Project GD STI: Getting More Power with Precision Turbo and Cobb Tuning! - Part 1

by Mike Kojima

Our STI has been driving around as a daily for two years now with all of the typical bolt ons.  It has been reliable at 318 whp with good driveability and fuel economy on regular pump gas, in other words a hassle free car that is very easy to live with.  Of course if good was enough, more is better and we could not leave good enough alone.

One of the drawbacks of the EJ257 engine is that the bottom end is somewhat fragile and getting more power usually means more boost. We were close to maxing out the stock turbo and not willing to give up the daily driving convenience of 91 octane pump gas.  We did not want to simply turn up the boost but rather make more power with the same boost and less backpressure so the turbo could run close to a crossover situation.

With lower backpressure there is less reversion and less retained heat in the combustion chamber, the engine is happier, runs cooler and tolerates more timing which can also result in cooler conditions in the combustion chamber.  Less backpressure can allow a cascade effect of more power with less stress on the engine as a system.

It was time for an upgraded turbo.  We just didn't want any turbo either.  Yes we know that a rotated mount turbo system is best for power but we wanted to keep the stock location due to heat management and keeping some semblance of a stock appearance.  We wanted a turbo upgrade that was a direct bolt in.

After studying the market, we found that Precision Turbo made what we were looking for, a totally engineered from the ground up stock replacement ball bearing turbo with the capability to make up to 500 hp.  The Precision Turbo fits 2002+ WRXs, STIs and Foresters.  We called up Precision and soon had a test unit in our hands. 

Read more about Project GD STI !


Other stock position turbo upgrades involve carving the stock compressor housing and stuffing larger compressor wheels in the place of the stock part.  Yes this gets you more airflow and power but costs efficiency as the the larger wheels mess up the B height and diffuser configuration of the housing.  Less efficiency means more shaft power is needed to make boost and backpressure is created by the turbine when extracting the extra needed power from the exhaust flow. Less efficiency means more charge air heating and more lag as well. Precision's turbo has a CNC machined compressor wheel with a 55mm inducer.  The wheel is machined from a 2618 low silicone aluminum forging.  2618 alloy is known for it's toughness and ductility.  Machined wheels generally mean thinner compressor wheel blades and hub sections which further improves wheel aerodynamics and efficiency, Precision call this CEA or Competition Engineered Aerodynamics. The bigger compressor wheel is in an appropriately sized bespoke housing that was engineered for it.
The bigger turbine wheel is also in a housing designed for it with a larger A/R for greatly improved flow.  The turbine wheel is cast from highly heat resistant MAR-M alloy.  The wastegate flapper valve is much larger to prevent boost creep. The nib on the flapper valve is an anti rotation feature as wastegate valves tend to spin and saw themselves loose if not constrained.  Most direct drop in turbo upgrades are a larger wheel stuffed into the stock housing which again looses efficiency without optimizing flow.  The housing is cast from stainless steel for it's anti rust and low thermal conductivity properties.  The turbine shaft/wheel spins in a dual ball bearing floating cartridge.  The ball bearing speeds spool and transient response and the floating cartridge damps out harmonics for longer life and quieter operation.  The bearings are protected from coking by a water cooling jacket around the center section. 
To replace the turbo, Howard Watanabe first removes the intake charge pipe and the charge hoses to the turbo.  He also removes the oil feed, drain and water lines to the stock turbo.
Howard then removes the belly pan and removes the up pipe, unbolts the turbo, downpipe and exhaust manifolds.
Page 1 of 7 Next Page
Bookmark and Share
Monday, December 07, 2015 10:31 AM
Guys, you should use fender covers when working on cars.
Monday, December 07, 2015 11:44 AM
20hp and extra lag down low, not seeing any benefits to this upgrade if you are only going to run 91 octane.

Retune it for E85 and let's see the real potential of the turbo.
Monday, December 07, 2015 12:01 PM
i've been wondering what happened to this car. this article reminds me again why i'm glad i don't have to use that squirrel piss you guys get for gas.
Monday, December 07, 2015 12:18 PM
We've got the same junk here in CO. People think you don't need it because of the altitude? I don't see many carbs on the road anymore, not seeing the relevance there.
Monday, December 07, 2015 1:56 PM
judging from the brakes and coilovers, this is a track car, prolly not only track, but mods are track focused.

On track, it will never fall below 4k rpm unless a mistake is made, and it that case its not a big deal. The biggest improvement is that it makes this power at lower boost and less back pressure, meaning less heat and stress on the engine (little power increase on the top end isn't bad either), increasing reliability and decreasing chances of power loss from heat soak...

For a purely street car, this mod on pump gas only is more or less a waste of money, but for a car that see's regular track duty its a pretty good investment.
Monday, December 07, 2015 3:21 PM
Track or no peoples' objectives vary so to say this turbo is a waste on a street car is ignorant. I've seen far more aggressive setups on cars that will never see a track because their owners want to drive flat out.
Monday, December 07, 2015 3:36 PM
New Turbo, New Header + Retune + Installation costs... for 20hp.... track or street it's a waste.

Mike Kojima
Mike Kojimalink
Monday, December 07, 2015 3:44 PM
Its pretty obvious this turbo is a lot better, did you not read the article? More power than the stock turbo at 3 psi less boost. Way better transient response. 300 rpm more lag, that's not not even feelable lag and not in the normally used powerband. We even explained why the turbo made less boost and what we are gonna do to get the boost back to where it was. Once the boost is at the same level as it was with the stock turbo, it will make at least another 20 more hp, at the same boost level! Thats 40 more hp with the same fuel and same boost. Once we build the bottom end and raise the boost some more it will do even better.
Monday, December 07, 2015 3:55 PM
in this specific case, if this was a street only car, this setup would be a waste of money as the power gained is minimal. But this isn't a street only car (at least I hope not)... but as Mike pointed out, its an actuator away from more boost and power. As far as more aggressive setups for the street, doesn't mean its not a waste of money. There's nothing you can do on the street that will overwelm decently maintained stock STi brakes. For a street car that upgrade would be a waste of money.

for a track car, its not all about power. Reliability and longevity are important, temperature is important, making the same power on lap 17 as you are on lap 2 are important...
Mike Kojima
Mike Kojimalink
Monday, December 07, 2015 4:44 PM
I am paranoid because the stock EJ257 is pretty fragile on unforgiving 91 octane. I know plenty of guys who nuked their engines.
Monday, December 07, 2015 6:02 PM
@Supercharged111: It's not about carbs, it's about air density at high altitude and not needing high octane for most engines because of that. It's hard to get enough cylinder pressure to knock without forced induction.
Monday, December 07, 2015 6:39 PM
As a California resident I've been to maybe 20 track days in my life, and the only catastrophic engine failures I've seen have been 2.5L Subarus. 3 to be exact. Your caution is well placed Mike. This looks to be a great mod while tip toeing around our shite gasoline.
Tuesday, December 08, 2015 10:27 AM
With new headers, uppipe and turbo, why didn't you want to go to a twin scroll setup? Extra cost and complication? Wouldn't be able to keep the stock turbo location?

Also you brought back memories of that downpipe to transmission brace rattle. It really got going under engine braking...
Tuesday, December 08, 2015 11:44 AM
Mike Kojima
Mike Kojimalink
Tuesday, December 08, 2015 1:39 PM
I would prefer a twin scroll but I don't want complications like dual external wastegates, etc.
Mike Kojima
Mike Kojimalink
Tuesday, December 08, 2015 1:39 PM
Water meth and flex fuel will be coming once I build the bottom end.
Wednesday, December 09, 2015 4:32 AM
I love to see the ignorant comments on here day in and day out.

It's all about power but who cares about reliability right?

For a road course car this is an excellent upgrade. Less boost, less stress, lower temps!

The increased lag will also help put less stress on the motor.

I am using the new AEM Progressive Meth injection kit on my car to supplement the 91 octane. It really helps with charge air temps, especially at buttonwillow.
Wednesday, December 09, 2015 2:07 PM
Of course you made more power with less boost pressure...you used a bigger turbo and a larger diameter intake charge pipe. Comparing boost pressures between two different turbo setups is not useful for power discussions and is misleading. The amount of power you make scales directly with mass air flow, not necessarily with boost pressure. If you turn the boost up on the stock turbo, you expect to make more power because the volume of the system is unchanged and the boost pressure went up, so the mass flow will go up and thus the power will go up. If you change the turbocharger, the volume of the system has changed, so comparisons of boost pressures between the two systems are not a good indicator of relative power levels between the two systems. The same boost pressure on the two different systems does not represent the same mass flow, because the volume is different between the two systems. PV = nRT. SportCompact Car wrote an article about this years ago. With Dave Coleman on your staff, I figured this would've been caught.

Also, the STI is a car that already suffers from lag issues. This is a matter of opinion of course, but if you ask me the power comes too high in the rev range for street duty. If you really wanted to do this car a favor, you would make it reach peak torque lower in the rev range.

You added an intake, new bigger turbo, and full exhaust and got a professional custom tune, and only made 30hp. AND you moved the power curve further to the right. Is the peak power number better? Obviously yes. But I fail to see the return on investment in all the parts and labor. You could get 30hp from an STI with a downpipe and an OTS Stage 2 map. Maybe if you're building a track-only car, this setup is "better", but if you just want a faster street car and you don't have connections to Cobb, I seriously doubt this is worth the money.

I'm being critical here, and you can call me a hater if you like, but I love reading MotoIQ and I've learned to have high expectations of your articles. I'm just trying to make sure you live up to your motto... "More signal, and less noise."
Mike Kojima
Mike Kojimalink
Wednesday, December 09, 2015 6:56 PM
300 rpm lag with faster transient response feels like less lag when you drive it, at least to me. I drive this car everyday and would not say it has a lag problem at all as I stated in the article. Power created at a given boost level is useful when you talk about work and system efficiency. The turbo is turbine machine and power at a given boost pressure is a good way to look at overall system efficiency. Which includes VE and TE changes due to less work having to be recovered and cylinder conditions . I am better at turbo stuff than Dave. I could write a white paper on it but nearly everyone will get lost. Yes this works well and yes thus is worth the money under the parameters I am building it for. Stock bottom end 91 octane reliable and no E85 which is not conveniently avalible in my area. When I raise the boost to simular levels to before as noted in the article the power gains will be quite a bit more. I suggest you read this article more carefuly and the rest of the ones in this series that were engine related. The car already has a downpipe and other stuff and made really good power on Cobbs dyno for a bolt on 91 octane car before the turbo.
Dan DeRosia
Dan DeRosialink
Thursday, December 10, 2015 5:44 AM
... if you change the turbo, the volume of the system hasn't necessarily changed; the engine only takes in what it takes in, volume-wise. And lag isn't the same as boost threshold. And ugh. I'd personally go for having my WRX make boost at 4k rather than 2.8ish, but that's me.

And boy am I glad I don't have to deal with Cali "91" octane gas.

Mike, have you thought about something like Turbosmart's 2-port IWG actuator and a 4-port solenoid? I know they're offered for some stuff, not sure if there's something that will go to the Precision bracket. The idea of using pressure on both sides of the diaphragm, ala EWGs, always made a lot of sense to me.

Been wondering how this turbo would do. I know Precision certainly did their homework on production quality on the housings, but I'm never sure how much to trust compressors and turbines from companies other than Garrett, BW, Mitsu and IHI. Precision seems to have a pretty good rep, but it's not like many people do A->B testing on comparable products; I'm sure we can both think of lots of stuff that had great reputations that turned out to be not that well engineered, along with stuff that is as good as its made out to be. Don't get me wrong, I'm not straight out knocking Precision, I just wish there were available turbo maps.
Friday, December 11, 2015 1:14 PM
@Fabrik8: A little late now, but seeing how the computer recognized the less dense air and adds timing, you're right back on the threshold again. Back in the day of carbs, distributors were not smart enough to do this on their own.
Mike Kojima
Mike Kojimalink
Friday, December 11, 2015 5:08 PM
We added timing to the map, not the ECU.
Sunday, December 13, 2015 10:21 AM
Of course it "only" made 30 hp more on 91 with less boost. Mike your the man don't listen to the whiners
Tuesday, December 15, 2015 9:33 AM

Good article by the way. I actually read it and know that making a "only 20hp!?" post is dumb.
Wednesday, December 16, 2015 8:43 AM
If you live in ca like I do you register your car in South Dakota (which is legal even if you've never been there.) no smog!
Mike Kojima
Mike Kojimalink
Wednesday, December 16, 2015 12:42 PM
Details on how to do that?
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:

© 2018 MotoIQ.com