Subaru Engine Comparison - 6 Cylinder vs. STi
Updated: Apr 29
In this blog post we discuss our findings from our recent dyno day of our 6 Cylinder swapped RS to our stock 2011 STi. Check out the video below to see our day at the dyno!
It's always difficult to compare two engines since there are so many variables in how a engine is set up (turbo vs. non turbo, exhaust design, etc.) and the application that it is in (chassis weight, transmission gearing, etc.). To try to eliminate as many of these variables as possible we compared the actual engine performance stock to stock without worrying about the differences in chassis type. This is why we are heavily leaning on the dyno charts (see below). We know that the race between the stock 2011 STi and the H6 swapped RS would be close and we guess the RS would win because of the weight difference, but really this isn't what this is about. However, we know people want to see what a race between these cars looks like, so we do plan on putting them head to head as well so stay tuned. (2020 Update: we went to Mexico and did a race, the STi blew the H6 out of the water by 2 car lengths over a 1,000 ft race).
6 Cylinder Specs
To clarify which engines we are testing, when we say 6 cylinder we are specifically talking about the EZ36R engine found in the 2008+ Tribeca and 2010+ Legacy. It's a drive by wire engine with intake and exhaust AVCS. From what we've seen the EZ30R from the 2005-2007 Tribeca and 2005-2009 Legacy would be very similar to what we found with the 3.6. The EZ30D from the 2001-2004 Legacy and SVX are slightly different in setup and horsepower so our findings would not be comparable here (note that stock for stock these make much less HP). The only modification we made to our EZ36R engine was a replacement exhaust that eliminates the catalytic converters because this is a known reason these engines struggle to make power.
STi Engine Specs
The engine we are testing to compare is from a 2011 STi with intake and exhaust AVCS. We left the catalytic converters on this exhaust because we know that from the factory the 6 cylinder is about 50 horsepower less at that crank. To try to even the playing field we determined the best route was to leave the STi engine totally unmodified to start.
6 Cylinder compared to STi
There are a few variables we can compare to each other to get a better idea of which engine choice is best for your project. We believe both the 6 cylinder and the STi are good engines, it just depends on your overall goal for the car, budget, and mechanical capabilities.
Before we break down the cost comparison we need to talk about transmissions. Both engines make enough horsepower and torque that although a 5 speed could work, it would be pushing its limits. For cost comparison we are assuming that either choice would need a 6 speed transmission and thus the cost for this would be the same for either path, therefore we do not factor it into the analysis. All costs are ballpark. Of course deals can be found on anything based on availability and resources of the owner.
2.5 Liter STi 2004+ USDM
Engine - $4,500
Wiring and Stock ECU - $1,500
Misc Parts (radiator, exhaust, etc) - $1,000
Total Cost - About $7-8,000
3.6 Liter 6 Cylinder 2008+
Engine - $1,000-$1,500
Wiring and Standalone ECU - $4,500
Misc Parts (radiator, exhaust, etc) - $1,000
Total Cost - About $7-8,000
Overall they are pretty similar due to the relatively inexpensive cost of the 6 cylinder engine. As with any project we see, the cheaper the engine the more expensive the labor and vice versa. It just depends on where the project is going and what skill set the owner has. With that in mind, this pricing does not include using professionals or shops and also does not account for the amount of time spent fabricating and installing parts. Due to the 6 cylinder path being less established, there is more labor involved with the 6 cylinder because there aren't as many off the shelf parts readily available. For a 6 cylinder build there will be things that need to be customized like the radiator and exhaust. We managed to do it with minimal fabrication but it is a still a factor given that a 4 cylinder swap does not require any custom work.
It is worth mentioning the weight concerns. The 6 cylinder has a metal timing cover and timing chain which adds even more weight beyond the 2 extra cylinders. In addition, all of this weight is pushed even further forward because of the additional cylinders. All of this creates a potentially nose heavy car. In our RS, we have not driven it enough to notice this as an actual issue and one could say the additional weight of the turbo related components of the STi would probably even this out. The counter to that argument would be that even if the weight is similar (H6 probably still weighs more), where the weight is on the STi is going to be much more centered than the H6. Still, it's something to consider depending on what the plan is for the car.
As mentioned before, we know that from the Subaru factory the STi makes 300 horsepower at the crank and the 6 cylinder makes 250 at the crank. We also know that there are massive limitations with the 6 cylinder headers so we eliminated that problem since we had to tune the car anyway. We fixed the flow of the headers and tuned it to give the 6 cylinder a better chance.
6 cylinders of all varieties have a mystique about being "torque monsters." The dyno chart would disagree with that statement. Peak numbers between each engine is the same, but it's not just about peak numbers. Comparing through the entire RPM range is going to tell the complete story. At all of the key RPM spots, the 6 cylinder and the STi make essentially the same amount of torque. Take a look at 2500 RPM to 3000 RPM, this is where we would expect the 6 cylinder to make much more power. Instead, they both make about 160 lb/ft to 180 lb/ft of torque. They both peak at around 3500 RPM at 220 lb/ft where the 6 cylinder stays flat but the STi drops down and comes back due to the factory calibration. It should be noted here that the 6 cylinder has been tuned so it's not completely fair to single out the dip of the STi. When following the chart, note that they similarly start to drop off toward 6500 RPM. It should also be mentioned that both cross hp and torque at around 5500 RPM at 220 lb/ft.
We'll find very similar findings to the torque numbers with the horsepower numbers. They both max out at about 225 HP at about 5700 RPM with very close numbers at each RPM increase. Much like the torque numbers the factory calibration of the STi ECU from 3500 to 4000 RPM flattens out but quickly catches back up to the 6 cylinder after.
STi tuned blows away the 6 cylinder
This brings us to part 2 of the test where we put the STi on a matching playing field in regards to tuning. We hadn't intended on this when we brought the STi out, but Travis at Snail Performance offered to worked his magic on our stock STi and needless to say we were very impressed by the results. This is the point where my mind was changed. With updating the map we saw the numbers dramatically increase in both HP and torque. The real kicker here is when the turbo spools at around 3250 RPM, the STi jumps way out ahead. This goes to show that with the extra displacement of the 2.5 liter turbo over the 2.0 liter turbo models and an appropriately sized turbo the difference is vast, and not in favor of the 6 cylinder.
6 Cylinder compared to STi Tuning Notes
Although the 6 cylinder performed close to the STi on the dyno, it should be made clear that the 6 cylinder is maxed out in its current form. We increased timing, adjusted intake and exhaust AVCS to try to get anything more out of it and found that it just wouldn't make any more power. This suggests that the real horsepower gain was really just from the cat removal and the ECU tuning has a much smaller effect on gains. In our years of experience we have noticed this same phenomenon for all 4 cylinder non turbo applications. The tune from Subaru on these models is pushing it pretty much as far as it'll go. At the end of the day, the 6 cylinder is probably very limited in its stock form due to the heads just not flowing enough air. Could this be fixed, probably. Is it worth it? Hard to say.
Meanwhile the STi is just getting started in terms of it's capabilities. We gained 70 lb/ft of torque through 3500-4500 RPM which is where it's really going to be noticed and horsepower in the same range we gained more than 40 lb/ft. This is all while making no mechanical changes to the STi. With the addition of an exhaust we would certainly see even more gains. This shows that what the STi can do from the factory is pretty dumbed down due to smog and fuel issues that Subaru must contend with and that with a little massaging, we can really see what the engine is capable of.
OUR CONCLUSION - STi for Most People
With the power, cost, time, and goal in mind we typically recommend the STi for most of our customers. The availability of parts and the easy path to more horsepower, the STi is a hands down winner. The other factor here is smog. The only option for a 6 cylinder with a manual transmission right now is a standalone ECU so for those with emissions requirements, the 6 cylinder is not going to be a good option.
Where the 6 Cylinder Engine Might be a Good Idea
The horsepower per dollar up to about 500 horsepower is not in favor for the 6 cylinder. For those planning a high horsepower build over 500, this is where the extra displacement may turn the dollar per horsepower in favor of the 6 cylinder. Both engines will need to be built and although we don't have exact costs for a 6 cylinder build, the cost for an engine that can withstand these higher horsepower numbers are probably going to be pretty comparable.
Another avenue where the 6 cylinder could be a great option is specific racing application. The new ARA Rally rules have increased the displacement for NA models to 3.0 Liters which opens up the 6 cylinder models. This basically doubles the current power output for those running a standard NA 2.5 liter. Any other racing series that has a class component with displacement as a factor could lead someone to utilize the 6 cylinder as well. We even think that autoX would be a good way to show the merits of the 6 cylinder due to the lower speeds and lower gears where it's tough to generate enough load to create boost.
60 Day Update on Real World Driving Impressions
After getting a chance to drive the car with the new engine around town, some longer cruises, and spirited driving, it has definitely been a joy and the sound echoing off a canyon wall can't be beat. It's certainly not blowing any doors off though.
A couple things we have found is that the Raptor headers sit too low for regular street driving since they were built with a Tribeca or Outback in mind. As a byproduct we will likely look to something else or put on the stock headers for the time being.
One other unforeseen issue is the light weight flywheel. The RS has a twin plate OS Giken clutch with lightweight flywheel. What we realized is that it is extremely difficult to drive on the street. The reason is that the flywheel is designed for the rotational mass of a 4 cylinder so with 6 it becomes an extremely light flywheel. The combination of getting the car going from a stop as well as how fast the RPMs drop between shifts is something we will address in the future with a stock STi flywheel and probably a more standard clutch. We were able to mitigate the RPM drop between shifts with the Haltech software closing the throttle body more slowly when letting off the throttle.
1 year update
We are currently removing this engine for a dual AVCS 4 cylinder turbo setup to sort out the CANbus related issues we were not able to solve on the 6 cylinder application. Overall impressions were that it made decent power but the mod path for the future was too custom for us to follow through on. With that being said, unfortunately iWire is no longer building harnesses for standalone ECUs to focus on stock ECU applications. Although some people have made it work with the stock 6 cylinder ECU there are a number of issues that exist with this route and that we are currently unable to solve. Due to the fact the swap will not work 100% as it did from the factory, we can't put our name on this and so we no longer build harnesses for 6 cylinder applications.
Check out the Six Swapped Subaru group on Facebook for additional resources on 6 cylinder projects.
Do you disagree with our conclusions, have a question, or something to add? Let us know in the comments! We want to hear from you.