Project GD STI, Dyno Tuning for Big Power with Reliability


With E85 we were able to work our way up to 25 psi falling to 20, this was making 501 hp and 487 lb-ft of torque!  We tried turning up the boost more, but our ball bearing Gen 2 Precision turbo was maxed out and the compressor efficiency was falling off rapidly causing heating of the charge air temp.  We replaced our first-gen turbo Precision turbo with the newer Gen 2 when we installed the new engine.  The Gen 2 has a more efficient compressor wheel and did a lot better than the old first gen would have. The post intercooler charge air temp jumped from 125 degrees to 145 degrees as we turned up the boost, and the drop in charge density was not making up for the gains in boost pressure.  When turning up the boost more than this, our power actually fell.  We made our best peak power but you can see that the falling compressor efficiency is causing the power to rapidly decline on the dyno chart below.

The falling power at higher boost levels was to be expected because the Precision Turbo is a direct replacement mild upgrade stock position turbo and it is working like a champ, putting out power like a smaller rotated mount system! The power we are producing is about 60 more hp than what the turbo is rated for by Precision! We are very pleased with the performance of this turbo and we feel that it is an excellent drop-in upgrade and replacement for the stock turbo. It is the perfect turbo if you have a milder small stud, forged piston, and rod, open deck bottom end, 450 hp kind of build. Heck, with its low lag,  it’s a great stock replacement turbo too.  On E85, we are currently at 90 percent duty cycle so our injectors are at the limit and we cannot do anything more unless we go to bigger injectors. The good news is that our fuel pump is holding steady pressure and on paper, it looks like it should easily accommodate more power if we decide to go with more turbo in the near future.

We used the optional 20 psi wastegate actuator for our turbo and highly recommend this for you as well if you have a built bottom end. The large internal wastegate flapper needs the stiff spring to hold it shut at higher RPM.

To help prevent overheating when track driving, we made another map holding the engine to 17 pounds of boost, and surprisingly the engine still made 463 hp and 408 lb-ft of torque.  Note at lower boost the engine is making more power after the power peak! This is because the compressor is much more efficient at this boost level.  The waves in the power curve are because the small internal wastegate cannot precisely control the boost without a lot of overshoot.  With an external wastegate, we could probably get some nice gains from 4000-6000 rpm due to better control of the boost. Nevertheless, the car will still be pretty darned fast at this lower boost map.  All of the tuning and dyno runs were done on a 95-degree day and there was zero knock count so we feel that this is a pretty conservative tune.

The lower green line is a VA STI with bolt-ons and an OTS Cobb tune to give you an idea of how this dyno is compared to some others.  The newer version of the EJ257 has variable cam timing on both the intake and exhaust cams which helps the bottom end power some.  Our engine does not have much more lag or less bottom end than a stock STI GD STI at all.  At 3500 RPM it makes about the same torque as the lightly modded car makes at its peak.  At 3900 it makes the same power!

Our car is really fun to drive, in second gear it can break all 4 tires loose. it has a nice wide powerband and the Flatshift and Launch Control functions make the car super fast.  Sure a rotated mount bigger turbo is going to make more peak power and be faster but our car has a nice wide and friendly powerband.  Our car is a well-integrated, not quirky, balanced machine that would be as if Subaru had introduced a super version of the GD STI.

We have plenty of engine strength headspace with our IAG engine so if there are other stock position turbo options and larger injectors we will investigate, we want to flatten our torque curve and reduce charge heating, gaining power by improving efficiency, not necessarily just by increasing cylinder pressure.  We will try to maintain our friendly driveability and wide powerband when we make our changes. We are not sure if we are done yet but we are probably close to it!


Church Automotive Testing

Precision Turbo



    1. Funny thing, I was comparing results with a guy that had an ETS kit with a rotated mount G30-660 and the results were just about equal to our car and we were running slightly less boost. He made a little more power, we made a little more torque. This shows that at this power level, it’s not worth going rotated mount and that this turbo works pretty well.

  1. So it sounds like you basically ran the engine 500 miles then started doing power runs? What is your engine break in philosophy / understanding? The wide spectrum of ideas about “how it should be done” is confounding at the least. So I’m definitely curious to hear Mike’s break-in theory.

  2. I’m getting ready to do a very similar setup in my 2019 STI. How does that turbo drive on the street? What’s spool like in second? I have no access to E85 so that turbo at low boost looks very tempting on 93.

    1. Subaru’s are kinda laggy but it starts at 3500 rpm and is really pulling by 4000, look at the dyno chart and compare to the stock but bolt-on mod car. That stock car had variable exhaust cam timing so it had a little more low end but a similar year car with variable intake timing would have about the same lag as our engine.

  3. Love reading through your posts! Is there a page that has your full build list by chance? Couldn’t find one! Thanks!

Leave a Reply

Your email address will not be published. Required fields are marked *