1999 Legacy L Engine Swap Choices

[Mrrlyn]

I know this has been gone over countless times on so many sites, but I'm trying to run down the checklist (so I don't get killed by the wife)

 

Car:  1999 Legacy L

Engine/Tranny: EJ2.2 II - toasted, 4eat

Driving conditions:  Lots of hill climbs and eventually a small tear-drop trailer.

 

What I am looking for is feedback on my options: 

 

1st Choice: I can go with a 00-03./04 Forester/OBW (?) 2.5sohc (01/02 best?)

2nd Choice: I can go with a straight replacement - rebuilt EJ2.2 II

 

So, with the desire for a bit more torque for the frequent hill climbing and potential for future light towing, I really would like to drop in a 2.5.  Both come with 3 yr 60k warranty/same cost, so if there did happen to be a HG issure, it's covered.

 

What I know/expect to have to do if i do go with the 2.5:

Keep 2.2 Intake

Y-pipe from 2.5

...

 

So, what am I missing?

[Numbchux]

Assuming your 2.2 is a phase 2, the intake manifold should swap right over and make the project very easy. In that case, I would definitely go with the 2.5.

[idosubaru]

If it were me I'd get an EJ25 short block only and bolt your EJ22 heads to it (or buy new/other ones if those are toast).

then you can use the EJ25turbo headgasket - those never fail like the non-turbo gaskets do.  that's the best gasket to use on EJ25's...so if you're using an EJ25 block you'll need EJ25 headgaskets so i assume you can use them on an EJ25 block/EJ22 heads combo.

GLoyale had a place in Oregon building $600 EJ25 short blocks.

 

that avoids the issues i'm about to mention (which are all rather simple), better end product with you choosing the headgasket, and probably cheaper.

 

EJ22 is single port exhaust , EJ25 is dual port exhaust so you'll need an EJ25 exhaust manifold.  it bolts right up to the existing exhaust system just fine.

 

intake manifolds swap for Phase II's (and phase II's ONLY for future readers that will say they're confused LOL)

 

there are two different timing trigger mark set up around 99-01-ish era.  you can look and check, or just swap the EJ22 crank and drivers side cam sprocket onto the EJ25 to be sure.

 

what headgaskets are installed on the rebuilt engine?

[Mrrlyn]

I'm not sure on which head gaskets they use, but I was either going to go with Enginesus.com or SSI.  After looking at some reviews of SSI, and I think they are out of the running (although they have a A+ with BBB).  Enginesus.com sells/describes the gasket they use on their website, but no brand is given.  Whatever it is, they are confident enough to offer a 3yr unlimited warranty.

 

I actually have two 2.2's at home, the one in the car and the original before the swap (don't ask me how the swapped one got fried, it's embarrassing).  I was going to send the one that seemed to have valve issues back as the core, and keep the one that's in the car now (lower mileage too), as my local subi guy probed it when it was running and said it sounded like a crank bearing, all bottom end - oil starved.  On that engine I might do the 2.5 short swap/frankenmotorize, and keep it around as a spare or swap it in if I have a long weekend.

 

So, that's the plan.

 

As the engine is coming as a long block only, I shouldn't have to worry about cam timing, since it's all going to come from the donor motor.  Is there any reason I should request a particular year 2.5? Just the EJ251, right? Or would a 253 work as well/better/worse?  Since i'm using the 2.2 intake, I don't need to worry about MAP vs MAF, right?

 

On a side note, does anyone have suggestions on the 4eat 'downshift under load' problem that is out there?  I don't think it is in every model year the 4eat was used.  If you're not familiar with it, during hard accelerations, up hills, when the car goes from 3->2, it drops in pretty hard.  Feels like the power comes back in before the converter re-engages completely (however those torque converters do it) and gives you a kick in your pants.  I had heard a rumor that Subaru had a modified tranny chip that would fix it, but I don't know the veracity of the rumor.

[Fairtax4me]

I'm not sure of the phase 2 2.2 engines, but I know the SOHC 2.5 from that era leaves something to be desired when it comes to low end torque.

It does make more overall torque, but it has to run up to about 4,000-4,500 rpm to get there. The 2.2 makes its torque at a lower RPM range, which may suit you better for towing and hill climbing.

I would also check into a set of reground camshafts from Delta, to help bring the torque band lower. These make a noticeable improvement in the phase 1 2.2, I'm sure they would make a difference in a phase 2 2.5 or 2.2 as well.

[Mrrlyn]

I have been able to find a purported stock 2.5 sohc v 2.5 dohc dyno graph, but no luck finding anything for 2.2.  The 2.5sohc appears to have reasonable torque pretty low, and is almost flat until HP peak.  I'm not sure what dyno this was run on, but I've seen this report style before, so maybe someone knows it?

 

Does anyone have/know the torque curve of the 2.2 sohc? If it really is significantly better, I may go with that (the 2.2) and then build the 2.2/2.5 with some Delta cams to provide the torque.  I can't imagine that the 2.2 BEATS the 2.5 in the lower rpms, but I could buy that maybe its higher in comparison to mid and high rpms. 

 

Part of my deal right now is I want to have this car be reliable for several years, covered with the engine warranty, so I'd be going stock-only on the engine.  I would think I could throw a more open intake and exhaust (obviously the double vs single if I go with the 2.5), but it'd be hands off until we get a bigger car (kids and all that).

[idosubaru]

EJ25 = 14% more displacement, 16% more power, 12% more torque.

 

peak torque:

EJ22  149 @3,600

EJ251 167 @ 4,000

EJ253 167 @ 4,400

 

the graph above shows a maximum torque drop over a span of 400 rpms (the rpm difference in max torque readings between 2.2 and 2.5) of 8 ft/lbs (and that's in the exaggerated high rpm  area well over 5,000 rpm's).  or only 2 ft/lb in the 4,000 rpm range. extrapolating that data, if it's accurate - the 251 would have to loose 20 ft/lb torque from 4,000 down to 3,600 rpm's to equal the Ej22...seems unlikely?

[Mrrlyn]

Thanks grossgary!  That is the kind of thinking I needed.  Even if the 2.2 does 'better' at lower RPMs, it still doesn't have as much torque there, it's just peaks quicker.  Sounds like the 2.5 will be the choice for me...

 

Unless this part is true...reading on another forum, don't remember which one now; the point was that if you were changing engine volume on a MAP sensor car, you have to swap ECUs (and maybe other stuff) because the fuel tables are hard coded in, and it doesn't learn.  This is in contrast to the MAF years which calculate it on the fly.  Does this sound correct to any of you?  It's funny that they kinda went backwards going from MAF to MAP.

[Mrrlyn]

Anybody?  Let me rephrase my last question:

 

'99 Legacy with MAP sensor: Does this mean that any change in engine volume means a new ECU (as the air fuel density is hard coded in MAP editions)?

 

It seems that a lot of people say its just a 'plug and play' kind of thing, after bolting on the intake, but I'm wondering if they're all running really lean, with that 15ish% bump in volume...

[hankosolder2]

Here's my thought... the O2 sensor is going to adjust the mixture. Unless the difference in the long term fuel trims is outside of the maximum window allowed, it will work OK. The mixture might be a bit off when you floor it and it goes open loop, but I wouldn't be surprised if that is scaled based on the long term fuel trim values "learned" during closed loop operation. 

 

This is speculation, not based on actually having tried this. You could ask someone who has done the swap what their long term fuel trim values read to see how close you are cutting it.

[Fairtax4me]

The fuel tables on the MAF based cars are technically hard programmed as well. They all start out the same and the ECU uses MAP and/or MAF readings to determine air density and go from there.

How do you suppose vehicles that normally travel above 3,000 ft altitude handle the change in air density there? Pretty sure ECU has enough "wiggle room" to adjust for such a small change in displacement.

[Mrrlyn]

How do you suppose vehicles that normally travel above 3,000 ft altitude handle the change in air density there? Pretty sure ECU has enough "wiggle room" to adjust for such a small change in displacement.

 

Very good point!  Short of actually hearing from someone who's done the swap, this makes sense and makes me feel better.

 

On the other side (of the engine), since I'm going to be having to swap the header, I might as well swap something that flows well, get a bit more out of the engine.  Any suggestions?

 

*Edit: Nevermind, went with the Borla, so I can still mount the CATs and pass emissions testing.  Now for the other little bits - tranny cooler, free flowing intake and such.

Edited August 19, 2013 by Mrrlyn