posted on February 12, 2013 19:00
Project EVO X GSR- More power with Garrett, Full Race and Cosworth
By Mike Kojima
In the last edition of Project EVO X, we upgraded the fuel system in anticipation of more serious mods. Well the time for more serious mods has come and we are going to see exactly what can be done to the 4B11 with bolt ons. In this go around, we are going to install some camshafts by Cosworth, Garrett's direct bolt in upgrade turbo and a tubular exhaust manifold by Full Race Motorsports.
Click here to read more about Project EVO X!
|We obtained some of Cosworth's MX1 4B11 cams with matching springs. The MX1 cams have 266 degrees of duration on both the intake and exhaust with 10.8mm of intake lift and 10.4mm exhaust lift. With the tuning freedom of having MIVEC variable cam timing on both the intake and exhaust cams, we were anticipating a best of both worlds situation where we would gain power across the board.
|Before installing the cams, we sent them out to be WPC treated for reduced friction and longer life.
|We also used Cosworth's single valve springs matched to our cams. The springs are made of Japanese Kobe steel which is super trick electro purified and vacuum degassed steel. These processes eliminate inclusions and pockets within the steel which are the major cause of valve spring failure. Many high end valve springs around the world are made of Kobe steel for this reason. The Cosworth springs have double ground and tapered ends to reduce fatigue and are shotpeened for long fatigue life. The springs are good for 8000 rpm which will allow us to bump the rev limiter.
|After installation we did some tuning with Naji Dahi of Looney Tuning. After correcting the boost curve and A/F ratio we had to retard the intake cam and advance the exhaust cam to reduce overlap because the car was sneezing out of the intake on part throttle due to our previously programmed aggressive MIVEC map. Once we were done tuning, the results were impressive. The Cosworth cams got the turbo spooling earlier, greatly reducing lag. We had huge gains across the board with peak gains of 17 whp and 28 lb/ft of torque. Unfortunately, a few days ago Cosworth USA announced that they were halting operations and that the parts programs future was uncertain. We really hope Cosworth continues to stay in the market. The market needs quality well engineered parts. Cosworth bigshots, are you listening?
|With impressive gains from the cams, we next installed Garrett's GT direct bolt in upgrade turbo. This turbo is based on the proven GTX3071 ball bearing center section turbo with Evo X specific housings and internal wastegate. The turbo kit is very complete with water and oil lines included.
Wednesday, February 13, 2013 4:53 AM
Looking at the Garrett turbo from the outside, it doesn't seem like it retains the twin volutes of the stock turbo, but maybe they just filled it in for ease of casting? It would explain the major loss in low end if it's single scroll.
Wednesday, February 13, 2013 6:17 AM
Looks like a divided housing to me.
Wednesday, February 13, 2013 6:49 AM
It says it's a twin scroll or divided housing several times in the article and points it out in the captions for both the turbo and the manifold. There is even a paragraph which touches briefly on how twin scrolls work to reduce lag.
One thing I want to say is the turbo feels LESS laggy than stock. It simply zooms through the rev range where the dyno says its making less power. The turbo is really responsive and the car feels really fun to drive now, much better than stock.
When passing on the freeway, you don't need to downshift anymore, just roll on the throttle. Around town the car scoots. It is much more than the dyno chart suggests.
Wednesday, February 13, 2013 8:03 AM
Maybe try loading it at 2000rpm start point to see if it's some weirdness under dyno conditions?
Wednesday, February 13, 2013 9:15 AM
^ I agree. I know it moves through the lower RPM quickly and as such perhaps the loss is less noticeable, but that is a large amount of power and torque to not feel it.
Wednesday, February 13, 2013 10:01 AM
@ Steve & Wes: You guys are forgetting that these curves were not generated on a dyno but rather with the virtual dyno software that Naji uses. It would be interesting to see if the curves did alter any under dyno conditions but I think in the past Mike has mentioned that the virtual dyno results track fairly closely.
Wednesday, February 13, 2013 11:47 AM
If I were going off of the dyno results alone, that turbo wouldn't seem worth it. Large loss of low end and it is running more boost as well to gain the extra up top.
Wednesday, February 13, 2013 12:24 PM
The turbo feels like it has less lag than stock.
Wednesday, February 13, 2013 12:44 PM
Wrecked - "You're not thinking 4th dimensionally Marty!" If looking at things right at the moment, the Garrett might not seem to be worth the dollars spent, but as more and more is done to the car, the stock turbo would've only hurt and hindered future work.
Wednesday, February 13, 2013 1:11 PM
Boost threshold isn't lag either.
Wednesday, February 13, 2013 1:39 PM
@Wrecked, the car is only running ~20psi of boost at redline right now. Just turn it up to 28 psi and see what happens ;) There's another 100whp in the turbo with better fuel whereas the stock turbo is peaked out. I actually think the boost could be turned up to 23-24psi on 91 octane which would give another ~30whp or so.
I've driven an Evo X with the Garrett GTX3071 and the larger 0.94 A/R housing which adds about another 500rpm before boost comes on (This particular car was making 500whp on a dynojet with E85. Needless to say, it was very very fast). Even then, you still don't really notice the lag. If you're in traffic and intend on making a move, you're probably in 2nd or 3rd gear sitting at 3k or 4k rpms already. Lastly, looking at the fully modded car as it sits compared to the completely stock car and tune, they spool up the same.
Wednesday, February 13, 2013 2:01 PM
Good points related to boost threshold. Driving my old school Evo I rarely floor it below 3000 rpm or so.
Wednesday, February 13, 2013 5:46 PM
What was the seat to seat overlap, if any, after the cams were adjusted?
Any idea what the back pressure (drive pressure) might be? I didn't see any sensors so I'm not expecting any actual measurements, just a WAG.
Wednesday, February 13, 2013 6:39 PM
It is hard to tell exactly from looking at the map. I tune by trying and seeing.
Saturday, February 16, 2013 11:16 AM
I would be interested what is the gain at the same boost level and and AFR...that is a real compaction.
The only real compaction was the cam changeover.
Sunday, February 17, 2013 10:41 PM
No it's not, the whole time we are tuning for the maximum power repeatable several times on 91 octane with no knock count.
The Garrett turbo allows for more aggressive tuning due to less crossover pressure and less reversion. Making power with a turbo isn't identical boost levels and AFR.
Tuesday, February 19, 2013 10:19 AM
The better the flow the easier to make more power, yet its interesting what difference it makes when just the flow is bigger and the pressure is the same.
Most of the time not the bigger power what we aim for yet the better and easier breading of the engine, can rev harder and fester so the car in the end is faster.
Tuesday, February 19, 2013 11:12 AM
We are talking about crossover pressure not boost pressure.
Wednesday, February 20, 2013 2:00 AM
Yes, we are mapping for emap to (max 1.5 x boost), basically the same deal.
On the same boost pressure you can make 300hp and 500hp on an evo engine (different turbo, cams, exhaust and the rest), or any other one just because it has better flow, less emap.
This is what I was interested, how much it gained on the same boost level, just because it has better flow (less emap or like you say crossover pressure). Anyway I compare every modification this way, so I know how better it flows or run cooler after the mod.
Friday, February 22, 2013 10:35 AM
A few things that I would like to note:
First, here in CA we are limited by the fuel that we have. Give me 93 octane gas and I will run more boost and gain at least 25 hp from the higher octane alone. As it is on 91 octane we were having 1-2 counts of knock which is considered the edge that you should not cross over. If I added more boost, then I would have to pull timing or richen the AFR. We already did make the AFR slightly richer. Pulling 1-2 degrees of timing simply to add a 1lb of boost wouldn't leave us with much more power. The redline on the car is 7000 rpm, the cutoff is 7700 rpm. At redline we were already at 23.x psi and by cutoff we were around 21.xx psi.
This turbo needs better octane fuel. Those who are running it in Australia on higher octane are getting better results. Unfortunately, we do not have that luxury. Hopefully we can try and tune the car on e85 and see what it can do. I would love to run 25 lbs of boost by cut off, a shitload of timing, and a lean/mean 12.2:1 AFR to see what this turbo can do.
Second, this car is a daily driver. We were shooting for a total package that is smooth and powerful. You have to drive this car to appreciate its stockish nature. It is simply amazing. It just revs all the way to cutoff with no hiccup in the power band. It pulls in high gear from low rpm and does not stumble thanks to the mivec tuning that Mike and I did. And it still gets 26 mpg. It is a pure pleasure to drive daily.
Monday, March 25, 2013 6:54 AM
Honestly these numbers are kind of disappointing. I'm no expert, but I tuned my own Evo X so I am pretty familiar with the car.
With the money spent on the turbo, and the lag that it introduces, the horsepower gains up top should be a lot more than 17hp for it to make sense. Also, because the turbo you used is based on the stock frame, you still have the boost taper that occurs with the stock turbo because the exhaust housing can't flow enough to keep the boost up at higher RPM's. So unfortunately, you'll probably never be able to sustain 25lbs of boost at redline with this setup. If I were planning on replacing the turbo and the exhaust manifold, I wouldn't have used a stock frame turbo for that very reason.
The costs involved to make this power (under the curve, not just peak) don't make sense. I'd be interested in another update with a different turbo. Something's not right.