This test is a little OT becuase it was done on an evo, but we will be performing the identical test on an STI in the next 2-3 weeks.
The idea behind this test was to prove if a larger (and much less expensive) non-BB turbo can outperform a smaller BB unit. After having extraordinary results with the new BorgWarner/Airwerks S300SX turbochargers, we found as good or better performance with the non BB BW turbos than from duall ball bearing Garrett GT turbos. Sean Ivey at Iveytune in NJ was very interested to find out so he graciously lent us his dyno facility and tuning prowess.
Long story short, we have a customer who purchased our T4 Twinscroll GT35R kit for his almost entirely stock daily driver evo9.
pic of his car
The only other mods on the car are cams/springs/retainers clutch and fuel mgt. The owner daily drives it on 93octane pumpgas straight, and occasionally puts c16 race gas in it to turn the boost up and make 540+whp. While he loved the powerband and performance of the car, the twinscroll garrett 3582R was not satisfying him 100%. his complaints were:
-the twinscroll gt35R spooled very fast, but at WOT in 5th gear and low rpms it would occasionally surge (and that is WITH the tial ported shroud housing)
-the powerband was great, but he wanted even quicker response and more midrange with more high boost potential for when he builds the motor next year. His cosworth M3 cams are lacking a little midrange, so anything we could do to get the lowend-midrange back was welcomed
-the extremely narrow gt35R center section sunken into a huge twinscroll housing makes getting to the water and oil lines very difficult as well as tightening the CHRA backing plate bolts can be tough.
-the low production qty twinscroll turbine housing casting is not as good quality as the other garrett gt castings and is a pain to get wrenches on
enter the BorgWarner/Airwerks S300SX 83-75 (basically their version of a twinscroll t4 35R with slightly larger turbine wheel and higher flowing compressor) This turbo has a 7 blade 60mm inducer that moves 68-69 lb/min (good to 45+ psi boost). The exducer is 83mm with the extended tip compressor design. (look at the comp maps to compare and you can see how well the extended tip works to deliver solid airflow at high boost levels) It uses a large 75mm 10 blade inconel turbine wheel (the 35R is only 68mm) in a 1.00 a/r housing modified to 3" vband. This turbo is NON-bb and non-watercooled only!
the first dyno compares this stock 2.0L mitsubishi evo 9 on 93octane pump gas at 26psi boost. The S300SX 83-75 with 1.00 a/r TH (red line) vs the garrett gt3582R 1.06 a/r twinscroll t4 (blue line).
Its pretty amazing to see that despite having a 7mm larger turbine wheel and non-ball bearing CHRA, the bw spoolup almost mimics the garrett 35R. The bw even has a slight bump in midrange tq and hp.
the 2nd dynograph shows the above car with race gas at 30psi (red line) vs the above plot of pump gas at 26psi (blue line)
The owner noted that on the street the BW feels noticeably better in and out of boost particulary the transients like between gearshifts. also now there is zero surge. it has better fitment, and is easier to work on than the ball bearing garrett center section. and best of all it sounds muuch cooler
These are non ball bearing turbos, but the development that goes into them is incredible. They have journal bearings designed to hold their speed at particular levels and offer ball bearing control and friction levels but with journal bearing durability and ease of repair. very very impressive.
As soon as this was done, we unstrapped the 35R 2.0L evo9 and sean got started tuning a customer's car. this one has a Buschur engine and our Twinscroll 4088R turbokit. Overlaying this graph on top of our 2.0L twinscroll BW testing really expemplified how well the increased displacement from 2.0 to 2.3 affects the powerband. this 3rd dyno shows the 2.3L Garrett GT40R turbo at 26psi on 93octane (GREEN LINE) vs the S300SX 83-75 with 1.00 a/r TH (red line) vs the garrett gt3582R 1.06 a/r twinscroll t4 (blue line).
the 4th dynoplot shows this same 2.3L mitsubishi evo8 with the garrett GT4088R at 26 psi on pump gas and 40psi on race gas. it made 694whp and 559tq but the injectors were running out of out fuel (only 1000cc inj) and the motor out of cam (only buschur 272s) so the run was aborted before we could make over 700. that will have to wait until next week. the car is another psycho fast street car with AC. Sean took it out todays cold weather and in 2nd gear it spun all 4 wheels instantly he had to get out of the throttle
Sean will swap in some 1200cc injectors, some big tomei cams + cam gears and then retune, but after that i am most excited to install the borgwarner 91-79 turbo (their version of a 40R).
last but not least here are a few pics of the turbos and the compressor maps overlayed for comparison sake
turbos from L to R -- TSGT40R, GT37R, BW S300SX 83-75, TS35R, BW S200SX 75-70, TS3076R