NorthWet

There is no "overdrive" function. It sounds like your transmission is not shifting into 3rd (high) gear, and the usual cause is a problem with the governor. Either the valve is sticky or the drive gear is worn. Check out this repair article: http://www.indysworld.com/80s/general/USRM/Transmission/Governor/ATgovenorclean.htm Other common issues involve dirty/burnt/low ATF, and failed vacuum modulator. These are less likely to cause your problem.

Less base compression means lower thermal efficiency when not super-atmospheric (<1bar). *Edit - OK, it lowers TE throughout, but it will be noticeable when you are off-boost.

87 and up should be interchangeable, but connector may be different.

If it smells like water/moisture without the glycol smell, it could also be just a clogged ducting drain line allowing evaporator condensate to back up. Glycol smell kind of clinches it, though.

I agree, just unplug old, plug in known-good controller. The case is not used as a ground, so no need to mount or even ground it. (My 92 Leg used a plastic case with slide-fit controller board.) Forgot that wagon's connectors were behind the kick panel... 2-doors are hidden next to rear seat...PITA to get at. If the connector cleaning and/or controller swap doesn't help, the controller can show you its error code(s). IIRC, this involves jumpering 2 connectors behind left-side kick-panel.

If you still have your SVX, then the controllers will interchange between the Leg and the SVX. IMHO, it would be better to rule-out the controller before you start pulling interior trim hunting for a bad connection or limit switch that is not working right.

The controller located by the speaker may only be for wagons. Most of my experience has been with 2-door vehicles (XT and SVX) and one Leg wagon. The SVX are trunk-mounted and the XT is below right-rear window. The connectors have been an issue in the SVX, so are likely to also be an issue with the Legacy: Bad contact, possibly due to loose connector engagement, despite having lock-tabs.

Your problem may be simple (e.g. - a fuse, a limit-switch, or a connector), but i suspect that it is seatbelt-controller related. I have come across several failed/failing belt controllers in the early Subarus, and it seems that most of them are related to failing solder joints on the power relays within the controller.

*Edit: My apologies to Indrid Cold, as I did misread/misunderstand what he said about shaft collars. Sounds like something very useful, and I am going to look into them. endedit* EEEK! Maybe I misread Indrid Cold's post but it sounded like the advice was to pound on the inner race of the bearing... NEVER do that, unless you are discarding the bearing. All pressure/force should be applied to the race that you are trying to fit to the housing or whatever the bearing is being pressed upon. Look to see if the housing is scarred where the bearing used to be. Quite often, someone has used something harder than brass and has scarred the housing.

Flywheel/starter depends on how you do it: Starter is relative to the tranny housing, and flywheel teeth its relative to the engine's crank flange (and, indirectly, the engine case). Can't remember offhand which side the clutch fork goes through.

I have been lucky so far with a 16oz ball peen, but that thick brass punch is needed. Steel punches will tear up the housing, and thinner brass punches don't look too good afterwards, assuming that they survive at all.

When going the other direction (EJ engine to EA tranny) they add a 12mm (or 1/2") spacer between the two. What it sounds like you want to do would require removing that much material (12mm or 1/2") from the spacing between the engine block and the tranny body... it would have to be cut from the bellhousing sections. If this is not done, then the input shaft won't engage the clutch and the pilot bearing. I imagine a possible (but sounds dangerous and bad engineering) way would be to add a 12mm spacer between the crank flange and the flywheel. This sounds so bad that I hesitate to mention it, but figure it might show how hard it would be to try something like this. (I think that this would also require some clutch fork reengineering)

I was astonished how well that clean/polish/seal kit worked on my SVX. The lenses looked almost brand-new afterward; very clear. What type of cleaner were you using? You NEVER want to use a cleaner containing ammonia (like most glass cleaners) on plastics as it will cause crazing on many of the common plastics.

Over the last 30 years, I have bought several JDM pull-outs and have been satisfied. I deal with companies local to me, and they offer enough warranty time for me to have a good feel for the engine (or tranny). I feel that most of the larger importers depend on word-of-mouth and repeat business, so they try to ensure a decent product. (Some have large containers full of rejected parts.) I would not hesitate to buy from my current company.

That should get SOMETHING out... might be you!

This situation is more common than you can imagine. I have lost out on several deals because the seller was talking about wht he was selling and I would point out that the belt issue was probably the reason. Honesty has its drawbacks. GD might be opinionated (usually correctly), but his technical advice is excellent.

The purpose of either valve is to relieve the pressure on the turbo's intake impeller when you close the throttle. The sudden backpressure can cause a drastic slowdown of the turbo core, increasing lag when the throttle is reopened. (It also can cause some physical issues with the turbo.) The recirculation valve will route the intake flow back between the MAF and the turbine's impeller, keeping the metered-air mass within the system. (Non-MAF systems don't care and can use a simpler BOV.) At the low flow/pressures of the stock system, there is not a lot of effect from closing the throttle.

Regarding "Gen3" heads: There is a lot of talk about how they are better and less likely to crack. The only thing that i have seen for certain is that they have different casting identifiers. I make a distinction (when I remember to think before talking/typing ) between what I know and what I think that I know. i do not KNOW that the heads with what we refer to as "Gen3" heads is any different or better than any other casting run; I do not think that anybody else KNOWS this either. (No offense to those that believe that they are better... especially you, Rob ). To the best of my knowledge, nobody has opened up these castings and done a comparison as to which looked better designed. AFAIK, people look at the "Gen3" heads and think that they have not encountered as high a rate of exhaust-port divider cracks as with earlier casting runs. "Earlier" might be a very significant word. It might just be that gen1 and gen2 casting runs have had more miles, more heat cycles, more cooling system issues, and/or just plain more abuse than the castings from the last year that they were made. (I am assuming that Turbone is correct on where to find gen3 heads.) If this is the case, then these heads are LIKELY to still be your best bet. But on an individual, case-by-case basis, that would not guarantee anything. Or they might be superior castings. My advice (glad you asked for it! ) would be to concentrate more on the cooling system and its hoses, and an intercooler to keep intake-charge temps down rather than go on a Grail Quest.

OT: It is common to use core designs for different applications. The 4EAT is also known as the R4AX-EL and EC8. The R4AX-EL core is used by Nissan and others. The 3AT uses an M41/M41A core design that was also used by at least Nissan. The OP might have a good reason for wanting an automatic. For instance, due to joint issues, my wife finds it painful to drive a manual transmission in town. Other than the truly poor governor design of the 3AT/M41, it is not that bad of an automatic. Very few automatics seem to last 200K, so the inability of the 3AT to make it that far is only bad in light of the general longevity of the rest of the drivetrain.

They had a tendency to stop working. i think the general suspect is a voltage regulator or similar semiconductor blowing out in the dash.

(OK... First off, please pardon my non-precise terms for the transistor terminals. Their functions are probably more meaningful to most than the various names given to them based on hole-conduction theory and internal construction.) You are showing voltage at the "input/power" terminal of the transistor, and then you see pulses at the "control" terminal? When the pulses appear, what happens to the voltage on the "input/power" terminal? Does it go to zero? Does it sag a bit? (Might be hard to tell with pulses anyways.) Come to think of it, are you seeing voltage pulses on the "signal/control" terminal of the transistor? I think that the ECU grounds this terminal instead of giving a voltage pulse... pretty typical of power applications. Regardless, the third terminal of the transistor will be it's ground through it's bracket mount to the body. If this ground is poor then you won't see much if any spark... and you shouldn't see the voltage sag on the "input/power" terminal. Something you may want to consider is checking the coil terminals for steady voltage on one side and pulsing voltage on the other. (A dwell meter might also be useful.) The basic circuit would be: Ignition-switch power to one side of the coil, the other side of the coil going to the "input/power" terminal of the transistor, the "output/ground" terminal of the transistor being grounded to the chassis, and the ECU sending grounding pulses to the "signal/control" terminal of the transistor based on signals from the distributor. BTW, and this may be a TOTAL false-trail, but I recently came across a no-start, seemingly no-spark situation after an engine swap. Both recipient and donor car had been sitting for many months, but both ran prior to storage. We were getting a single spark for every approximately 10 that we should have gotten. (10 rotor rotations yielded 4 sparks, evenly spaced.) Replaced disty, no joy; replaced coil/exciter transistor/bracket assembly, no joy; replaced ECU and it fired right up. These were all known working parts prior to sitting. ECU might have only required reseating the connectors, but that is 20/20 hindsight.

You really don't want an actual Blow-Off-Valve (BOV), as it will dump already-metered air to atmosphere and screw up your mixture. A recirculation valve is what you want. However, unless you are planning on increasing your boost to broken-engine levels (i.e.- above the stock 7PSIG) than it is not really needed.

IMHO, this is pointless: It will only make the car run cooler at low throttle, and will hurt mileage, emissions, and engine life. Even if you stay at 170 degreesF at cruise (shudder!), a second or two of WTO will use up your "reserve" temperature buffer. Turbone is a good source for what has worked and what has not. I do feel that using a pre-87 engine in a 87-and-up chassis will provide you with many interesting hours trying to get things working the way that you want. Others have done the switch from digi-dash to analog dash, and their write-ups show much frustration... BUT they were doing spot-rewiring instead of wholesale work as you have planned. (Might be easier for you.) I have my own views of the exact cause of the exhaust port cracks, but it basically does follow what Turbone said. Make sure that your cooling system is tip-top and always full. My OPINIONS follow: If there is a simple way for you to get the turbo water-jacket nominal "supply" line from somewhere else (maybe all the way from the disty-side head?) you MIGHT lower risk to head and HG. (This line does have to come from a physically lower point on the engine than the nominal "outlet" going to the T-stat.) It might also be beneficial to not dump the hot turbo-drain-oil back into the non-disty head; however, there may be few if any practical ways to accomplish this. There will be plenty of nay-sayers. As long as you are aware of the difficulties and seeming impracticality of what you are doing, go for it! Logic and playing do not mix well. Play on!

The ingested oil lowers the effective octane of the charge. (Engine oil decomposes into detonation-prone molecules at a lower temperature than does the gasoline.) RE: kit - yeah, but the only time I remember the kit is in the middle of long right-hand sweepers...

My guess on the "ticking" is detonation caused by the ingested oil. My SPFI pings heavily when it ingests oil due to "righthand sweeper" issues on the SPFI version. Follow GD's suggestions. BTW, it is easy to overlook gunk-narrowing of the metal nipples on the valve covers, which I believe can cause oil-pumping into the PCV system.