Most major racing series would disagree with you on the power per mm. Most of them have class limitations and rules based on inducer sizes of the turbo you run because it locks you into a particular "power band" that then produce a close set of times. This is for drag, indy, f1, rally etc.

For example, in the 62mm range, you have the 6266 and all of its siblings with various sized turbine wheels from Precision, you have the comp 62 with various sized turbine wheels, an EFR 8374, and the Garrett GTX3582. Out of this whole range, the Garrett and EFR bring up the bottom of the power band in power per mm. You don't see people cracking 800hp with a GTX3582, partially because of the compressor wheel and because of how bad the GT35 turbine wheel is. You're absolutely correct that inducer size doesn't tell the whole story, but it gives you a pretty good idea of what band of power it's going to be in. There usually aren't really many outliers to power per mm unless you get creative, like cutting your compressor cover to extremely tight tolerances and hoping that it doesn't blow up if you push it too hard, or running extended tips or custom exducer sizes.

As for the turbos that make 900+hp out of a 62mm, call up Joe at locash and ask. His car broke the FIS records if you were so inclined to research it. http://honda-tech.com/drag-racing-36...untry-3268855/ If you took your blah blah blah what did it trap crap attitude and actually did your homework you'd see that there's cars that make absurd power for the size of turbo they're running. It's actually pretty amazing stuff.

Next, I didn't shift it as hard as possible. Lets say I did, what arbritary number would you ascribe to "shifting fast"? The real question is, do I have to have a sequential or automatic transmission with sub 100ms shift times to appreciate the supposed "transient response" from the EFR? I still had some pretty decent boost recovery times with my "slow" 400-500ms shift times.

In the end, I really just want to validate if the "transient response" claim. So far to date no vendor or end user who pushes these turbos have posted any real world data on this. It's all just been marketing so far.

 

The transient response/time-to-torque benefit due to reduced inertia is most felt at lower engine speeds. Say in the 1500-2500rpm range. As you go higher in engine speeds, the time-to-torque benefit due to lower inertia is reduced.

 

That would mean you have more power down low, or the boost threshold is lower because you have torque faster. In my graphs I see around 3psi around 4000 or so rpm so it'll be interesting to see what the other turbos will make at low flow rates.

I'm going to do a quick dyno pull on Monday with the Efr before I pull it off on a pack or jet to have final 30psi dyno numbers and then I'll slap the precision on and retune.


Edit: Actually, how would you want me to test this? Go to 1500-3500rpm in 3rd gear or so and cycle to 100% throttle and see what I make? I can do this test and repeat it on the other turbos if you can help me try and devise some test to produce the behavior your thinking

 

I usually test by decelerating lightly, in gear, at zero throttle, and then go full throttle until it is fully on boost.. I do it for 3000, 4000, 5000 rpm... this gives me the transitional data that relates closely to the way we drive on the road and track... boost recovery after shifts is not so important for me as by using NLTS changes everything.....

 

+1

 

Does your S2000 with the 9174 still have that 3rd party "QSV"?

 

Because people who's opinions I trust (Chad at CBRD/Jon at Full-Race/Jeff at Perrin) have all tested the EFR line and were impressed with the results. So much so that they design kits around them. These are established tuners who have developed great reputations. I don't know you from Paul, so I simply asked if you had any data to support your position. You bash others for not doing side by side apples to apples comparisions, yet do not do them yourself.

 

I forgot, there's some gear dependency too, but I can't remember how strong of a lever that is.

For a test procedure, since S2k is so shortly geared, I'd probably use 3rd gear. You want step throttle inputs to measure the time to torque. So maintain constant speed at set rpm and then go WOT and see how fast the boost comes up. I'd probably do 500rpm increments if you have the time: 2000, 2500, 3000, 3500, 4000, 4500, 5000.

Maybe use cruise control to maintain constant speed and then go WOT. What you should see is two things. Time-to-torque reduces as your start engine speed is higher. The percent advantage of the TiAL turbine wheel reduces versus standard inconel at higher engine speeds.

Other variables that affect time-to-torque are turbine A/R and turbine efficiency among other things, but at least your testing will show the difference between one package against another.

 

Literally, its their job to sell you and other people parts. Of course they "tested the EFR" line and were impressed. The only other incumbent of OEM grade turbochargers was Garrett, and they made the worst power per mm up until they decided to introduce the GTX lineup which is now still just ok. Of course they sell kits around turbos that they get at a massive discount on (far deeper than garrett/honeywell would ever provide) because it's more money in their pocket.


It seems that you just don't fundamentally grasp that the goal of these vendors is to sell you and the rest of the market turbos and sell kits built around said turbos. It especially makes a lot of sense when you consider that the EFR lineup comes with an integrated BOV and WG that actually does work very well. This is extremely attractive to anyone looking to develop products on any platform because you lower the cost of the kit and open up the market to more people since they don't have to buy a BOV or WG cutting down costs for all parties involved.

What I have been saying this entire time is that no vendor has actually provided any proof for their claims of transient response. In terms of them being more reliable, it's been debatable. As for the comparison, I'm literally doing it right now.
I'm going to provide all my results and data from my testing so that it will either validate what BW claims, or prove that it's not the case. I'm not interested in discussing anything further with you if all you can do is appeal to authority.




No, I've completed that testing and am now running it straight on the manifold. What I found is that the QSV had a very limited impact at the lowest of low end torque numbers and choked it in the mid and high RPM range. This compounded with how well the honda heads flow, means I didn't really gain much. With a 4 port solenoid I'm within 100rpm of the spool without a QSV and I had to add fuel everywhere. I know the 2js don't breath as well as the F or K engines, but I really have to wonder if the QSV isn't just another case of a company trying to sell you something. I think if it was designed inside the housing itself like a Koenigsegg and was smoothly flowing it'd work exactly like we'd expect. If BW decides to release the QSV housing for the EFRs or Airwerks I'll definitely retest them.



I have a 3.63 final drive so it helps quite a bit and actually helped spool across all the gears because of the load. I don't have cruise control, but I can definitely do these tests, It'll be interesting to see how it plays out

 

So we should take your word for it, over theirs because they are in the business of selling turbo chargers? Doesn't that make them more qualified to speak on the subject?

Do you really think they are so willing to trash their reputations to make a quick buck selling junk?

 

I hate the financial gain/ulterior motive analogy. While there are grains of truth to it, nothing would ever get done if the only motivation was curiosity. These large companies that develop these turbos wouldnt even exist. We probably wouldnt even be driving sports cars/coupes/sedans, etc etc.

I would also not call the 35R turbine awful. It is a bit dated but there was a time when it was the standard. For its size, its still a very good unit to build power on and have reasonable spool characteristics. Its just about the only unit that you can make 700+hp on and weighs sub 20 lbs.

 

 

I mean it's not awful awful, but it's pretty dated. You make a lot more power with the p-trim turbine wheel which is older on most platforms, but Garrett doesn't really put any bigger turbines with the 62 or 58mm compressors.

As for the sub 20lbs, you'd be suprised -- Comps, Xonarotors, Precisions and EFRs can come with aluminum CHRAs. I'm not sure if you can get them on the airwerks platform yet. Combine that with any of the good quality lighter vband turbine housings and you're in business for a light turbo -- a far cry from the old school massive cast iron housings that weigh 25lbs by themselves. I have held a CT6X turbo with a big boy 120mm compressor wheel and was shocked at how light it was.

 

So you're running a single scroll manifold with twin scroll turbo? If so your "QSV" data directly contradicts URQaudiguy's extensive testing, where he gains nearly 1000rpm of powerband with the QSV versus an open manifold alone . I'd have to call into question your boost control and QSV control systems. Headflow doesn't completely change a turbos operating characteristics.

Same. Sure, full-race helped create the EFR line, so they're going to tout them, but if they didn't work, they wouldn't be as successful as they are. BW is currently revising the entire Airwerks lineup, I'm sure they'll be back to the EFR line in a few years to revamp it with the latest advances.

Talking with the BorgWarner guys at PRI, they said several times that they emphasize reliability over performance in their lineup, which is why you don't see them running extremely thin hubs and blades on their compressors. They build those things to go a million miles. They are tested for blade deflection at operating speeds and beyond. Pushing the boundaries in turbine wheel tech has cost them quite a few setbacks, but you can't really fault them for that. We need and want companies to push the boundaries, otherwise things become stagnant.

So while I get what you're saying x622, to toss out the entire EFR line as sub-par, and to deny the advantages of the TiAl turbine wheel is simply overreacting. Maybe the gains aren't as good as they were hyped up to be before their release, but they do what they're supposed to when it comes to rotating assembly characteristics. Give credit where credit is due.

Precision would never be able to carry out the kind of extensive R&D as Borgwarner, simply because they're a small company and can't afford to. Garrett seems to take a much more incremental approach to retail turbo development, with nothing being completely redesigned besides the compressor wheels in the GTX line, and possibly the new GTW turbines?

BTW if you're looking for a 35R with a higher flowing turbine, the GTW3684R has a 71mm inducer, significantly larger than the 68mm 35R. Unless it's a old design borrowed from the GT line, I'm guessing this is a higher performing turbo. Why it's not available in a T4 divided turbine housing is beyond me.

 

I know! I want the GTW lineup to be ball bearing and come with a wider selection of hotsides. It looks really promising. I have my suspicions that the new 71mm turbine might be the new iteration of the older p-trim turbine wheel, which still makes amazing power to this day.


As for the gains with the QSV, please send me his thread. Here was the setup and what I found, keep in mind all numbers stated are in 4th gear.

Originally, I ran a 3 port on the Tial 44mm gate with boost reference plumbed in to the gate to keep it shut and ran another 3 port to straight to the QSV. Both were mac valves.
With this setup and the resulting lazy spool of the 3 port, the QSV picked up a more respectable 500rpm and gave me 25psi @ 5000rpm vs 25psi @ 5500rpm.It's also important to note that with the QSV vtec was set to 4800rpm, as it wasn't happy down any lower because of the backpressure.
Attachment 326775
While this tune was garbage, you can still see what the valve did. The cyan line was the result with the valve open whereas the other ones were with it closed.



I then swapped the 3 port on the gate to a 4 port, and the end result was I was making 25psi @ 4900rpm regardless of if the dutycycle sent to the QSV. I then completely removed the QSV and was then reaching target at around 4850rpm, in addition, fuel had to be added everywhere. At this point in time I had to swap to a spare turbosmart 45mm gate I had lying around because the Tial gate's diaphragm decided to start sticking and I was hitting 35+ psi with a 25psi target. Fun.


I'll go get my datalogs and post the immediate after of removing the QSV so you can see the leaning out, and the final run after being on the dyno so you can see the before and after. If you'd like to see any other datalogs let me know, I have... too many.


As for writing off the EFR lineup, I'm not doing that! Not yet anyway. Not until I have back to back data. I have a sneaking suspicion that it might reach target boost a bit faster, but we'll see.