I'm not sure why everyone is choosing the 0.92A/R for a 2.0L-2.3L engine. The 0.92 A/R flows like a GT30 in a 1.01 A/R and no one would ever put a GT30 1.01A/R on 2.0L; or very few I would say. Drop down to the 0.72 A/R which flows like a GT30 in a 0.63 A/R and you'll pick up a ton of spool. And if you're on E70-85/race gas, you'll give up almost nothing in peak power.

 

Same. I was not given the chance to give my input before it was ordered.

 

The .63 GT30 /.72 G25 seems on the small side for a ~60lb/min turbo really, doesn't it? Most guys looking for >500whp run .82a/r GT30R hotsides, I thought - which fit basically between the .72 and .92 G25

 

I agree with Mr Lith. .63 will choke before you utilize what the compressor is capable of. Especially on pump gas.

 

Not necessarily. Don't forget, the G25 has like 5-10pts better turbine efficiency than the GTX30-35. Better turbine efficiency = reduced back pressure. So, in the case of the G25 with the 0.72A/R, the increased turbine efficiency offsets the smaller A/R of 0.72 vs. 0.92. So you gain the spool with little back pressure penalty and therefore little power penalty. Also, smaller diameter wheels tend to like smaller A/Rs and vice versa. While the GT30/60mm diameter wheel has relatively poor turbine efficiency compared to the 0.82 A/R, the smaller 54mm diameter turbine wheel may have it's optimum turbine efficiency with the 0.72.

 

I suppose some of that is plausible, fwiw though - someone is running a pair of G25 660s on their R34 GT-R time attack car on an RB26 stroked to 2.8litres and they apparently ran into backpressure issues when they pushed it quite hard. Granted, they made over 1100hp @ crank on an engine dyno with it still, but if a pair of them on a 2.8litre can get to that point then I'd expect that if someone wants to justify the compressor flow on a 660 with a 2-2.3litre engine using a SINGLE one may be better off with a .92 than a .72 hot side.

 

Yeah, I chatted with PMC Race Engines. Notice they did a 40psi run but they were well off the compressor map at that point. What happens when you overspin the turbo off the maps? You drastically increase the backpressure. Notice they said they backed down the boost to 34psi, ran a lot more timing and gained a ton more torque and powerband. Looking at the compressor map for the G25-660, 34psi is ~3.5PR (with some assumptions for compressor inlet pressure depression and pressure drop through the IC plumbing), which looks to hit the top speed line at ~57lbs/min. This is basically on a 1.4L engine (twins on a 2.8L). So, on a bigger 2.3L, lower pressure ratio will be required. Probably in the 2.75-3.0 range where the compressor can do 59-60lbs/min.

And like I said, the missing part of the equation are the turbine maps. My hunch is the 0.72A/R has better efficiency than the 0.92A/R based on GT30-35 turbine efficiency trends I've seen. So the higher turbine efficiency will offset a lot of the effective reduction in turbine efficiency from having to wastegate more (increasing percentage of wastegate flow reduces effective turbine efficiency). I would only use the 0.92A/R in lower boost applications, such as a pair on a big V8 (5.0L+). Back to my point, on that 550whp Evo on race gas, I bet dropping down to the 0.72A/R shifts the torque curve at 500rpm to the left and gives up maybe 20whp. If on E85, maybe nothing at all.

 

You guys are arguing Turbine housing size, I look at that graph and am simply not impressed. I think a Precision 5858 or gen2 5558 would destroy the g25-660 on the same motor. Or even if it spooled a touch later, they'll both make over 600whp.

 

I noticed none of that as I never got any of that information - didn't know they went off the compressor map, that could explain things. I was just told that the hotside was holding them back, so at face value and giving the guys who are tuning the credit that they know what they're dealing with I took from it that they were confident that the compressor was keeping up.

Regardless, that result sounds very successful - they had tested a Precision 6870 on the same engine and the G25s came on sooner, made more power and the 660s held well over 1000hp past 9000rpm while the 6870 apparently nosedived just after 8000rpm.

 

Keep in mind the purpose of the G25 was same power in a smaller package with reduced rotational inertia. The G25 is physically smaller than the GTX30s and equivalent Precision turbos and has much lower rotational inertia for better transient response; which is also the whole selling point of the BW EFR.

 

Yup, figure each turbo is doing 57lbs/min, that's 114lbs/min total where the EFR9180 tops out at 95lbs/min. 3 cylinders per turbo is an ideal configuration; 4 cylinders per turbo and you get cross contamination hence the reason for twin-scroll. 2 cylinders per turbo and you get big, widely spaced impulses which isn't great for the turbine performance. 3 cylinders is a sweet spot where the degrees of crank angle between power strokes of each cylinder is enough to avoid cross contamination. That just got me thinking, the new Ford Fiesta ST has a 1.5L, 3-cylinder turbo engine to replace the old 1.6L 4-cylinder turbo. The old 4-cyl had a single scroll turbo to save money vs a twin-scroll.

 

I've never felt like a 5858 for 30r had poor transient response on a X. And I've only experience them on a 2.0L, not a 2.3 which would only improve it. I just don't see the point when power falls so short for comparable boost/torque threshold.

 

I'm unsure of your point here. Are you saying that better transient response is a bad thing? These are very small turbochargers, they're clearly not made to compete directly with the current 30r and larger GTX or larger Precision lines. We all know that dyno numbers only tell so much of the story. If you've ever driven a car with an EFR on it, you'll quickly realize that the response doesn't show on the dyno. Are the power figures Garrett has given a bit inflated? It would seem so, but these are a starting point for a new line of highly efficient turbos. There's no free lunch in physics, and we have yet to see what the smaller turbine housings and twin-scroll versions will do.

 

For the size of the turbo, the torque plateau starts late, especially for a 2.3L

I have driven EFR turbo cars. Yeah, The transient response is "better", but the way some people describing as some profound life changing insane difference is not based in reality. And to that point, no one has ever posted any form of a data log showing this "amazing" transient response. All they do is talk about it.

My point is you can build something with a similar torque curve down low and it will make a lot more up top...

 

Again, to which I counter, putting on the 0.72 A/R would shift the torque curve good 500rpm to the left and not lost much of anything up top with the G25-660 on a 2.0-2.5L. I agree people have not done a good job showing the transient response benefits of the EFR. MotoIQ did a test recently with headers for a 991.2 Carrera and they did show transient response improvement in addition to the dyno chart.
https://motoiq.com/testing-vektor-pe...911-carrera/9/

Another reason for the G25? Engine bays are getting tighter and tighter. Therefore, people want power in a smaller package. The G25-660 makes the same power as the GTX3071 Gen2 in a smaller package. Where does this come in handy? Check out Chris Forsberg's drift Z for one. You can't low-mount GTX30 size turbos on the VQ like he has the G25s low mounted. All of the German V6 and V8 engines have the turbos inside the V. Yeah, there's one guy that crammed a pair in the GTX3576Rs into the V of the M5 4.4L V8, but they BARELY fit. Think a pair would fit on the AMG 4.0L V8 or on the Audi V engines? C5 Corvette turbo kits used GT28s because there's not much space in the engine bay. The lack of space is why people did the rear remote turbo setups.