SlickDizzy
SlickDizzy UberDork
6/13/13 8:48 p.m.

Subaru XT EA82 turbo, ancient platform with little potential, yadda yadda yadda.

I am lucky enough to have an XT6 transmission with the car that will support 400HP+. I could really easily sell some of the rare bits off the car, drop in an EJ, and make a decent showing in the sub-$1000 class at the Challenge. Naturally aspirated EJ22s make as much power as the heavily modified EA82 turbo I already have, but are much less finicky and have a much broader torque curve. Not a bad option.

...or I could go all out and run an "NA-T" EJ and have something really, really, quick. I have a lot of experience with EA82 turbos and naturally aspirated EJ-series engines, but the prospect of an NA-T build is somewhat intimidating.

Early EJ-series motors will drop right in with an adapter plate and I am still familiar with the wiring after my old EJ22 Brat project from a few years back. In a few ways, it would make a lot more sense than fighting the fragile and incredibly complicated EA82 turbo setup currently in the car.

Has anyone here ever attempted this, or at least have any insight?

Ojala
Ojala HalfDork
6/13/13 10:18 p.m.

The main limiters are the heads, pistons, and fuel. 10:1 compression ratio and boost over 8lb dont live together for very long in a Subaru motor. But thats fine because the heads will only support enough airflow and the injectors only flow to support that boost level. What I have done in the hazy past was use a stock Subaru turbo along with a manual boost controller and a rising rate fuel pressure regulator and bigger fuel pump. The older JEC and Denso ECUs will work fine with a rising rate fuel pressure regulator at the boost and flow levels this engine will support.

Many people have also used M45 and M60 superchargers, but I cant comment on that because I have never done it myself. Also there are many documented build threads on RS25 about adding boost to older Subarus.

icaneat50eggs
icaneat50eggs Reader
6/14/13 7:33 a.m.

I'm a little lost as to what the point of this would be. I skimmed the xt thread, if you are not going to stick with the ea82, and go to the ej series, why not just start with something already turbod? Do you have a NA ej sitting around? By the time you turboed it couldn't you buy a JDM turbo, and not have the bad compression ratio?

SlickDizzy
SlickDizzy UberDork
6/14/13 8:27 a.m.
icaneat50eggs wrote: I'm a little lost as to what the point of this would be. I skimmed the xt thread, if you are not going to stick with the ea82, and go to the ej series, why not just start with something already turbod? Do you have a NA ej sitting around? By the time you turboed it couldn't you buy a JDM turbo, and not have the bad compression ratio?

The idea is to stay with the lowest budget possible, so I am just exploring options outside of an EJ20G swap. Plus, I'd hate to get a $1000 JDM motor and find out it's been abused or otherwise in bad shape.

Aspen
Aspen New Reader
6/14/13 9:05 a.m.

Turboing an NA EJ22 will most often end in tears. Sure you may save a few hundred. If you instead swap in a good turbo motor you could quite possibly end up with a good car that you actually like to drive and can depend on. Other people may even offer you money for it. I have had a turbo swapped Impreza for 9 years now and autox it regularly. Current day cheap prices make this an appealing route. You are also looking at 300+ hp vs a max of about 200hp with the former. If your tranny and clutch can handle it then get the real turbo. The turbo motors have better pistons, oil squirters, better cooling, better heads, better valves, better headers, etc that their n/a simblings.

SlickDizzy
SlickDizzy UberDork
6/14/13 9:15 a.m.

Sounds like a factory turbo motor may be the way to go. Figured I'd ask; I know a lot of Nissan SR20 guys rag on turbo'ed KA24s while they actually aren't that bad so I was wondering if this was a similar thing. Apparently it isn't!

fornetti14
fornetti14 HalfDork
6/14/13 9:23 a.m.

Could you drop the compression ration by swapping out a set of pistons?
Inquiring minds want to know.

Vigo
Vigo UltraDork
6/14/13 9:51 a.m.

I honestly think the whole 'dont turbo a non turbo motor' thing is a dinosaur from back when people didnt have real tuning or monitoring options.

Most n/a motors dont have such high compression that you just cant turbo them. IMO it doesnt even become a dealbreaker until you're talking like 11:1+. Ive turbo'd a non-turbo 9.x:1 motor and run 16 psi on it. I have a friend who dyno'd 440whp on stock 10:1 pistons with one cylinder misfiring. He broke the crank in that motor with detonation without hurting the stock 10:1 pistons!

N/a motors generally only have 2 special needs vs factory turbo versions of the same thing: They generally need to have the factory ring gaps opened up, and they need a slightly different timing curve. Once you get past that, i think the differences in reliability for 'mild' power setups are insignificant. You can break a forged low compression piston at low power by berkeleying things up, or you can make a stock non-turbo piston live at high power by doing things right.

Almost all stories about breaking non turbo motors have to do with mismanagement of timing and heat. You're not supposed to detonate on real turbo pistons EITHER, they just do a somewhat better job of covering your ass if you suck at tuning.

icaneat50eggs
icaneat50eggs Reader
6/14/13 10:08 a.m.

I got my ej20 twin turbo for $475.

penultimeta
penultimeta New Reader
6/14/13 5:28 p.m.
SlickDizzy wrote: a lot of Nissan SR20 guys rag on turbo'ed KA24s while they actually aren't that bad so I was wondering if this was a similar thing. Apparently it isn't!

Yeah. They're entirely different motors though. The SR is all aluminum, 2.0 liter while the KA is a heavy iron block, 2.4 liter, "truck" motor (as they're pejoratively called). But people get real power from the KA-Ts and, to me at least, a turbo KA is a much more interesting project than an SR20 swap. Check out KA-T.org

I guess, my point is that if a turbo version of a motor exists in relatively high supply, I don't really see an NA+T as a particularly economical route.

SlickDizzy
SlickDizzy UberDork
6/14/13 6:54 p.m.
icaneat50eggs wrote: I got my ej20 twin turbo for $475.

Dear lord, where?

icaneat50eggs
icaneat50eggs Reader
6/14/13 7:29 p.m.

Tiger jdm, I just checked they are currently 650.

Ojala
Ojala HalfDork
6/14/13 9:23 p.m.

The twin turbos are fairly easy to convert to single turbo if you decide to go that route. They are usually pretty cheap because they are getting kind of old and the twin turbo. the twin setup doesn't really work with left hand drive unless you made some fairly involved changes to the brakes and steering.

ValuePack
ValuePack SuperDork
6/15/13 12:07 a.m.

Someone say cheap? EJ18s are all but free in running condition, and the bottom ends are incredibly strong. Add dual port EJ22/T heads, should net you something like 8:1 compression(the numbers are out there, Google the hell out of legacycentral.org). Similar results are possible with single cam EJ25 heads on a NA EJ22.

icaneat50eggs
icaneat50eggs Reader
6/15/13 8:47 a.m.

I forgot the twin turbos are hard for this type application. Sorry to raise false hopes

Ojala
Ojala HalfDork
6/15/13 11:30 a.m.

In reply to ValuePack:

EJ18 and EJ22t are great...if you can find them in good shape.

There really are a lot of options and mixing and matching one can do on a Subaru motor. But many of the options are not as cheap or require machining. Putting a turbo on a n/a EJ is really not that hard and can be done cheaply as long as you keep the boost low. Turn up the boost and you need to address the pistons, injectors, intake, ECM, and etc.

Probably easier to turbo a phase 1 OBDI ECM vs the phase 1 OBDII ECM because it is so dumb. The gotcha is that the phase 1 crank and rods are fine but the pistons are cast. Keep the boost low and you wont have problems with the pistons.

Did I mention keep the boost low?

icaneat50eggs
icaneat50eggs Reader
6/15/13 11:33 a.m.

Microsquirt, megasquirts cheap brother ftw!

yilEKBsl7i5cmroDSMTIXBAI17lCiVa5