R&D is not cheap, but looking at data only tells you so much so we regularly push our cars as much as possible. Then we can find out the limits before we send any tunes out.
In this instance we were in the process of testing exactly how much boost we could get away with on as low of an e-content as possible, and at 27 psi we found out. 40% and 27 psi. A single pre-ignition event lead to head gasket failure that blew out a freeze plug in the head, which then turned the oil into a milkshake and instantly took out everything but the rods and cams. Worst possible time too as we had 3 customer FA20’s waiting and we knew we weren’t going to get to rebuild it for a good while. We put that time to use though. Knowing that the WRX heads likely flowed poorly based on just eyeballing them, we spent some time working on the heads to see how far we could go without going as far as moving to BRZ heads. While the D-4S toyota heads are incredible, they make the whole thing quite expensive. Instead we tested the stock WRX, BRZ, and mild port WRX, and then mild port WRX without the metal splitters. Here’s the data utilizing stock valves:
The flow numbers show that it is likely very worth the effort. Once we get it back on the dyno we’ll know for sure. From purely a financial standpoint, this is achieved for around $500, versus $1000+ for BRZ heads. Obviously the overall cost to convert to the BRZ setup is much higher than that. Not surprisingly the exhaust side is near identical.
Given the turbo we selected is the GTX3076R, with this, WMI, and an auxiliary injector setup we should now be well equipped to push 24-25 psi to 7500. That ought to put us right at the GTX capacity of about 450 torque and 550 hp without hopefully resorting to race gas. With 40% ethanol we should be able to get similar knock resistance to 98 octane, and with WMI/auxiliary injection push it even further.
Anything bigger than that then likely a BRZ head conversion would make a lot more sense, but who knows. If we can keep making power without having to ramp up boost with revs then we may be able to break 550.
For those that know us in the Miata world, you’ll recognize that we operate very differently than your typical performance shop. We always look at everything empirically, and ensure we are 100% confident in our product before pushing it out to the consumer. This means sometimes it takes longer to release, but the end result is that quality and reliability are always better. This is one of the reasons racers use our tunes and our products, and why you’ll find our tuning used across several platforms, racing classes, and most of the Pirelli World Challenge MX5’s that run today. We also supply hard parts for racers that are specifically designed for race use, with many of those features trickling down to the consumer products (i.e. our intakes, turbo kits, mufflers, flex fuel kits, intake manifold, header).
For the ND we partnered with Eurosport Racing and GuestDriver.com in Joliet, IL next to the Autobahn Country Club race track. For this round we are doing two of the cup cars, one without a header and one with. Dyno used is a dynapack. This particular dyno gave us a baseline of just over 132 WHP on the totally stock cup car.
We started by first gathering a good bit of data from these cars and some road cars. This is a very important first step in understanding exactly how this ECU operates. To make power, we first worked the air fuel ratios to find out exactly where we are making best power. From there we worked on the cam timing to find the most optimal combination of overlap. Once we did that we moved on to the ignition timing. Reason for this sequence is to ensure that knock retard is minimal and we can differentiate what is actually getting us the gains.
Initial impressions? The stock tune is VERY good as is. This is no surprise, and as already discussed previously the car is very much designed to “ride” the knock sensor and always be at peak timing. However, with good gas (93 or better) you can start to get the more substantial gains.
After the first day we got the first car done (stock cup car), and the results are outstanding! From 132 we went up to 145 to the wheels. Below is a version with the stock rev limit.
Header Testing/ Day Two
Day two we had set aside for the header equipped car. With direct injection, scavenging is critical in making power. One of the reasons we are so successful on the BRZ/FR-S platform is our ability to fine tune for each header on the direct injection FA20. This proved valuable experience here and saved us a good bit of time. The Mazdaspeed header is a great design, with maybe just slightly smaller primaries than what we would have liked. But being a truly equal length, it has the advantage of making more peak power at race revs (4500-7500) but that will come at a slight cost for the street user—the “torque dip” that every BRZ/FR-S owner is intimately familiar with. 20+ ft/lb dip in the low range. These can be slightly massaged with cam timing, but they are actually caused by the physics rather than tuning, and can only be worked on so much. Towards the end you can see we got rid of the worst of it at 3500, and really made the most of the the header throughout.
The simplified explanation is that the same pressure waves that are responsible for scavenging and the >100% VE at high RPMs, resonates at low RPMs and interferes with power delivery. With all this in mind we took the tune from the day prior, and then began to iron out some of the areas where the header needs changes. The result? Yet again we were blown away with how well the car did. In comparison the 132 to the wheels factory, 145 tuned, we hit 156 with the header and tune! The ND does not fail to impress here, showing slightly higher gains from both tune and header than what we saw with the NC.
At this point we are going to keep tabs on the cars and see how they perform throughout the season. During testing we continued to make power with every bit of timing we added, however, we were approaching the range where diminishing returns tell us that we are very close to MBT and it is best left just a hair under that to leave some margin for heat etc… So with this we are 100% confident that folks will find great value for both street and track users with tunes, especially with the addition of modifications. Given this data we are officially taking our tune off of “beta” and it is now officially released.
If you have an 08+ WRX/STI (including 2015 and up) then this tune will make it perform as it should have from the factory, but with minimal impact on reliability. For those covering 100k+ miles on their vehicles, raising boost to 20 psi and running AFRs for best power is not the way to go. While they do run great that way, reliability will be reduced due to the increased load on the engine. What we set out to do with this tune is to create a “factory replacement” tune, that leaves boost levels modest, but addresses the lean AFRs, delayed closed loop, and lackluster cold weather performance.
Requires EcuTek ProECU kit and EcuTek license. This tune is available in any of our STI/WRX EcuTek packages.
After having an unfortunate situation with our 2015 WRX at just shy of 5,000 miles on the clock, we figured we’d make the most of the situation. After doing a fully built forged motor, we decided to use the old engine as a jig to make a turbo system. For this one we used a GTX3076R, same as the FA20 BRZ we built last year. With a powerband that hits from 4000-6500, it’ll be perfect for the track. Once we have access to an upgraded HPFP, we should be able to stretch that powerband to 7500+. Until then we’ll use water/meth injection on the street to get the higher numbers, and simply limit boost at high RPMs for the track. On with the show….
After driving the car a few things became clear. The TiAL BOV opens up under idle vacuum, which is normal for this valve. Converting to speed density solved the tuning issues resulting from that. Thanks to EcuTek, we have the RaceROM functionality which allowed a full speed density conversion which runs as smooth as OEM. Amazingly enough the VE table they supplied worked 100% dead on compared to our MAF tune. This allows us to run a softer spring and minimize compressor surge and preserve our turbo.
Now for the most important thing, how driveable is this GRX3076R on a 2.0l? Actually, it is not half bad. It makes much less boost below 3500 compared to the stock turbo, but this turbo at 5 psi is more fun than the old at 12 psi at 3000 RPM. So yes, slower below 3k, about even 3-3600, then completely takes off after 4k. Meat of the powerband is 4500-6500. After that the direct injection runs out and we have to drop off the boost. Hopefully we will see some upgraded HPFP upgrades soon, if not then the Aquamist water/meth system we have waiting for us in the box will go a long way to maximizing the powerband.
We’ve been at it again with CAD. This time we addressed a weakness of the Duratec/MZR oiling system, where the factory cast oil filter bracket can break off where the nut mounts, and leave a big puddle of oil in the middle of a track session. We’ve now made it easy to relocate your oil filter as well as add an oil cooler/turbo feed/or oil temp/pressure senders. This is 100% US made and machined, designed completely in-house.
The two prior turbo kits utilized a sandwich adapter and a banjo feed adapter. The banjo method actually works well, but it takes a good bit of effort to ensure that this connection is leak free and aligned correctly. The alternative method we have seen is the sandwich adapter. This is not something we would do given the fact that we have had customers actually lose the entire filter assembly due to the additional stress the weight of the adapter adds. Our solution is simple yet reliable. We custom machined fittings in-house that utilize the correct thread size for the factory sender, then integrated a -4 AN feed. This is the preferred method but the fitting size is not a readily available part.
Once we finished making the first prototypes we had to ensure that you couldn’t break it if you tried. So with a specified tightening torque of 15-20 lbs, we pushed it as far as it could go. Results? 106 ft/lbs before failure! Needless to say this final hurdle has been overcome!
We are going live with the pre-order!
After a year in development we have our own MX5 Turbo Kit. While similar in many respects to the hybrid kit we already carry, this adds several premium features over the standard kit.
This is the first track proof setup. Thick CNC 1/2″ head flange, 3/8 turbo flange (not plasma cut 1/4″ as we have seen), precise fitment, and fully back-purged. All US made parts where possible, and genuine Garrett turbos with genuine Garrett wheels and intercoolers–not some Asian knockoff hybrids. While Bell uses genuine stuff also, we are taking a different approach with turbo inlet and IC end tanks which give us a greater flexibility for higher power applications. Manifold and downpipe designs are somewhat different as well.
Turbo options will utilize the T25 turbine housing. We’ve debated this over in great detail, however, this spring a gamechanger came around. The GTX29 series from Garrett. This puts a clipped GT30 wheel in a hogged out T25 housing. Now you can have your cake and eat it too! The same fitment with near instant spool on a 2l (GT2560r/GT2860rs) will support 550 HP with a GTX 2976r! Which is perfect for the 2.5. Best of all is it does not have the boost creep issues you see with the 30 series. So if you don’t want 12 psi, simply use the GT2860rs and it will stay under 10 pounds with the stock exhaust or 2.5″ exhaust.
A few details.
So these are the basics of what we will be offering. We’ve evaluated all the other options out there and clearly there is a spot for a premium kit that doesn’t skimp on quality, but we’ve picked carefully which features to splurge on, and what is best just as it is. This is the wish list we always had for an NC turbo, and what our customers have asked for as well. With this iteration of a turbo kit, we are shooting for a perfect balance of responsiveness at 280 HP, with the potential of 550 without changing anything other than the turbo. The current form will do just that.
We get this question almost daily! What pulley to run on 91 octane? What about intercooled? Can the 335 system be used on a road course? The worksheet below summarizes our experience running every version of the kit. The limitations of the system are the smaller laminova intercooler core, and the heat generated through compression. Combined with the 12.5:1 CR of the BRZ/FR-S FA20 motor, the chart below offers the best balance of knock resistance to performance gains. If your car does both double duty with e85 and pump gas, pulley changes can be made in about 5 minutes with common tools. Alternatively a map can be set up using EcuTek RaceROM to pull out more timing for pump gas operation. This allows you to run e85 at the track while still remaining relatively safe on pump gas during regular operation. For built motors, you will require a smaller pulley to achieve the same torque if dropping compression. So a 75 mm pulley on the 335 kit is nothing drastic when running 9.5:1 CR. With any of these kits, we don’t recommend dropping the compression much below 10.5:1. With the direct/port injection combination of the Toyota D4-S system and dual VVT, you can get away with a good deal of static compression and still run fairly high boost levels.
The other day we dyno tuned a turbo kit for a local customer on an MX5. Setup is an off the shelf turbo kit (shall be unnamed), injectors, fuel pump, water meth injection. The system would creep to 17 psi in the cold by 6800 RPM, once warm, 14 psi. This latest time he had brought it over to get water meth put in, and an EBC to just set the boost to 14 psi as it was creeping there anyway. This system unfortunately uses a 3″ downpipe from the turbo back, and an oversized GT30 turbo. The vehicle is a high compression 2.0l (10.8:1) and very limited space to put in a larger wastegate. For this reason a 2.5″ downpipe and a T25 flanged turbo is much more appropriate. Initially we tuned it over the winter working around the boost creep issues. That put us at 275 WHP. However we had to utilize advanced cam timing strategies and throttle management to reduce the boost creep at the top end. The plan was to do water/methanol injection come spring, and this was completed just recently. The dyno chart below shows the gains from doing just that. With the water/methanol injection, we were able to just set a level boost and not have to limit. We’re well over 300 to the wheels with this combination.With this setup now nearly mirroring our car from 2009/2010, we went ahead and overlaid the standard Bell Engineering turbo kit at 14 psi and water/meth injection. Both cars were running similar AFR and timing numbers, and the only difference being the turbo kit design. Now what you see is that the smaller turbo (GT28RS) makes over 50 ft/lbs more torque below 3800 RPM! So for a stock motor where you are highly unlikely to ever exceed 325 WHP, it makes very little sense to go with such a large turbo. A note on the plots: the dyno graphs are on different days under different conditions. So you if you normalize both the HP levels should be similar. However the gains down low are still 100% there.