Project Gen 3 STi:

Testing the COBB AccessPORT.

By Wes Dumalski

As it turns out owning a project car is not easy! First and foremost these are real cars owned by real people that pay for them with their own money. Life as well as well as vehicle practicality and reliability have ways of getting in the way of project car progress and/or completion. Take our Gen 3 GR STi project for example... We had had the car repainted, put back to stock all just so we could start fresh... OK so that is a total lie, the real story is that the previous project car was sold by the owner after blowing up it's totally stock motor on more than one occasion. While I personally do not have all of the gory details the demise of the previous test vehicle put a complete HALT on our build and left unfinished business under our project car tab. As luck would have it I was looking for a new all around practical car in January of this year and that practical car turned out to be a 2008 STi. 

Rather than re-hash the juicy details that Mike already covered in the first two installments HERE and HERE let's get right down to business. That business is the stock tune. It for lack of a better term SUCKS! The power band is horrible with peaks and valleys that you can not only see on a dyno. graph but also feel under power delivery. Honestly I wonder how the car ever left the factory with this being acceptable. As I began to research the platform more (yes I started by reading Mike's previous articles) I realized I was not the only one that noticed this. Apparently MANY of the cars that had reliability problems were linked in MANY cases to the STOCK TUNE. Apparently in the name of emissions Subaru decided to do some funky things! The OEM tuning strategy was to lean the car out low in the RPM band between the crossover from closed to open loop. This lean condition was good at reducing emissions but bad for piston rings, especially if the car was modified in any way without a re-tune. Like many other cars there are always horror stories and while I would not consider the EJ25 the most durable of engines, if things are done correctly and conservatively it should hold moderate power just fine. In many cases the start of this project reminds me of Project 200SX where we were dealing with an engine that nearly everyone had written off and plenty of people lined up to tell us what was not possible, and in the end we proved just how much reliable power it could make! Also similar I am sure there are going to be S.P.E.'s who will tell us we can make more power by running more boost, running a more aggressive timing map or AF ratio, but alas that shit will fall on deaf ears. Just like our other projects our goal is reliable USABLE power and the FIRST step towards that goal was, believe it or not, getting RID of the factory tune. 


This sweet little handheld number is the key to your dreams...

Enter COBB AccessPORT. Chances are you have suffered head trauma or been marooned on an island and only recently returned to civilization if this is the first time you are hearing about the Cobb AccessPORT. But in the spirit of giving you the benefit of the doubt we will offer some fundamentals. The AccessPORT (hereinafter referred to as the AP) is a handheld module that allows you to interface with the OEM ECU through the OBD II port. The primary goal is to allow re-flashing of the OEM ECU calibration to achieve a desired goal. Notice that I carefully chose my words in the previous sentence, OF COURSE the primary goal is to make more power through more aggressive parameters, but one can also choose to flash a Fuel Economy or Valet map in addition to the numerous performance maps. That's right folks, in less than a minute you can cripple your car's performance to prevent a re-creation of the airborne Ferrari scene from Ferris Bueller's Day Off. In our case, the main goal of using the AP was not just to make more power, it was to keep our motor from shitting itself. I find it rather ironic that to avoid reliability problems we actually used a product that INCREASES PERFORMANCE but alas we will just have to suffer that side effect.  

Not being someone who qualifies themselves as ECU tuning savvy I was reluctant to mess with it until I could get on the dyno but upon receipt of the COBB AP I cracked open the package to see what it was all about. After reading the simple instruction manual (yes I really read it) I thought to myself 'hell even I can't screw this up". I connected the provided cord to the OBD II port, snapped the other end in to the AP and turned the key to ON. After a few seconds the AccessPORT identified the vehicle and presented some options. It comes pre-loaded with different MAP choices for various stages of modifications. In my case the car was box stock so I chose the 93 Octane Stage 1 MAP. After backing up the existing ecu data it installed the stage 1 map. I turned the key off to finalize the install and that was it. Seriously I did nothing else and guess what that few minutes resulted in?  

An increase of 53 lb ft of torque to the wheels. HP was also up across the entire power band with MUCH more usable power in the area under the curve despite a peak difference of 15HP. In addition to the large increase in power we also gained much needed reliability through less time spent at Stoichiometric due to zeroing out the Closed Loop/Open Loop delay. All of this assessment performed by my rear end diagnostic sampler on the street outside my house... OK OK so we used a real dyno with real numbers but we will get to those in a second...  

I did not install the AP at the dyno. It was done one evening in my garage when I had about 20 minutes to spare between work and family time. After doing the install I went for a drive and the first thing I noticed was that it was a completely different car. Gone were the flat spots and choppy power delivery and in their place was stump pulling torque that was instantly available and fell off much less as the revs climbed. In the week and a half I drove the car before I had the chance to dyno, I also noticed increased fuel economy by an average of 2 MPG. And this was on the Stage 1 MAP NOT the fuel economy map. Honestly I was amazed by the change and my posterior dynamometer told no lies, the power was real. 

The AMS facility is MUCH nicer than our posterior dynamometer!


Page 1 of 3 Next Page
Bookmark and Share
Tuesday, March 27, 2012 7:07 AM
I'm interested more on why you chose a dedicated TMIC cooling variant to most open element cooling that many other shops use.

I've heard shops explain that cooling off the engine is just as important as cooling the TMIC, but I'd like to hear your take on cooling the TMIC and Radiator "only" vs. opening up the hood and using two fans to cool the entire engine bay.
Wes Dumalski
Wes Dumalskilink
Tuesday, March 27, 2012 7:46 AM
If you look at the pictures we did not only cool the TMIC, there were two blowers (including the TMIC blower)and one large fan on the front of the car as well as an evacuation fan at the rear of the car.
Tuesday, March 27, 2012 7:56 AM
Okay, thanks for clearing my confusion.
Tuesday, March 27, 2012 10:36 AM
Great write up Wes.

I am assuming that the car was driven in "S#" mode when measuring the power increase after the flash and the calculation of increased fuel efficiency.

Is this a correct assumption?

Tuesday, March 27, 2012 10:41 AM
Good write-up, and yes, Subaru's stock map SUCKS. SUUUUUUCKS. It sucks to the point of being dangerous to drive the car on the stock map. People wonder why the STi doesn't perform as well as its numbers suggest, and that's why. I had a brand new 2011 STi and I went through the same process you did. I got the Cobb AP, and it easily increased HP and tq by 25, probably more. Then I had it dyno tuned by Shawn Church, a respected import tuner in SoCal, and he got a lot more out of it. In the end, I had 50 more wheel HP and 100 more wheel torque over stock, and this is with my only mod being the Cobb SF intake. And the curves were smooth, the power delivery was smooth, that's how it should have come from the factory. And this was on crappy Az/Ca/Nv 91 octane piss water gas!

Bottom line, if you have an STi or a WRX or are considering getting one, GET THE COBB AP, even if you do nothing else. The stock Subaru map is brain-punchingly terrible.

Tuesday, March 27, 2012 11:33 AM
Cool, glad to have a Project STI back, regardless if it's the same car or not. I did enjoy your clever intro on how you guys repainted & went back to stock. :)

Do you guys have all the old Project STI parts sitting around or is this a clean slate?

Curious about how you went about buying your car (I've always been told to be extremely wary/afraid of used turbo cars). Any tips?

@Xaqtly - how long did you wait to flash your STI?
Tuesday, March 27, 2012 12:30 PM

If it helps, I waited until I reached 1000 miles (my conclusion on average break in period) and never looked back since. To be quite honest I should have flashed even sooner knowing what I know now about the factory tune. Night and day after the flash too, very smooth power delivery and you can feel those irregular points that were through out the power band all cleaned up.

7600 miles here, I drive my car daily and in a spirited manner when I can. No problems yet and no excessive oil consumption yet.
Tuesday, March 27, 2012 1:59 PM
@ speedball3, I had the Cobb AP bought and delivered to my house before I took delivery of the car. I did a normal 1000 mile break in, didn't go over 4k rpm etc., but I did flash it at about 300 miles. I was that concerned about the stock map. After 1000 miles I romped on it and it felt good. My STi with just the Cobb SF cold air intake and the tune from Shawn Church, made 303 HP at the wheels, and 354 lb-ft of torque., also at the wheels. But more importantly, it was great to drive after that. Smooth and just a wall of power through the whole rev range.
Tuesday, March 27, 2012 4:23 PM
@ Xaqtly - This article is specifically about the STi but in your comment you said "...if you have an STi or a WRX or are considering getting one, GET THE COBB AP, even if you do nothing else. The stock Subaru map is brain-punchingly terrible."

Is the Subaru stock map for the 3rd gen WRX (non-STi) also so bad that you'll see significant knock on a bone-stock car running non-AZ/CA/NV 93 octane? I have one as my daily driver and am considering a tune (either off-the-shelf from Cobb or from a pro tuner) and I'd love to hear more about this from anyone with first-hand knowledge to share.
Wes Dumalski
Wes Dumalskilink
Tuesday, March 27, 2012 4:55 PM
The power and torque measurements and differences were measured in sport sharp.

The Mileage increase was a mix as that is how I drive the car. I had the chance to log several tanks on the stock tune for mileage and now several tanks since the tune. The increase in economy has remained consistent.

This car was a clean slate. No previous parts were left over or used. In choosing a used car I do the following.

1. Look for ANY sign of modifications:
Parts still on the car
loose or cut wires
paying attention to bolts, threads, coloring, and corrosion consistency

2. I get the VIN and have a different dealership run it and provide the vehicle history.

3. I spend time with the carfax in looking at previous title transfers and mileage between them.

4. I take a risk if I like it.

In the case of STi VS WRX the STi is notorious for having piston ring failures. Not that the WRX tune is any better but you may see less failures depending on if the WRX is a 2.5 or not. In any case I would consider the AP as Subaru appears to be using some techniques that may not be good for longevity or reliability in the interest of emissions.
Tuesday, March 27, 2012 6:20 PM
It's almost like Subaru is stating "If you don't reflash this car, you deserve what's coming."

I wonder what kind of impact this has on emissions? Are the requirements so difficult for the manufacturers to meet they have to resort to this kind of stuff? Obviously, even in CA the "after sale" emissions tests aren't as thorough (no checking how long cat light-off takes, etc) as certifying a new car. But I can't imagine the WRX's/STi's not passing with a reliable tune.
Tuesday, March 27, 2012 6:37 PM
Subaru is in a bad position with the EJ257 motor I think. It's based on the same EJ25 that was in the original '04 STi, and that was long before the government started cracking down on MPG regulations. Subaru is caught between designing a whole new engine (which is what they did with the new Impreza and BRZ motors) and trying to get the old EJ to meet all the stricter emissions standards and avoid the gas guzzler tax (minimum 23 MPG highway). Believe me, 23 MPG is optimistic in the STi.

So it looks like they've sacrificed power and good power curves to try to meet all these regulations for another year or two until the new STi comes out for 2014 or whenever. The new one will almost definitely have a variant of the FA20 engine in the BRZ, which can hit 30 MPG. Subaru is going to do what everybody else, notably BMW, is doing; smaller engine size, direct injection and turbocharging. And for the same reasons - emissions and federally mandated fuel economy minimums.

Still, the FA20 seems like a peach of a motor so it's not all bad, right? But yes, if you're getting a WRX or STi before 2013 or 2014 or whenever the new ones are coming out, then make sure you get the Cobb AP for your EJ25. It'll cure what ails you and your car.
Wednesday, March 28, 2012 9:44 AM
Hey Wes you seem to be using rings and ring lands interchangeably. They are very different. The stock RINGS are just fine.

As for what to do when looking at a used sti, a missing or modified turbo heat shield is a good thing to check for. cars brought back to stock are rarely put together with new shields and all the hardware for them.

You also should have a leakdown test done. I have seen cracked ringlands pass a compression test many times, and a car with a cracked ringland will generally run and drive fine. Subaru says burning 1qt every 1000mi is acceptable, but if you are going through that much, do a leakdown test. I suspect there are many, many people out there driving around with this problem that don't even know it.
Wednesday, March 28, 2012 10:15 AM
It would be cool to see a shoot out between other off the shelf maps, such as XPT, with the base Cobb AP maps.

I just drove to Las Vegas from LA in a stock 2008 STi. I drove in 'I' mode and stayed at the speed limits (65-70). I made 25.5 MPG one way and 27.3 MPG the other way. I have daily driven the same car around LA and OC in 'I' mode and get 22.5 MPG (average over about 8 tanks). You can get decent MPG, but you have to drive it like you robbed a bank and are trying to not get pulled over. Shift before 3k, cruise at 2k, downshift every time you want to accelerate so you don't bog the engine and very slight accelerator pedal changes.

I used to get 17.5 MPG when I drove it like you should drive an STi, but I wanted to experiment and save a few bucks for a bit.

Also, for those who want to stay stock (or modded), you can buy a Tactrix OpenPort 2.0 cable and hook your car up to a notebook. Then load a 3rd party utility called "Learning View", which can show you A/F learning and knock correction.

It's a useful tool to make sure you are not having the ECU compensate for a large swing in A/F correction or knock correction.
Wednesday, March 28, 2012 6:01 PM
+1 on "x01011000x" I would also love to see the average MPG on the economy map, I'm trying to convince my wife that I can save some money on gas if I can get the AP for my birthday. She thinks I'll never run it on "Eco" but would just keep it in POWER! What can I say she knows me :) Wes I would also love to see the use of Synapse Engineering's Diverter Valve on this project, Mike did I great article on it in a EVO IX, I would love to see how it stacks up on the STi.
Wednesday, March 28, 2012 6:33 PM
@y@ssel, Xaqtly, Wes, Jamal: Thanks for the good, solid feedback! This all helps restore my confidence in the 08+ STI and the hope that it wouldn't blow up if ever I got one. It's one of those "realistic dream cars" for me. :) Looking forward to seeing how this project progresses!
Wes Dumalski
Wes Dumalskilink
Wednesday, March 28, 2012 7:46 PM
@ Jamal
You are right and I will re-read and correct where necessary. I was and am aware that the rings themselves are not the issue, but when the lands shit the bed it typically doesn't bode well for the rings themselves hence the shortening... But we should be correct about the shortcoming and will check it out.

At this point I do not have a reason to change the stock valve. If I do ever change it it will be for the Synapse!

All cars have an inherent weakness and some reach that weakness easier than others. In the case of the STi it is more likely to break when mistakes are made, but when things are done the right way it should last just fine.... At least we HOPE it will ;)

Friday, March 30, 2012 12:57 AM
Wish that Cobb (or hell, anyone) would crack the toyota ECUs :(

Being able to tune the stock ecu or reflash sounds awesome!
Friday, March 30, 2012 6:18 AM
I also wish someone would crack the '01 Miata ecu. Trying to find a simple obd2 compliant solution for turboing that car is a pain.

What all this really makes me realize is how amazing is the ability to tune these cars. And how powerful the stock ecu is.
Saturday, April 07, 2012 5:25 AM
Nice job Wes, looking forward to future articles...
Salt Wagon
Salt Wagonlink
Sunday, August 12, 2012 7:19 AM
Well written, looking forward to seeing the product in Sept.
Wes Dumalski
Wes Dumalskilink
Sunday, August 12, 2012 7:08 PM
^ thanks rusty!
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:

© 2018 MotoIQ.com