can't be made to work well without a LOT of time and money. It's got one sensor (Manifold Pressure Sensor) that costs so much that it's worth it to buy a Weber just on that sensor alone. On top of that it's always the one that's bad - I pulled half a dozen at the junk yards here and all of them were shot. Learn from my mistake - chuck it in a dumpster. Or at least convert to a non-feedback Hitachi (and watch it devour choke springs ). GD

They should be installed dry, and the only ones to use are the Fel-Pro perma-torques. They do not require retorquing later (and haven't been sitting on a shelf for 20 years). GD

Pressure plate design is entirely different and the EJ pressure plate won't work with the EA transmission. GD

Definately not EA81. I know them well. GD

Well - we are going to need a lot more information to assist you. It sounds like you have some issues with fuel delivery (Hitachi carb right?), and possibly some ignition problems. We need to know if this thing is feedback or not for one thing. Fuel, Air, and Spark. Give us something to work with. What tests have you done - what was good and what wasn't? Do you know how to read a vacuum gauge? You can tell a lot from a few vacuum readings.... GD

The next time it acts up, perhaps try disconnecting the knock sensor. The knock sensor may be sensing whatever is causing the noises and retarding the timing enough to make the symtoms so pronounced that it's impossible to pinpoint where it's comming from. If the ECU doesn't see the knock sensor it will retard the timing only to safe (limp mode) levels and that may be just enough difference to locate the source of the problem. It's an idea anyway. If you can get it to act-the-fool again for you try pulling some plug wires. If both plugs on one bank don't affect the way it's running then something is wrong with the coil pack - since it's a bank-fired wasted spark ignition there's only two sides to the coil pack. I worked on a Hyundai recently that had the exact same symtoms. The coil pack had failed (only 130k on it) and it would work fine for a while, and then all of a sudden would miss on two cylinders. When it was running right it was as if nothing had happened. Being it was an OBD-II car it would throw a misfire code (to protect the cat).... but I don't think the OBD-I legacy can detect that. It ran very rough when the coil pack would act up, and had no power at all. Perhaps the differences in noise can be explained by the way the OBD-I's ECU handles the misfire - what it does to the timing, and the nature of the boxer design. Perhaps..... The amount of noise it's making though sure sounds aweful. The Hyundai sounded rough, but didn't sound like a bag of hammers. Pulling the plug wires one at a time while it's performing for you could also track down what cylinder it's related to. Perhaps a valve is sticking open. GD

"Bad Gas" is extreemely uncommon. As in when it happens, TONS of folks will be complaining, and it will make the news. I vote for checking on your knock sensor/circuit. GD

And I've never had a Subaru ignitor fail. It happens, but it's a lot more rare than the Honda's. And Subaru eliminated Distributors in '90. That's entirely based on your experience with the crappy dealers in Indiana. I've NEVER needed a VIN at my dealer, parts are cheap, and they are happy to print me off sections of the service manual if I need them. Also there's several guys here in the community that can order service manuals for us for cheap. And the dealer prices aren't really much more than other brands. There's a lot of good aftermarket manuals as well, and there's this web forum and others. I've never needed more than that - over 30 Subaru's to my name (6 right now) and never bought a factory manual yet. They are simple for people that can turn a wrench. Parts are cheap and plentiful - my dealer happily gives me wholesale prices on all the parts I buy, they are friendly, helpful, and some of the parts guys have worked there for 20 years. Perhaps your Honda dealers are better than your Subaru dealers, but out here things aren't quite the same. Never the less, even if I lived in Indiana I wouldn't buy a Honda. Not with the experiences I have had with them. They can't handle the nothwest driving conditions, their reputation exceeds their abilities to an extent that it abnormally inflates their value in the used market. I would rather spend the same money and get a used Audi - they are just a crappily built but at least are FAST and have AWD. Plus they are DIRT cheap used. GD

You did use OEM gaskets, and torqued the bolts to 12 ft/lbs yes? Make sure all traces of old gasket was gone (I use a wire wheel chucked into a drill motor)?.... clean bolts and chase threads? ..... just eliminateing possible hickups. Perhaps pop the carb loose and make sure the JB weld is still in place. At least that's easy to verify. Did you have to use multiple applications to fill the hole or did you use some screen to keep it from oozing down into the passage? JB can be a pain if it's on an angle or trying to fill something like that where gravity becomes an issue. GD

Does it throw any codes when you put it in read mode? Sure would be nice if the ECU could help us out a bit... I can't off the top of my head remember the OBD-I procedures, but if there is a "dealer check" mode it may report things that normal driving mode does not. The tendancy is to program them to be less strict with component diagnostics when they are not in "check mode" so as to not confuse and alarm the consumer Sounds ignition/wireing related by the intermittant nature of the issue. The coil pack you mentioned.... 1. How much did that seem to change it's behaviour? 2. How long before it came back? Abnormally long? 3. What type of replacement did he use? In my readings at legacycentral.org I've found a good many folks swear by the "Diamond" brand coil pack from the AWD manual transmission EJ22's.... the auto's were apparently equipped with an inferior brand pack. Another possibility - you said he replaced "one" idler pulley. What about the others? I'm wondering if a change in the timing belt tension, or the phase of the moon caused one of the other idlers that were not replaced to go sour. A bad idler bearing might come and go like that making a nasty racket.... Some thoughts anyway. What say ye? GD

How long did you let the JB weld set up? GD

There's a 93 Impreza on there in Hood River that needs the rear diff replaced (so he says). $400 - excelent deal. Fix it, sell it, and buy a nice legacy. GD

Interesting - all aftermarket radiators come with the AT cooler so they should fit an AT car no problem. The difference must be that the stock AT radiators have the cooler, but the *stock* MT's don't have one at all. They all fit the same though so as long as you route the AT cooler lines properly you should have no problem at all. Radiators get replaced all the time, and mixed and matched all over the place. Since no one has ever complained here that an aftermarket didn't fit their rig you should be fine with any aftermarket that comes with the AT cooler lines (they all do from my experience). You may have to lengthen the lines a bit, but that's no problem with a few $ of hose. That said, the 3AT is a peice of trash, so you might just consider swapping it out for a 5 MT if you really like the car. GD

It's only the engine section and the engine parts manual. Not useful for the rest of the car.... and not terribly useful otherwise unless you are into rebuilding EA81's. GD

Distributor vacuum advance is always present. It's the EGR and Evap. purge that are thermo-controlled to only activate once the engine is warm. The engine ALWAYS needs vacuum and centrifugal advance to run properly. GD

I wouldn't want the liability if I were a professional mechanic either. And frankly a good looking mid-30's nurse needs at least a Legacy. They make enough money.... As for the car - drive it till it snaps in half. That's what they do in Alaska. Usually it's the rear end that gets really bad and one way or another you lose the rear suspension. Either the whole suspension tube folds, or the rear coil-over punch through the wheel wells into the cabin. I've also heard of EA81's where the frame rails were gone like yours and the car folded in the middle where the B pillar is.... doors wouldn't close and the roof buckled. GD

Early Maxima alt's are the same power as the XT6 but bolt-in. Same deal with swapping the drive pully though. GD

That's not what I said at all. I said TRANSVERSE engine's are harder to work on. That goes for transverse inline engines as well as transverse V8's, V6, and so on. Alternators and starters are harder to get at, transmissions are harder to remove, timing belts are often nearly imposible to get to, and so on. Having 2 heads is no problem - V6's and V8's have two of everything just as a Subaru engine. And I very much enjoy working on longitudinal inline 4's - like the Toyota 20R, 22R, and 22RE. Same with the Samuri. Transverse means the engine is sideways with the transmission to the side and underneath. This layout sucks from all perspectives IMO. And most mechanics will agree with me. GD

There is a relay of sorts - it is NOT a typical relay. The block box you are refering to is a fuel pump control unit. It is designed to allow the fuel pump to run ONLY durring these conditions: 1. For a short time (3 seconds I think) when the ignition is first turned to "run". 2. When the ignition is in the "start" (cranking) position. 3. When the engine is running. To do this it uses the tach signal, and the ignition circuits. Internally it's got a timer circuit, and a relay, and some other circuits to supply power based on the tach signal. What I know of these is they are fairly simple, and pretty rugged. They are neccesary saftey items to prevent the engine from running after an accident. I have also yet to see one actually fail - although water intrusion would probably kill one as well as it kills any other relay. EA81's and EA82's with carbs (non feedback) all have them and they don't cause any issue. Try bypassing it and see if that makes the pump work with the stock wiring. At least that will tell you if it's internal relay is damaged. IIRC they also come apart pretty easily and you could check the guts and see if you can clean the contacts, etc. GD

All tensioners I've seen come compressed with a pin installed. I haven't bought any EJ series tensioners, but I don't see why they would be any different from all the other hydro tensioners I've bought. Always pinned. GD

Yes - that will tell you if the drivers side belt is good. Best way is to remove the oil filler tube and shine a flashlight in there while it's cranking. The inspection grommet on the cover isn't real helpful as it's common for the belts to strip teeth off them rather than just break. Verifying that the cam is actually turning is much better and easier anyway (2 12mm bolts). No compression is a good indicator. Either the timing has jumped a lot, or the head gasket is blown. Only way to really tell is to re-time it and check the compression again with known good timing. Yeah - no problem. GD

It does sound like a head gasket issue to me. Both cylinders are very low, and the 20 in one plus the presence of spark at the plugs indicates the timing belt is not broken and the cam is turning. I would say you are looking at a best case of about $100 in parts for a head gasket job, or a worst case of internal engine damage and you'll need another motor (ulikely). But it's not as if you can't find these engines under every rock.... I would say $500 is *ok* if the car is otherwise immaculent. Otherwise pass. I got a low mileage 94 Legacy (in Portland) for $750 that just needed a water pump. Craigslist is full of them weekly. If you get it and need some help I'm close, and I've done dozens of Subaru head gaskets. It helps to have someone with experience with these engines as they are a bit quirky in their design, and some important points are not immediately apparent to the newly initiated. GD

It's YOU that needs a reality check. You are basing your opinion on two vehicles. Hardly representative of anything. I know of plenty of Honda's that were just as bad as anything Subaru has poorly built. Or for that matter any other company. My girlfriend had a 2004 Accord - tranny went out at 78k. And I just helped a guy on the side of the road who's distributor died on his 95 Accord at 80k. He was a mechanic - has owned a shop for 22 years. He knew exactly what was wrong with it of course because it happens all the time with those engines. And if you think that anything transverse is easier to work on than a Subaru you are dead wrong. That's just plain silly talk right there. There's no Honda on the planet that's easier to work on than an EJ series Subaru. I'm a mechanic - I do this every day and have been since I was a kid. I've seen a lot of makes and models and the reason I DO NOT drive Honda or Toyota cars is because transverse vehicles are simply more difficult and time consuming to work on. And I hate FWD. Give me 4WD for traction, or RWD for fun. FWD is for grandmothers in Florida The "cost to repair" argument is baseless unless you live in an area where one or the other is not popular. There are some areas of the country where Subaru's aren't as plentiful - for me that would mean driving probably a Toyota truck or a Samuri etc. Here Subaru's are cheap and plentiful so that's what I drive. GD

Very bad idea to drive it. You have a 4 cylinder, 2.5 liter, Single Overhead Cam, Boxer engine. Your particular model (Phase 1 2.5) has a known problem with the head gaskets leaking coolant and oil. This WILL cause it to overheat, and should you continue to drive it, WILL result in the complete destruction of the engine. You are going to need to have the head gaskets replaced, and the vehicle will need to be towed to the shop or person that does it. The prices listed above are accurate so expect to spend around $1500 or more to have it done. The extent to which you may have already damaged it is what concerns me. When the temperature spiked all the way to the top of the gauge why didn't you stop driving it? You very well may have warped or cracked the heads at this point - which will end up costing more to fix. Probably around $250 to $300 per head (there's two) so may well put the car beyond reasonable repair costs. It may be cheaper to source a used engine if that's the case. Unfortunately there's no way to tell until it is torn apart and checked. Very bad indeed. Next time when gauges show irregular readings, PULL OVER and stop driving it. You're just doing more damage if you don't. GD

No - I haven't been looking on there recently. Just started a new job and haven't had time.... I would look it over real carefully though - owners rarely know what's really wrong with their vehicles. Especially on craigslist. GD