CorkSport CST4 vs. OEM K04

For the last four years, we’ve been asked countless times if the CST4 is a direct replacement for the OEM K04. With the number of cars experiencing their 2nd, 3rd or even 4th owner, this question is being asked more frequently.  While CorkSport’s intent is to freely share information across the Mazda community, we cannot be everywhere all at once, and we will most likely miss the exact moment the debate explodes on your favorite Facebook page. For that reason…

Today, we’re setting the record straight: The CST4 is a drop-in UPGRADE from the ground up, and by no means is it an apples to apples comparison with the asthmatic K04.

How Does It Work?

Before we dive into the dirty details, let’s go over how a turbo works. A turbo is comprised of three major sections: the turbine, the center housing rotating assembly (CHRA), and the compressor. The compressor brings in clean air, and as the name implies, compresses the air before sending it through the CHRA and into the cylinder intake. With extra air in the cylinder, the engine is able to burn fuel at a faster rate. This process creates exhaust, which leaves the cylinder and is sent past the turbine, causing the turbine to spin. The turbine and the compressor are placed on the same shaft, such that when the turbine spins, the compressor will also spin. By burning fuel at a quicker rate, we are able to create more power in the engine, which eventually finds its way down to your tires, propelling you forward. If you’re craving more details on how turbos work, make sure to check out our white paper on the CST4. Now let’s get back to what this means for the CST4 and K04.

CST4

Is The CST4 Reliable?

The CST4 continues to prove itself as one of the best bolt-on options coupled with the needed reliability to withstand the additional demands of chasing 400 WHP  – something the KO4 cannot accomplish. This is due to an upgraded center housing rotating assembly (CHRA) which has a larger center shaft and larger bearings than the OEM turbo. The CorkSport turbo also sports a performance journal bearing with a full 360° thrust collar, which is what allows the turbine shaft and compressor to spin freely. The OEM K04 turbo comes standard with a limited 270° thrust collar.

Does The CST4 Have Increased Airflow Over The K04?

CST4 and K04 compressor comparison

Next comes wheels, which is where the CST4 really shines. Shown in the image above, we have the CST4 on the left and the OEM K04 on the right. The CST4 is 12% larger on the compressor inducer, and 21% larger on the exducer than the OEM K04. Combined with the use of a taller wheel (green line), every revolution of the CST4 not only brings in a greater quantity of air into the compressor, but has a higher airflow capacity, thus moving a greater volume of air. For those of you that need a few more key specific numbers; The OEM K04 uses a 45mm inducer; 56.25mm exducer cast compressor wheel, while the CST4 uses a 50.5mm inducer and 68.1mm exducer forged billet compressor wheel.

On the turbine side, the OEM K04 uses a 50.1mm inducer and a 44.5mm exducer 12-blade cast wheel. The CST4 uses a 56.2mm inducer and 49mm exducer high-flow 9-blade design. Again, the CST4 outshines the K04 with the turbine inducer being 12% larger and the exducer 10% larger. The 9-blade design has two key benefits: more peak exhaust flow as there is less material in the way of flow, and 21% lighter for a faster spool time.

CST4 Side View

The final component is the compressor and turbine housings. The K04 uses restrictive housings that cannot keep up at higher RPMs, and especially at higher boost levels. You can feel this as your stock turbo “runs out of steam” up above ~5200RPM. The CST4 housings may fit exactly in the OEM locations and use the OEM hook up points but that is where the similarities end.

Both the compressor and turbine housings were increased in size, increased in A/R, and optimized for the stock inlet and outlet sizes to provide better top end capabilities. The compressor ended up at a 0.53 A/R while the turbine ended up at a 0.66 A/R. This combo of housing and wheels keeps power all the way to redline, and in initial testing showed a 50WHP gain at the same boost pressure.

How Much Power Does The CST4 Make?

So what does all of this mean in terms of power? We’ve seen the OEM K04 pushed way out of its comfort zone and make in the 350-360WHP range with the right supporting mods. This is far out of the efficiency range of the little K04, and it’s a ticking time bomb when running at this power level. The CST4, on the other hand, is perfectly happy running in the 400WHP range all day, again, with the right supporting mods. We’ve even seen it pushed to its limit in the 450-460 range.

For those of you more interested in boost pressures, the K04 can hit a max of ~24-25psi in the midrange before it’s out of its efficiency range and starts producing just heat.  At redline, the K04 is typically at a max of about 17-18psi. What you feel as your car seems to stop accelerating after ~5200RPM on the stock turbo. The CST4 does a lot better, hitting a max of ~29-30psi in the midrange but carries the high pressure into higher RPMs, with peak boost pressure at redline of ~26-27psi. This keeps you pushed into your seat with a smile on your face!

CST4 Mounting Point

How Quickly Does The CST4 Spool?

We get a lot of questions on how fast this turbo spools, so let’s take a moment to discuss both. The OEM K04 spools very quickly since its housings and wheels are so small. If tuned incorrectly it can spool almost instantly and kill blocks with an extremely low-RPM torque spike. The CST4 also spools quick, making full boost by approximately 3300RPM on most cars. The big difference is that the CST4 carries power out to redline instead of falling off as the K04 does. To be clear, you still have to be careful with the CST4 as it too can kill a stock block with too aggressive of a tune.

Is The CST4 A Drop-In?

So bringing things full circle, the “drop-in” aspect of the CST4 means you can run it with almost no other supporting parts, only a HPFP upgrade, access port, and a tune are required. It also means that it hooks up directly to the OEM inlet and outlet flanges so that there is no excessive modification required to make the turbo fit. We even include new studs, lock nuts, gaskets, a custom upper coolant line, new coolant and oil crush washers, and the correct oil feed banjo bolt so there is no hassle of finding replacement hardware, gaskets, or lines to make your turbo function. We do strongly recommend picking up a CorkSport EBCS to best control boost on your CST4. We have also found that the stock intake size will be maxed out at around 18-19psi on the CST4. To get more power from there, a 3 inch or 3.5-inch intake will be needed.

By no means is the CST4 a K04 though as it’s larger and has much higher horsepower capabilities. For those of you more familiar with Garett turbos, the CST4 is just a smidge bigger than a GTX2867.

If you want even more info on what makes the CST4 tick be sure to check out the white paper on the subject HERE. As a final afterthought, remember that the CST4 is getting an EWG housing option in the coming months, for added features (and noise!) that just don’t come with the K04. Stay tuned for that, and be sure to ask any questions you may have.

The Design – 2.5L SkyActiv-G Exhaust Header

CorkSport 2.5L SkyActiv Header

A few months ago we broke down the complicated design of the exhaust manifold found on the 2014-2018 Mazda 3 & 6 2.5L SkyActiv.  Mazda put extensive R&D into the design and packaging of the OEM header to optimize the exhaust gas pulses and overlap.  

In this blog we are going to explain some of the design features in the CorkSport 4-2-1 header and why those features are important.  

Below is a diagram showing the primary, secondary and collector routing of the OE header.  

Mazda 2.5L SkyActive Header
The OEM header for the 2.5L SkyActiv engine has a 4-2-1 design.

When designing a performance header we have to ask ourselves, “what is the goal with this performance part?” and then fulfill that goal.  With the performance header for the 2.5L SkyActiv our goal was to increase mid-range torque, retain good fitment and user installation, and improve the sound output of the exhaust system.  

CorkSport Aftermarket Exhaust Header
CorkSport 2.5L SkyActiv header design.

Immediately you’ll notice a significant difference in the design of the OEM header and the CorkSport Header.  There are three major differences:

  1. Primary, secondary, and collector diameters have been increased to promote better exhaust gas flow.
  2. Primary and secondary runner lengths have been increased to optimize power/torque lower in the RPM range.
  3. The design is two-piece to drastically improve the installation process.  

The primary runners (these are the runners that mate directly to the engine) have been increased in diameter from 1.55” to 1.75” and the secondary runners (these are the runners that combine only two cylinders before the collector) have been increased in diameter from 1.87” to 2.00”.  Both of these changes improve peak flow per cylinder throughout the RPM range. Lastly, the collector has been increased from 2.00” to 3.00” to be paired with the CorkSport 60.5mm or 80mm Cat-Back Exhaust Systems.

CorkSport Exhaust Header Installed
CorkSport Header Installed.

Here’s where things got a bit tricky.  Increasing the length of the primary and secondary runners forced us to be a bit creative in routing all the piping.  In order to achieve the primary runner length we wanted, we had to route the piping upward first (as you can see below) then back down between the engine and firewall.  The results were better than we expected with a “Medusa” style header peeking out of the engine bay and the lengths we wanted.

It makes us grin every time we pop the hood open, we hope you love it as much as we do.  

CorkSport 2.5L Exhaust Header broken down for install.
The final design of the CorkSport 2.5L header is installed in two pieces.

However, the complicated CorkSport design did create a new problem.  Installation! We always try to create a performance part that can be installed by the average enthusiast in their garage and this was no exception.  In a one-piece design, the header was nearly impossible to install. We went to the drawing board and realized that separating the upper and lower halves of the header was the best option.

We considered a conventional flange, gasket and hardware setup, but realized it to was far too complex in the close quarters behind the engine.  We then moved to a v-band connection that proved to be the best setup for installation, weight, and sealing ability.

That wraps up the design, next we’ll breakdown the testing and results! Let us know if you have any questions or thoughts down below.

-Barett @ CorkSport

CorkSport CST6

Testing & Validation of the CorkSport CST6

As we get closer and closer to announcing the launch of the new CorkSport Turbo Line-Up we want to share the testing and validation we put our turbos through.  You may not realize it, but we’ve already shared a lot about the CST6 without really saying so, check out Barett’s Built Gen1 Here.  

So we’ve talked a bit about the design intent behind the CST6; defining the wheel sizes, wheel size ratio, and the ball bearing CHRA.   If you’ve seen the teaser listing then you’ve already seen the 633 whp dyno graph, so we’ll look at the data to support it!

The First Look at the CST6 Performance

CorkSport CST6 dyno at 28psi
CST6 running at 28PSI

First let’s look at the CST6 at a more moderate boost pressure.  Above are the results of back-to-back testing comparing the XS-Power V3 Exhaust Manifold and the upcoming CorkSport Cast Exhaust Manifold.  All dyno runs were performed with the same 28 psi peak pressure tune.

So the exhaust manifold testing is exciting, but it’s not what we’re here to discuss.   What I want you to know is that the CST6 is fully capable of providing mid-500 whp power at 28 psi.   While we have and will continue to push the CST6 to its max ability, the 27-30 psi range has proven to be a sweet and efficient spot for the CST6.

Testing the Limits on the CST6

CorkSport CST6 Dyno Graph running 34psi
CST6 running at 34PSI

Searching for the limits with the current fuel system we can easily push past the 600 whp mark plus some.   The efficiency of the CST6 at this power level is still very strong and the turbo continues to pull through the RPM range.   What really makes the CST6 shine is the power under the curve. This is a BIG turbo and will respond like one, but the loss of early spool is easily compensated for with the abundant power curve and power that carries past 7500 rpm.  

It’s important to note that testing for the CST6 is not finished because we are currently limited by the fuel system on the vehicle.   The current fuel system is OE DI injectors paired with a boost based methanol system flowing 40 gph peak. In the near future, we will continue finding the limits of the CST6 with a true port injection system and Split-Second controller flowing E85.   This will give us headroom for 8000+ rpm and boost levels past 34 psi (let’s see what 40 psi give us!).

Looking at the CST6 Data Log

CorkSport CST6 Data Log
MAF Voltage and Actual AFR of the CST6

This is a datalog form the 633 whp dyno run and was recorded on the chassis dyno.   Because of that, it is not a perfect example of street driving… let me explain why. The dyno dynamics chassis CorkSport uses can control load and thus the rate at which the engine can rev through the RPM range.   In order for us to dyno a vehicle at this power level safely, we need to find the right ramp rate for low RPM and high RPM. The biggest factor this affects is the spool RPM of the turbo.

On the graph I marked ~200 rpm shifted to the left for the boost curve.   On the street, the CST6 spools about 200 rpm sooner due to the higher load on the street vs the dyno.   This puts the CST6 @ 20 psi around 3800-3900 rpm.

Also shown on the graph are MAF voltage and actual AFR.   Both of these are important because they provide real data about how the vehicle is being tuned.

Target AFR is set for 11.76 which is neither rich nor aggressive for this setup.  The slight up and down of the AFR curve from 3500-4000 rpm is due to the very high amount of auxiliary methanol starting to spray along with the DI injectors.

Looking at MAF voltage you can see us get well past 4.50v.  Actually, we are consistently seeing MAF Voltage around 4.65-4.70v using the CorkSport 3.5” Intake which has a true ID of 3.50”.  This is just further validation that the CST6 is flowing enough air to support 600+ whp.

There’s more to come from the new CorkSport turbo lineup so stay tuned for more info on the CST5, CST6, and EWG housings.

-Barett @ CorkSport

Let’s Get Chilly: CorkSport Intercooler for SkyActiv 2.5T

It’s time to break down our design for the CorkSport Performance Intercooler Upgrade for the Mazda 6 2.5T. We have covered both the OEM intercooler and piping, and our design plan for the upcoming Sky-T intercooler piping upgrade in previous blogs, but today’s focus is the intercooler itself. Intercoolers are a delicate balancing act between size, cooling efficiency, and pressure drop so naturally things can get a little complicated. Buckle up and stay with us, and be sure to drop any questions you may have down below.

Taking a look at the stock intercooler mounted on the Mazda 6 (shown above) shows us quickly where our size constraints lie. With the large crash bar, we cannot go too much larger in height without trimming the crash bar, bumper, or both. However, there is a ton of room for added thickness and better end-tank design that can really help increase the width of the intercooler. The stock intercooler core is 24.5” wide, 5.5” tall, and 2.625” thick. Our plan is to fit a 27” wide, 6” tall, and 3.5” core without any trimming. This sizing combined with a low-pressure drop will be good for 400WHP with no issues! Because the Mazda 6 comes with just around 200WHP from the factory, this sized core provides plenty of room for upgrading down the road without causing excessive boost lag that can occur if an intercooler is simply too big. Check out a prototype CorkSport intercooler mounted on the car below.

Now that size is taken care of, let’s move on to cooling efficiency and pressure drop of the CorkSport intercooler for the SkyActiv 2.5T. These are tied closely together as getting extremely high cooling efficiency usually means high pressure drop and vice versa. Just so we’re on the same page, cooling efficiency is how well the intercooler cools off the pressurized air that passes through it. So a highly efficient intercooler will be able to bring the boost temperatures down similar to the ambient air temperature. Pressure drop is exactly what it sounds like, a loss in pressure from the inlet to the outlet of the intercooler which can be caused by a number of things: poor end-tank design, too many intercooler fins, or simply poor flow distribution in the intercooler. Too large of a pressure drop means lower boost pressures reaching your engine and/or your turbocharger working harder to achieve the same boost levels.

Pressure drop and cooling efficiency are influenced primarily by two things: fin density and end-tank design. Fin density is basically how many fins the boosted air must pass over when traversing the intercooler. More fins = better cooling efficiency, but also more pressure drop. To choose the best core for the SkyActiv 2.5T we plan to use multiple different fin densities and test each for power, cooling efficiency, and pressure drop. While we can get pretty close based on our work from the CS Mazdaspeed Intercoolers, it’s always best to test and identify the best one for each platform. With this extensive testing, we can reach our goal of improved cooling efficiency, lower pressure drop, more power, and no CELs.

End-tank design is critical as it determines how the air reaches the core of the intercooler. Sharp bends, poor air distribution, and small inlet/outlet size all adversely affect the performance of the intercooler. To fit the core size we want, we had to do away with the plastic inlet and outlet pipes of the stock intercooler. This was advantageous as it gave us more room to have a smooth flowing end-tank that distributes air well to all the runners and does away with the sharp corners present in the OEM end-tanks. In addition, we were able to increase the inlet and outlet size of the intercooler to 2.5”. This is a fairly standard size that has shown to work well for the Mazdaspeeds with stock power and without choking flow way up to Barett’s 600+ WHP.

Those of you with a keen eye have realized that the connection between the CorkSport front mount intercooler (FMIC) and the OEM Intercooler is not the same. As shown in the CAD rendering above, each intercooler kit will come with the silicone and custom adapters that are needed to work with the OEM piping. If you decide to upgrade to the CS intercooler piping kit, later on, the CorkSport Intercooler for SkyActiv 2.5T will not need to be removed, and you will only need to change some silicone parts.

We will have more info on this kit coming soon, with the next blog covering our testing of the different core designs using a few new toys from AEM Electronics. Be sure to check out the product listing for more info, and to be notified when the intercooler is available. Last but not least, CX-9 Turbo and CX-5 Turbo owners, we are 99% sure this kit will also work on your rides but we plan on validating fitment before release!

-Daniel @ CorkSport

CST5 Spools!! Testing and Validation

We’re back on the new CorkSport turbocharger lineup again with today’s blog, this time focusing on the testing & validation of the “medium big” turbo, the CST5. Just in case you missed it, the CST4 (formerly known as the CorkSport 18G) is getting some company to go along with its new swanky name. Check out the full lineup here and the design behind the CST5 here. Now that you’ve read all that, let’s get into what you’re really here for, testing & dyno numbers.

We started with the internal wastegate option, to validate the CST5 for drop-in fitment. Since we’ve had great experience with the drop-in CST4, we knew how to design a turbo around the tight confines of the Mazdaspeed engine bay. The CST5 fit great in the OEM location with just a few minor revisions for proper fitment. It looks pretty good in there too if we do say so ourselves!

Next the car got put on the dyno for tuning and to push the new CST5 to its limits. With a little help from our friend Will at PD Tuning, the CST5 was soon putting down some impressive numbers. We started off with a “calm” boost level of ~25psi. This netted us 450WHP and spool time that surprised us, achieving 20psi by 3500-3600RPM. Turning up the boost and pushing the turbo to its limits, we achieved 519WHP at ~30-31psi on Barett’s built GEN1 MS3. Check out the dyno graph below.

Taking the car out on the street surprised us further at just how early the car was building boost for this size of turbo. Road logs showed that we were making 20psi slightly sooner than on the dyno (3400-3500RPM) but even more surprisingly the CST5 was making 30psi by 3700-3800RPM! Obviously this is an aggressive tune that would most likely kill a stock block, but, the CST5 can be tuned to be stock block friendly and still make good power.

Then came the testing on the EWG variant of the CST5. We had developed fitment for the CST6 which meant the CST5 had no issues upon install on both MS3 and MS6. Next was a quick retune and some power runs. The larger swallowing capacity of the EWG housing meant some extra power at peak, yet spool was nearly unchanged. We made 525WHP at the same ~30-31psi.

Comparing the IWG and EWG turbine housings you can see a small variation in the graphs.  This variation is mainly due to the change from internally waste-gated and externally waste-gated.  The EWG setup provides more precise boost control through the RPM range. The EWG setup allows us to better tune the “torque spike” around 4200rpm vs the IWG setup.  For peak power the IWG and EWG housings are within the margin of error which makes since because they are both 0.82 A/R housings.

Further supporting the IWG and EWG setups, both options allow you to tune the spring pressure so you can better setup your CST5 and Speed for the fuel and boost levels you want and of course the most noticeable difference is what you hear. What’s an EWG without a screamer pipe!  

Wrapping up testing showed exactly what we were hoping for with the CST5: a great middle ground between the existing CST4 and the upcoming CST6 that can be used on both high powered stock block and fully built cars. Our testing continues as this blog is written as the CST5 is being beta tested by a close friend of CS with a freshly built Dankai 2.

There’s more to come from the new CorkSport turbo lineup so stay tuned for more info on the CST5, CST6, and EWG housings.

-Daniel @ CorkSport