[STINick19]

Another update & teaser..

Just got the turbo back from SoCal Porting. They gasket matched, ported and smoothed out the turbine scrolls. Afterwards, they applied Cerakote High Temperature Ceramic Coating to the turbine housing. They claim that their porting can improve spool by ~200 RPM and result in 10-20 whp up top depending on the turbo and application. Obviously, the ceramic coating will help keep the heat inside the turbine housing maximizing the exhaust gas energy. This should help further decrease spool and lower under-hood temperatures resulting in lower IATs and a denser intake charge for maximum power. This was a no-brainer for me as this build has been about injecting some modern engineering and efficiency into the "archaic" EJ257. I wanted to have the ultimate, holy grail of OEM+ turbo setups.

Porting the turbine inlet yielded a ~16% increase in the cross-sectional flow area. These aren't the most accurate measurements, but they were able to port the scrolls from ~40.4mm to 43.6mm (~2980 mm^2). For reference, the single scroll VF48's P18 turbine inlet has a non-ported cross-sectional area of ~1570 mm^2. That's ~40% less compared to ~2560 mm^2 for the non-ported P25 twin scroll housing. Also, the wastegate design is significantly better on the twin scroll P25 housing with its dual flappers (one per scroll).

A quick re-cap on the turbo setup:

• OEM JDM S208 IHI VF58 Twin Scroll Turbocharger
• Ceramic Dual Ball Bearing CHRA
• Forged Billet Aluminum 6+6 Compressor Wheel w/ Extended Tips (Reducing rotational mass while also improving strength and flow)
• OEM Titanium Aluminide (TiAl aka Gamma-Ti) 9-Blade Turbine Wheel & Shaft (Significantly reducing rotational mass while improving flow, spool, transient response)
• Lightly Ported/Polished Compressor Inlet
• Ported Turbine Housing Inlet Scrolls (~16% larger cross-sectional flow area improving spool and top-end performance)
• Cerakote Ceramic High Temperature Coating applied to the Turbine Housing (Retaining the heat, maximizing exhaust gas energy, improving spool and reducing under-hood temps & IATs)
• Forge Motorsport Billet Aluminum IWG Actuator (Improved durability and wastegate control/response along with higher boost levels)
• OEM JDM S208 Equal Length, Twin Scroll Header & Uppipe

At this point, the VF58 has essentially been modernized with the same features as EFR turbos and even the IHI turbos supplied to Ferrari! This is the turbo setup that the S209 should have come with

Between the shadows and black coating, it was difficult to get pictures of the ported scrolls.. but I tried. I apologize for the poor image quality. It doesn't do SoCal Porting's work justice. I wish I could have gotten a picture of the porting before they ceramic coated it. The ports are very smooth!

 

[Archer1.8]

Absolute pornography right there. Tough to not mount that on my desk instead of burying it in the engine bay lol.

How does that P25 housing cross section compare with a p20 housing

 

[dlheman]

Great post!

I always wanted to give that Forge actuator a go on my VF56. Ah hell I may aswell just send it to Italy.

 

[STINick19]

Great post!

I always wanted to give that Forge actuator a go on my VF56. Ah hell I may aswell just send it to Italy.

Thanks! Let me know if you decide to and I can help save you a bit of money

I look forward to comparing notes and results once I get this installed!

 

[STINick19]

Absolute pornography right there. Tough to not mount that on my desk instead of burying it in the engine bay lol.

How does that P25 housing cross section compare with a p20 housing

I feel the same way!

I'll see if I can dig up the turbine inlet diameter for the P20. But here's a visual comparison of the single scroll P20 (left) vs twin scroll P25 housing:

I believe the P18 is ~0.63 A/R whereas the the P20 is ~0.71 A/R.

 

[STINick19]

How does that P25 housing cross section compare with a p20 housing

The OEM single scroll turbine inlet gasket is ~2.0" (50.8mm) meaning that it limits the single scroll inlet diameter (even when ported). This translates to ~2025 mm^2 (vs the P25's ~2550 mm^2).

However, the P25's scrolls will narrow more quickly. They have a different A/R (Area/Radius) and I believe the P25 falls somewhere between the P18 and P20 A/R in this regard. Below are some illustrations in regards to the meaning of A/R vs the cross-sectional area. In theory, the P20 should have superior top-end performance over all of the other Subaru IHI turbine housings. However, finding any real data on P20 or P25 turbos on the EJ257 is pretty rare.

Again.. in theory, I could directly swap a P20 turbine housing onto my VF58 if I was for some reason unhappy with the results of the twin scroll P25 housing. But then I would need a Killer B ELH and COBB GESi downpipe for the conversion.

I don't think the WOT dyno charts would look significantly different for these setups, but the twin scroll setup should have superior spool (time) and transient/part throttle response. However, it may sacrifice a few HP in the higher RPM range.. but this is mostly speculative.

 

[Archer1.8]

I highly doubt you will be unhappy with this turbo. Yea, it may not be a dyno queen turbo, but thats not what you are after.

The P20 housing i think will be fun to experiment with on a EJ257.

I think you will be limited by fueling. I decided that fueling is one area of my OEM+ build where Im gunna go aftermarket. Spending good money on S209 injectors is a tough call when they aren't that much bigger

 

[dlheman]

VF48 is not even worth mentioning next to VF56 let alone VF58.

 

[STINick19]

Decided to do some baseline performance testing before I start with modifications while the car is still 100% stock. Unfortunately, I'm limited to ACN91 fuel (aka 91 knock-tane) here in California which hinders performance.

The shift to 3rd right before 60 mph kills the 0-60 time. Bumping the redline up a few hundred RPM with a tune will eliminate that shift making a high 4 second pass much more attainable.

0-10 mph: 0.57
0-20 mph: 1.14
0-30 mph: 1.84
0-40 mph: 2.99
0-50 mph: 3.96
0-60 mph: 5.46

1/8 mile: 8.68 @ 80 mph
1/4 mile: 13.51 @ 103.15 mph
60ft: 1.93

For reference:
2018 STI Type RA (Car & Driver/MotorTrend)
0-60 mph: 5.7 sec
1/4 mile: 14.0 @ 101 mph

2019 STI S209
MotorTrend (91 oct):
0-60 mph: 5.9 sec
1/4 mile: 14.1 @ 101.2 mph

Car & Driver, Motorweek, Autoblog, etc. managed better performance on 93 oct fuel.
0-60 mph: 4.9 sec
1/4 mile: 13.4 @ 105 mph

 

[dlheman]

I thought you have your VF58 fitted when I saw the post notification.

So when will it be fitted and have you decided on who will calibrate the engine?

 

[JFAZ]

Very nice! Can't wait to see how much improvement you get. What app are you using to capture the times?

 

[STINick19]

I thought you have your VF58 fitted when I saw the post notification.

So when will it be fitted and have you decided on who will calibrate the engine?

Sorry to disappoint! I have all of the parts now, just need to order the wideband O2 for tuning. I plan to do the install in two phases:
1) Install the turbo & tune on stock injectors (will max out IDC).
2) Upgrade injectors & install flex fuel kit for E85.

I plan to do the first round of calibration myself via road tuning. I will rent some dyno time once I upgrade to E85 for torque optimization since I won't be knock limited.

I also plan to do a "lag torture" test with the VF48 & VF58 to measure the lag time to peak boost & torque.

 

[dlheman]

Sorry to disappoint! I have all of the parts now, just need to order the wideband O2 for tuning. I plan to do the install in two phases:
1) Install the turbo & tune on stock injectors (will max out IDC).
2) Upgrade injectors & install flex fuel kit for E85.

I plan to do the first round of calibration myself via road tuning. I will rent some dyno time once I upgrade to E85 for torque optimization since I won't be knock limited.

I also plan to do a "lag torture" test with the VF48 & VF58 to measure the lag time to peak boost & torque.

Ahahaha. Is alright man.

Well we look forward for the results!

 

[STINick19]

Very nice! Can't wait to see how much improvement you get. What app are you using to capture the times?

I'm using the dragy performance meter. It's been tested at the drag strip to be accurate within a few hundredths of a second. The key to getting a good time is abusing the drivetrain to get a decent launch. I've also been using it to measure my 5-60 mph time (no launch) which is around 6.1 seconds.

 

[JFAZ]

I'm using the dragy performance meter. It's been tested at the drag strip to be accurate within a few hundredths of a second. The key to getting a good time is abusing the drivetrain to get a decent launch. I've also been using it to measure my 5-60 mph time (no launch) which is around 6.1 seconds.

I'm glad too see your getting a 5-60 mph time better than the 7.0 seconds that R&T got. What RPM are you launching to get your 0-60 time?

 

[STINick19]

I'm glad too see your getting a 5-60 mph time better than the 7.0 seconds that R&T got. What RPM are you launching to get your 0-60 time?

I've been launching between 4500-5000 RPM with consistent times in the 5.5-5.7 second range. I think there's a bit of room for improvement, but definitely not getting a sub-5 second time with the stock redline.

 

[Archer1.8]

Sorry to disappoint! I have all of the parts now, just need to order the wideband O2 for tuning. I plan to do the install in two phases:
1) Install the turbo & tune on stock injectors (will max out IDC).
2) Upgrade injectors & install flex fuel kit for E85.

I plan to do the first round of calibration myself via road tuning. I will rent some dyno time once I upgrade to E85 for torque optimization since I won't be knock limited.

I also plan to do a "lag torture" test with the VF48 & VF58 to measure the lag time to peak boost & torque.

So pumped to see these results. Giorgio has been sending me examples of what he calls their VF58 on steroids haha. Still don't know if a custom VF58 is the same as a true VF58 in the OEM+ game haha.

Anyway, be careful on those stock injectors. My car on Cobb OTS stage 2 91 octane and low waste gate map had injectors at 102%.

 

[STINick19]

So pumped to see these results. Giorgio has been sending me examples of what he calls their VF58 on steroids haha. Still don't know if a custom VF58 is the same as a true VF58 in the OEM+ game haha.

Anyway, be careful on those stock injectors. My car on Cobb OTS stage 2 91 octane and low waste gate map had injectors at 102%.

Excited to see which route you go!

I already have the injectors, fuel pump, fpr, flex fuel kit, etc. ready to go. I just want to make sure everything with the turbo is running well on wastegate spring pressure before the dreaded injector install.

 

[JFAZ]

I've been launching between 4500-5000 RPM with consistent times in the 5.5-5.7 second range. I think there's a bit of room for improvement, but definitely not getting a sub-5 second time with the stock redline.

I'm very interested in your build and I want to thank you for posting everything your doing up on the forum. In the age of Instagram and FB, we need more in-depth content/discussions.

I'm really interested to see how your upgrades change your real world performance. We always gauge performance upgrades with dyno results, but to me, I can't comprehend how that equates to real world driving performance.

I'd love to see some testing where auto journalist take a New stock STI and compares it's numbers against something like a Cobb Stage 3 w/OTS tune (as this is something, in theory, that should repeatable vs a customized tune).

 

[STINick19]

I've been working on optimizing the OEM/Stage 1 tune before installing the new turbo setup. I want to have solid before dyno & acceleration/performance numbers for a solid before/after comparison.

The OEM '18 RA/19+ STI tune is surprisingly good. It actually makes more peak power than the COBB OTS stage 1 tune (ACN91/91oct), but COBB makes significantly more low to mid range torque. COBB's higher redline also allows getting to 60 mph in 2nd gear which significantly drops the 0-60mph time. I also tried the MAPerformance stage 1 91 oct tune which performed very poorly and consistently experienced feedback knock at WOT.

OEM = Blue
COBB = Red
MAPerformance = Purple

The COBB stage 1 tune significantly dropped my 0-60mph time by eliminating the shift to 3rd (5.65 -> 4.85 seconds). Of course, if you go by the magazine/media outlet standards (1 foot rollout), that time drops to ~4.65 seconds. I'm also collecting 5-60mph, 3rd gear (~30-80mph & ~40-70mph), 1/4 mile, and "lag torture" (~2000-4000 RPM in 4th, 5th & 6th gears) performance metrics for comparison as the VF58 should really shine in responsiveness and real world, every day driving conditions where I'm not dumping the clutch to launch the car or exceeding 100 mph..

COBB Stage 1:
0-60mph: 4.85 sec
0-60mph (1ft rollout): 4.65 sec - this is the standard for performance testing

I'm still working on my custom stage 1 tune. I've been able to make solid mid-range and top-end gains compared to stock and the COBB OTS map. There's not much room for improvement over the OEM tune past 6,000+ RPM. I'm just trying to carry as much power to the ~7,000 RPM redline as possible. Unfortunately, I'm quite knock-limited on ACN91 "knock-tane".

Orange = My custom tune (in progress)
Blue = OEM Tune