SteamSpeed STX 71R BB Prototype Tested: 470 WHP

Hey guys,

Sam here from SteamSpeed.  We super happy to get back to you with the results of our SteamSpeed STX 71R ball bearing turbo for FA20 applications, ie. the 2015+ WRX.  The basic specs are:

  • Stock frame meaning it just bolts up in place of the OEM turbo
  • Utilizes a Garrett GT ball bearing CHRA (center housing rotating assymbly) sourced from Garrett Japan
  • It is GTX2971R spec meaning the we make our own compressor wheel of that size, and reuse the Garrett GTX29 turbine wheel.

[STX 71R for FA20 Prototype Pre-Test Recap]

Several months ago, I wrote a longish article discussing challenges with the JB CHRA and how the new BB CHRA.  It is a great read, and I highly recommend that you read it, but it lays out why we think the BB version of the 71 will be a superior turbo to our JB version.  The next section outlines, the basic points.

[STX 71R for FA20 Performance Performance Hypothesis [Pre-Test]]

  • The OEM style twin scroll turbine housing is restrictive for turbos larger than the OEM unit and our STX 67 JB turbo.
  • This means for our big turbos, exhaust back pressure ratios can get well beyond 2:1 that is efficient for a JB CHRA.
  • If you are going 5:1 or 7:1 makes it hard to build power and it wears out the journal bearings and especially the thrust bearing.
  • Tuners that have a good strategy to manage this backpressure can make good power with the larger STX 67+ and 71 turbos, but if they didn’t, sometimes our customers would be disappointed.
  • The main point of that post was that a BB CHRA actually solves all of these problems:
    • The CHRA can stay efficient even if the pressure ratios are 5+:1
    • This means, it will be a lot easier for all tuners to build power with the BB version of our turbo.
    • The turbo will just make more power under the curve in general
    • The the thrust bearing is more durable, so the CHRA will stand up to more punishment.

[STX 71R for FA20 Test Results]

We set out to actually prove if the BB version of the STX 71 would perform as well as we had hypothesized, and solve the issues we had with the JB CHRA on the larger turbos.  The short version is that yes, the SteamSpeed STX 71R BB Turbo for FA20 did exactly what we thought it would.  It was a lot more efficient that the JB version of a similar size; therefore, it made more power everywhere.  I suspect that tuners all over will be having an easier time getting results their customers want.

Here is the dyno result.  470 whp on E50 and 410 on 91 octane pump gas, and not measured on this chart, a ton more response everywhere.  Note: this is at high altitude in Utah.

SteamSpeed STX 71R for FA20 Dyno Plot. Green – E50, Black – 91 pump gas

[Technical Notes From Jessie at FNP]

Jessie: “Hey, First let’s go over what we saw, liked and disliked with the unit.

Dislikes:

It’s big, if the new unit has a clearance for the oil pan and obviously ships with hardware we are good.  Obviously as the prototype it is going to have clearance issues, and fitment issues.  There were literally no other issues noted from Luke on install.”

SteamSpeed: This will actually be a non-issue for retail units.  We have actually already resolved all fitment issues on our production 71R.  The production model has a modified turbine housing with a cutout to clear the OEM oil sump without modifying it.  We also designed and manufactured custom studs that we include with the 71R install kit.  This is how the retail unit will work:

Notice we modified the turbine housing to work with the OEM oil pan without modification. We also include custom studs with the install kit.
Notice we modified the turbine housing to work with the OEM oil pan without modification. We also include custom studs with the install kit.

Jessie:

Likes:

Sounds epic.  Do yourself a favor a crack open the boost nipple when running.  The ball bearing turbo sounds incredibly mean at idle.  With a catless exhaust it should sound great out the tail pipe.  Think diesel turbo, screaming at idle.

Response, Response, Response.  This turbo is incredibly responsive compared to the previous version.  Transitions in and out of boost are much quicker.

More linear boost curve via WGDC input.  What do I mean?  Check out this boost profile compared to WGDC on the old vs new turbo.  The new unit is much MUCH more linear with interrupt cycle.  This tells us the effects of back-pressure are far less of an issue with this upgraded unit.  You can also see the old turbo have more “Creep” under the curve.  The new unit doesn’t not do this.  The compare for RPM isn’t valid, as the previous tests were done in 4th, the current in 3rd.

67+ JB WGDC

71R BB WGDC

Makes more boost in the upper RPM’s.  This also is a direct causation from the upgraded cartridge.  It seems to be able to operate at higher levels of back-pressure with ease.

Red JB vs Yellow BB

Less oscillation of MRP than the outgoing cartridge.  Just one of those anecdotal observations, normally we see much greater fluctuations in boost on the FA20 with our incredibly fast sample rates.  This unit fluctuated much less, the average was 18% realized lower fluctuations.  This is a great indication of how much more air is being delivered.

 

Much more efficient flow from the turbo.  Check out the new vs old charge air temps!

New unit held much better boost.  From 1.9bar avg on the old unit at redline on 100% interrupt, to 2.3bar avg.

Output: Was increased by 12.35% over the older unit.  This was also impressive as the turbo could have easily generated around 8% additional output, but the owner of the test vehicle was very specific to “Take it easy”.  Based on the airflow averages and their changes, I’d say this observation is fairly precise.

Check out these airflow differences:  Old Turbo 229 average max, New over 300!  (it was 309 average when extrapolated up).  That is an increase of 35%.  Same intake, and injector scalings were used on both turbos.  Compared to the stock turbo this is over 56% increase in flow!

Well, hope this helps you guys.  I poured over all the data and these were the things that popped out at me.

Jesse @ FNP Tuned”

SteamSpeed Focus RS Intercooler Upgrade: Final Design Approved!

We are excited to announce our final revision to our Focus RS intercooler! We’ve gone through three design revisions and this is the perfect design.

[Mounting & Fitment Improvements]

The first prototype core design showed huge efficiency gains vs the OEM unit; however, we saw room for improvement.  In the next two prototypes we iterated and improved the end tanks and mounting tabs to further increase the efficiency of the core and ease of install.

After the intercooler had been mounted and driven nearly 1000 miles of road conditions including freeway cruising, data log pulls, dyno time, and spirited drives, we found ways to improve its mounting to the chassis vs the OE.  In our final version, we increased the thickness of the mounting tab material from 3mm to 5mm and adding a perpendicular support to the lower tabs for greater strength and durability.  This insures a rigid mount to radiator chassis.  We also found that utilizing openings in the bumper frame to support the intercooler from the top was much more effective then the OE mounting points; the factory cooler is only suspended from the lower tabs and held vertical by the upper tabs that engage hooks into the casting of the cooler tanks.  Due to the increase in weight of our cooler we knew the lower supports would never be sufficient without the extra bracing.  To utilize the bumper frame, we installed a bolt and nut through the opening. This was fine for our final prototype but we look forward to using a plastic insert and screw to secure the upper mounts. This will ensure ease of installation with minimal effort.

Here you can see just how massive this thing is. There’s also a good view of the upper bumper frame mounts. This holds the weight nicely and gives great support for the cooler.

[Intercooler Performance Testing]

Our upgraded Focus RS intercooler, uses quite a bit more heat removing aluminum vs the OE unit.  It is about 30-40% larger by volume and more than 50% larger by mass and surface area vs the OE unit.   The performance of the intercooler core of our first prototype was stellar, so final prototype’s core remains the same.

Our Focus RS intercooler upgrade brings intake temps well below 100 degrees F.  It doesn’t get terribly hot here in Washington State but we did manage to have a day with an ambient temp of 80 degrees F and a resting 101.9 degrees F. You can see in the graph how the Charge Air Temp (white line) drops rapidly when the throttle body opens. And makes its way back down to 86.7 degrees F when grabbing 3rd gear.  That’s great!  Many tuners are happy with and consider an intercooler “good” if it can manage to get below 100.  The temp does start to climb as expected in the longer gears but as shown in the second graph the charge air temp only climbs 5 degrees to 92 degrees before lifting off the throttle.  In comparison, the OE core is much less efficient.  We’ve recorded increased charge temps of around 30 degrees F.  That means our upgraded Focus RS intercooler around 25 degrees (6x) cooler than the OE intercooler!

If you have any more questions please let us know! sales@steamspeed.com

[Testing Data]


Above is the data log of a 1st through 3rd and into 4th gear acceleration. The white line shows Charge Air Temp. The Charge Air Temp is around 100 degrees F when the throttle plate is closed.  The induction air is moving slowly and heat from the engine heats the air.  When the throttle plate opens is when the largest drop in Charge Air Temp occurs.  Lots of fresh air is let in and the heated induction air passes through quickly.  As the vehicle hits 3rd gear it is apparent the turbocharger is now working very hard. There is a slight increase in Charge Air Temp towards the end of 3rd gear. If this data was done without the cooler we would see.
Data log continued (same log as previous graph)

#ford #focus #rs #turbo #intercooler #prototype #presale #keepitcool #steamspeed

Project Domino: A New Turbo and a New Tune

It has been over a year since we’ve done anything with the old girl.

As a refresher, our 2001 Impreza RS had(power wise):

  • a full version 8 JDM WRX STI swap
    •  The the version 8 EJ207 engine is still stock; it didn’t seem right break down a perfectly health motor at the time
    • Version 8 STI 6-speed
    • We did the 5×114 conversion from a 07 STI
  • Custom 3″ catless turbo back
  • Custom FMIC with a ridiculously large IC core
  • Big MAF intake
  • SteamSpeed silicone inlet
  • ID1000 injectors + a Walbro 255 lph pump
  • SteamSpeed STX 71 JB turbo for JDM STI twin scroll

In that iteration, it made ~360 whp at PRE facility on pump gas.   For comparison, the SteamSpeed STX 71 on a 2015 STI made about 420 whp on the same day.  With that turbo it was rewarding to push the car, and if you kept the revs up, you could keep it in boost, but realistically it was a 4.5k+ RPM turbo on a stock EJ207.

[Domnio V3]

We wanted to do something a little more responsive.  Enter the SteamSpeed STX 67R+ BB turbo for JDM STI.

We would manufacture small batches of this turbo, and we would sell them all out before we could install one in Domino, but we able to finally get one installed.  It has the smaller 67mm compressor which should make it more responsive, but the “+” means that it still has a GTX30 spec turbine which is a little better suited for a 2.5l.  I think a GTX29 size wheel would be perfect for the EJ207.  Next update to the turbo we can make the change.  🙂

The STX 67 wheel did get us close to 4k RPM, but it still feels big.  In fact, doing pulls, it feels similar to the 71; however, everywhere else, it is night and day.  The BB CHRA is super responsive.  Going on and off the petal, the turbo is a lot more responsive.  It could also be related to the fact that we went to speed density from a MAF based tune.

What is the final result?   It made about 10 whp less with a 4 psi more boost.  Not a bad trade off.

[Dyno Plot]

[Next Steps]

The turbo still feels a little big.  Andy at FTW thought going to a top mount could give back about 500 RPM of spool.  It would probably be best for our usage of the car, but for vanity reasons we like having the huge FMIC.  We could just go to a 2.5l displacement, but it seems like it goes against the spirit of the ej207 swap.

Where do you think we should take the build next?  Let us know what you think.