WoodsWagon

Sad update to this, according to a thread I ran across on Pirate4x4, "Garner" passed away. http://www.pirate4x4.com/forum/showthread.php?p=14073502#post14073502

Giving advice that will cause more issues down the road is cause for being rude. Compromising high tension wires by poking holes in the insulation is a bad diagnostic method. That sounds like rod knock in your video. If it was piston slap it would do a steady smack smack smack at idle. Floating the RPMS around 3k and having the noise get worse or getting a back-rattle as you let the RPMS come back down are often signs of rod knock. A flex plate will often clank at idle too. My friends WRX sounded like yours does for about 5 months before it finally spun 2 bearings. He was hard on it too. So you may have some time before it dies.

I'd be more suspicious of a failing ignition switch than the seat heater draw. If the starter solenoid isn't getting enough power from the ignition switch the draw of the starter motor will pull system voltage down to the point the solenoid releases. I think the chances of a battery being marginal to the point that seat heaters being on or off would affect it to the point it couldn't crank is pretty unlikely. Put it in park/nuetral with the parking brake on and try running a jumper wire straight from the battery termial to the starter relay contact (a spade connector on the back of the solenoid) while having all accesorys like lights, heaters, and blower on and see if it cranks.

There's a dropping resistor that's used for controlling the shifts. If it's removed or died, you get harsh shifting. It doesn't usually cause hunting though, but worth a check. I think it's mounted on the passenger side strut tower, but it would be best to check a FSM.

The ECU usually grounds outputs to things like A/C when it wants to turn them on. Power usually comes from a fuse, through a relay coil, and then gets grounded at the ECU to complete the circuit.

You can weld the center diff and it won't do anything to the VC. With the center diff no longer able to differentiate, then the VC will never see any slip so it will do nothing.

With a speed-density system (which is the strategy a MAP engine uses) the VE and displacement of the engine are critical parts of the programming put into the ECU. It will adjust the fuel trims using the O2 sensor, but the open loop and idle programming will be wrong. Speed-density is dumb fuel injection. It takes a guess on how much fuel to throw in based on the tables it has stored in memory for the size and efficiency of the engine it's fueling. A MAF measures the actual amount of air getting pulled into the engine, so it makes the fuel injection much more adaptable to changing engine paramaters. I'm sure your engine swaps run fine, but they aren't running right. Even with the tune 12% off, an engine will still run well enough, but the 2.2's are running rich if they're put where 2.5's were on the MAP system. I'm sorry, but on the point of MAF vs MAP being equal for engine changes, you're just plain wrong. A phase 1 SOHC is also a lot worse flowing head than a phase 2 SOHC, so it's not just displacement differences that throw the speed-density calculations off, the engine also doesn't flow as well for the same manifold pressure. It's good to know the cam pulleys are swappable between phase 1 and phase 2.

I hate to say this but I don't think the joints you used as an upper ball joint are rated for the side load your planning on putting on them. It looked like you were originally planning on using a lower ball joint from a toyota possibly? That would work much better because it would be a ball being pressed into the socket by the load. The joints you have now are a ball trying to press its way through a ring, and I think it will wear out quickly and pop through if the ring doesn't shear off the shank first. You see upper control arms done exactly the way you have your done pretty often, but they aren't load carrying. The just take the forces of keeping the knuckle upright and alligned, but don't support the weight of the vehicle. With your coil over attached to the upper control arm, that joint now carries the load of the vehicle.

A 99 SOHC 2.5 is a phase 2 already, probably out of a forester. I'm not sure Phase 2 SOHC cam pulleys fit properly on to Phase1 cams. You should call that guy in Conway. It's like a 2 hour drive each way, but he's selling an 01 2.5l with the 5spd trans for $500. It sounds like the combo you need, and looks really easy to pull out because the nose is already off the car.

They're speed density GD, so the computer will be expecting a 2.5l with a certain volumetric efficeincy, and it will be fueling a 2.2l with less VE to boot. It's not like the phase1's which are all MAF, and automatically compensate for the smaller motor. It's not a good swap, especially when he could sell the motor out of the wagon for $300, the body for scrap for $200, and buy the right motor to go into the car. It's not like he's in backwoods Alaska trying to cobble something together with whats available, subaru's are everywhere in the northeast.

Is it a 2 door coupe or a 4 door sedan? You'd be better off leaving the 2.2 in the wagon and just driving that. Swapping the wiring to the 2.5rs destroys what value it has left. There's plenty of good running phase2's out there. You'd be wasting your time to swap a phase1 in when you can readily find phase2 donors in junkyards.

The Samba is very much about keeping the cars true to OG specs. I'm on there because of the 68 Type3 Fastback I had. A much better site for engine swaps is http://www.shoptalkforums.com/

Why? Because they're better built, get better mpg, and are more fun to drive on the street than subarus of the same vintage?

Yes, but the 2.2's are harder to find and there's no advantage to them over the 2.5l. My cousins well cared for 2000 2.2 outback sport chucked a rod through the block while still full of oil.

Here's what a 95 RR HLA head looks like:

A 2000 2.5rs is a phase 2 SOHC 2.5l. Its pretty much a completely different engine than a 95 EJ22. The 95 EJ22's are great for swapping in place of the phase 1 DOHC 2.5l's in the 98 and older cars. You could sell it for probably $300 for that purpose. The wiring is different, the cam and crank sensor tooth patterns are different, and you'd have to butcher the hell out of things to get it to work. Not worth it. Just find a phase 2 EJ25 out of a 99+ forester or 2000+ legacy or outback. Here's one in Conway NH you could pick up. http://nh.craigslist.org/pts/2840903944.html He wants $500 for both engine and 5mt, so you could probably get jus the engine for $350.

A screwed up outer CV joint could do that. Does it shake the wheel while you hold it in the same spot while turning or is it hard to turn the wheel, then easy, then hard again? There's a pair of u-joints between the steering column where it pokes through the firewall into the engine bay and the steering rack. Soak them in a penetrating oil, not WD-40, and drive it for a bit. If it fixes the notchy steering for a bit, you need a new coupler. If it doesn't change, you probably need a new rack. The curb impact may have chipped a tooth or something inside the rack. Also check that the coil springs are turning smoothly with the strut as you turn the wheels from side to side. A seized strut cap bearing will make the spring wind up, then jump with a "bong" noise. That usually causes a steering pull to whichever direction you turned last though, not notchy steering.

Rear wheels wells are where the rust starts on pre 2000 legacys. The back end of the rocker panel, the corner of the wheel arch above the bumper, and the strut tower in the rear all rust out. The fuel tanks and filler necks rust out, as do all the brake lines and fuel lines under the car in back. 2000-04 legacys seem to rust through on the top of the rocker panel inside the back door opening.

I welded on the fender lips from an early 90's Chevy Cavalier: They looked better after I painted them. I got the 1/4 panels for free because they were bent in the front, that's the hammered area towards the front of the flare.

I've had multiple EA82's past 8,000 rpm with no issues then or after, so overrevving a tad won't hurt them. Does it smoke worse on long turns like highway onramps?

Is there anything really wrong with the engine in your car? The EA82's are fine. They all leak oil, most make some noise from the hydraulic lash adjusters in the heads, and they're gutless, but the one thing they are is reliable. That is, as long as you change the timing belts every 60k. Every single SPFI EA82 I've seen in peoples backyards or in junkyards has still run. It's rust or the 3at transmission that takes them down, not the engine.

There is an oil galley plug towards the rear of the engine under the intake, but I think it's on the passenger side. Uses an aluminum crush washer to seal.

As long as you go to loosen the bolt while it's still hot, yes a welded nut should get it.

Heat is key. It loosens locktite and rust. I never use an EZout without heating the drilled bolt up so it glows before putting the EZout into it. An EZout is weaker than the head the bolt had originally, so if you sheared off the head or rounded it over, the EZout isn't going to work without you heating the bolt up. With rounded over bolt heads, the first tool I reach for is my welder. By welding a nut to the head of the bolt, you give a new way to grab it with a wrench and you heat the bolt so it's easier to take out.