DaveT

There is a related thread on tuning EA82 engines on this forum. There are links to many threads about the attempts and failures to get any kind of power increase with this engine. The ECU is the first gen OBD system, so very simple. Reading the blinks from the LED on the ECU is about it for diagnostics. It will set the CEL for blatant failure of a sensor or solenoid, but will not catch a sensor that is wrong but inside the normal range of data.

Back when EA82Ts were fairly common, I read lots of threads where people were trying to tweak them. They would spend lots of time and or $ to get maybe small gains while trading off reliability, which there wasn't a lot of to spare. Long ago, I decided reliability is more important. And if I ever were to spend the time to try to get more power, it would be through an EJ22 swap, because for the amount of time and $, you get far more HP gain AND reliability than any of the other tried and failed to varying degrees other options.

Look for used parts. Lots of parts for these models are NLA, so, keeping an eye out for parts cars, extras from when someone decides to move to something newer, etc. The nut should be around 150ft lbs torque, and then cotter pin. It's also good to go somewhere, then re check after 10 miles or so, just in case. The differential is just a plain open diff, so if one side has no load, it spins, and the other side gets no power.

The 2WD vs 4WD power routing is done in the transmission, after the clutch. The most common won't move in 2WD problem is a stripped hub - due to the axle nut not tight enough. Theoretically, a failed CV / axle shaft, or a differential problem could also cause similar results, but I don't remember seeing a thread about that.

I have not had to troubleshoot one, but yeah, check those connections. Measure the voltage from each to the body, and across the 2. The car will have to be on to get meaningful readings.

There are wiring differences between 1990 and 1987.. when you go back further, the differences get more significant the further back you go.

I haven't had enough different CTS failures to have a solid answer for that. I did have another problem once that seems similar - the car wouldn't idle. Had to hold the throttle to keep it idling. It was also intermittent. Turned out to be the wire in the engine harness that goes to the Idle Air Control Valve. If that coil doesn't get drive, or it's dead, no [or not enough] idle air.

The CTS testing ok only at cold temp doesn't mean it is good.

In general, for EA82s, you want to start collecting spare parts whenever you find them. Many are NLA from Subaru or the aftermarket.

I just took a look at my 2. And remembered that the TPS I took apart was on an intake that was off the engine. I would unbolt the A/C compressor, and swing it out of the way if I had to get one off an assembled engine. The hoses are long enough.

One way the CTS can fail is to read incorrectly when hot. If the engine temp is not reported correctly to the ECU, it will not trigger the injector to supply the correct amount of fuel. The result is running poorly. A cold engine needs a lot more fuel to run well. A hot engine will not run well with that higher amount of fuel. I had a CTS fail so it randomly varied making all sorts of drivability problems. But it never went out if range to trip a code.

Here is my compilation of CTS data - Subaru CTS temps.rtf

I check the CTS with an ohmmeter. Run the engine till normal operating temp. Shut off. Log the temp immediately, and several time on the way back down. If you don't have one of those IR thermometers, a meat thermometer works, just have to mechanically connect its sensor to the block. Fill one of the unused holes with grease, stick it in kind of thing. I have a few threads with temps vs ohms I made from values I collected of the years. That other small sensor looks like the one on my CA car - it just verifies that the EGR is actually opening by measuring the heat in the passage.

A bad CTS can cause major drivability problems. And also, not nesisarily trigger a fault code. Fuel pressure is critical also, T a gauge into the line after the filter under the hood. Iirc, it has to be 21 psi minimum.

Check the coolant temperature sensor also. I have not had a problem with a TPS. I took one apart recently, they are incredibly well made. Been running and maintaining EA82 powered wagons since 1988.

+1 regarding the green connector and test mode.

Check all the fuses? Or at least that the ECU and coil gets power.

Ok... It's better. But something else is also contributing. Slight intermittent skip at half to full throttle is when it showed up. Maybe half as bad as before cleaning and testing the TPS.

Ok.. I decided to check the TPS first. Found one of my spare engine wiring harnesses, and separated the throttle cable. The pins slide out the back of the big round connector when you use a jewelers screwdriver to lift the locking tab. Got one of my spare TPS units. Carefully opened it up, these things are well made. 50 years of designing and building electronics with potentiometers, you get to know a well made pot.... Anyway, it looked good, so I cleaned a little glue off, and reassembled it. Trying to check for a flaky pot with a digital meter, as usual, is nearly useless, so I thought about it for a minute.... The easiest way to find a noisy pot is to listen to audio going through it, so I rigged connections from a music source through the TPS, to an amplified speaker. It worked great. Also checked the throttle closed switch, good. Bring the test rig up to the car, and plug the connector onto the TPS. Instantly, I have no ground continuity, music at full volume! wiggle the TPS connector, crackle, no volume, crackle low volume. Well, there's a problem! Never would have found it that quick probing a meter on the pins, since the probes would poke through any oxide, etc. I used my tiny wire brushes to clean the pins, cardboard to clean the sockets in the cable end, a little contact cleaner on both. The test connector now works as well as it did with the reference TPS. Volume goes up and down smoothly with no crackles with throttle position. Tomorrow, I'll take a drive to verify if this is the total fix.

No check engine light. No trouble codes. I forgot to check those before my first post, but yeah, none. [it's been a while since I had a failure that caused a CEL / code without some other obvious clue.] Injector wires, sounds good to check, TPS I was thinking of checking also. I keep thinking it's a 30yr old potentiometer, it could be worn out. Regarding the post from 6star - if the poor running is engine temperature sensitive, check the CTS. Search for my post about CTS, I have listed several values & voltages for good ones. They do fail in ways that do not trip trouble codes.

I've been hunting the cause.... I've got a new fuel pump, and regulator with a gauge I can read while driving - the fuel pressure is good. It will climb a long hill from a stop, floored, 2nd gear [3AT] but occasionally stumble once the RPMs get up. Not constant, just a glitch now and then. Sometimes on acceleration from a stop, a little trip before it gets going.

On your last picture- If you remove the bearing holder, you will have to re set the bearing pre load / pinion & ring gear alignment. Maybe if you carefully, mark the position, and count turns while removing you could get it back where it was. I have not tried this. I have read the FSM directions for assembly, it doesn't seem trivial / fun to do right.

If ypu never had the intake off, or it wasn't put together with anti seize, run it till it's up to temperature before loosening the bolts. Saves snapped bolts.

One of the intake manifold gaskets could have failed. Or the weird o ring under the throttle body. Either could supply enough coolant to make a big smoke show.

I never saw a 1 v pulley on an EA82. At least 1 pump I have, the hub for the pulley can be pressed to whatever height is needed.