Jump to content
SAU Community

Recommended Posts

Hi all, 

There have been all kinds of conversations on and off SAU about single turbo options for RBs (and other engines really) in the current state of the game and while I've not seen any evidence of compressor maps being inaccurate, different manufacturers use different ways to normalise and present the numbers, different cut off points, and different ways of estimating hp from compressor flow which can result in people expecting too much or too little from their turbos.

I've decided to semi-roughly translate it all into an even measure and shove what I consider "the most relevant" turbos for most people who would be talking about turbo choices in here, and split the compressor flow potential they have at 20psi, 25psi and 30psi - using "65% compressor efficiency" as a cut off.   Those are fairly familiar and reasonable reference points, and as you exceed 65% I view it that it's where the intercooler and hotside are starting to have to carry disproportionately more of the load to keep things happier.  This is the "you're pushing it" zone, imho - even if it's not really maxing the turbo out.    It seems to be a reasonable way of gauging them on a similar scale if you're going to being matching the turbos for real world performance.

It may not be PERFECT but I think it gives enough of an idea of how they may compare, I'm pretty sure all these turbos have hot side options which are pretty close to supporting the exhaust flow needed these days - give or take, but some of the numbers here may explain a few interesting things seen..

Hope it's vaguely interesting reading :) :)

Turbo 20psi 25psi 30psi
GTX3071R  50 52 53
GTX3071R Gen2 51 52 53
EFR7670 53 56 57
S257SX-E 54 57 60
G25 660  55 60 59
GTX3076R 59 61 62
Gen2 GTX3076R 59 61 62
GTW3476R 60 64 67
EFR8374 65 68 69
GTX3582R 70 72 74
Gen2 GTX3582R 70 75 77
EFR9180 74 79 84
GTX3584RS 75 80 84
Edited by Lithium
  • Like 1
  • Thanks 1

And because people like this kind of thing, a loose translation of what kind of area these flow figures may convert to in dyno numbers if everything else is up to the task.  
 

Bare in mind, these numbers are not what I'd say are "on kill", but where the compressor is starting to really earn it's keep - they should have more to give, but the things start working harder to get there and you start moving more into "how optimised is this setup" territory.  These numbers aren't a guarantee of what setups WILL hit using these turbos, it's what the turbos should be able to achieve working reasonably hard - but not too hard... and of course me making a bunch of assumptions lol.  

 

Turbo 20psi 25psi 30psi
GTX3071R 339 352.6 359.3
GTX3071R Gen2 345.78 352.6 359.3
EFR7670 359.34 379.7 386.5
S257SX-E 366.12 386.5 406.8
G25 660  372.9 406.8 400
GTX3076R 400.02 413.6 420.4
Gen2 GTX3076R 400.02 413.6 420.4
GTW3476R 406.8 433.9 454.3
EFR8374 440.7 461 467.8
GTX3582R 474.6 488.2 501.7
Gen2 GTX3582R 474.6 508.5 522.1
EFR9180 501.72 535.6 569.5
GTX3584RS 508.5 542.4 569.5
Edited by Lithium
  • Like 1

Interesting to see how little extra power is generated by some turbos with an extra 10 psi (as low as 14kw) while other turbos gain much more ...such as the EFR9180 with an extra 68kw

Just now, KiwiRS4T said:

Interesting to see how little extra power is generated by some turbos with an extra 10 psi (as low as 14kw) while other turbos gain much more ...such as the EFR9180 with an extra 60kw

100%.  One of the reasons I decided to break it down as often people just seem to look at the power claims or the peak airflow levels without considering the compressor efficiency.     

The EFR9180 doesn't really seem to be a turbo best suited to a setup which can flow a heap of air in it's own right, at least based off what I've done here it would indicate it'd roll over early if you treated it like a 90lb/min turbo and tried to flow that at relatively low boost.   On the flipside, you could potentially shove one on a solid RB26 and crank a heap of boost through it and potentially make more power than you would with an RB32 with big cams capable of making lots of power without heaps of boost

https://bmw.spoolstreet.com/attachments/c92a58335a9c242d64ec466cfc9172ee_zpsniwwkwsu-jpg.13075/

 

Man I wish the 8474 was a thing.. maybe too good to be true.

20-25-30psi

77-84-91 lb/min

also, G series hopefully scales well. Med and large frames are going to be beasts. if the G25-660 is smaller in every way than a gtx3071.

 

Edited by burn4005
4 minutes ago, burn4005 said:

Man I wish the 8474 was a thing..

20-25-30psi

77-84-91 lb/min

also, G series hopefully scales well. Med and large frames are going to be beasts.

 

Fark yeah, good call on mapping the 8474 in there - GET ON IT Borg.  That would change the game if it delivered on that :o    

Also +1 on the G-series scaling.  

This stuff is all on paper, but imho the simple job of just compiling some points into one list kind of explains some of the things we see happening in practice on some of these turbos.

The G25-660 with a .92a/r hotside could be quite a surprising beast on an RB25....

Edited by Lithium

yea the 8374 getting close to all in at 26psi/65%, but the 84 is just getting started.

literally off the chart... might be hitting a sonic stonewall or something so they stopped mapping to the right, still fine rpm wise.

Edited by burn4005

I agree Garrett has the potential to do great things with the G series . I'm thinking the compressors and turbines wouldn't have to be a whole lot bigger to put them in the power range that the fast road market is really interested in .

If they keep the same frame size and have housings similar to the G25s , in for example a G28/G30 series , then we can have efficient responsive turbos . Keeping them compact and having reliable hot sides is where they could have it over the EFRs .  

24 minutes ago, discopotato03 said:

I agree Garrett has the potential to do great things with the G series . I'm thinking the compressors and turbines wouldn't have to be a whole lot bigger to put them in the power range that the fast road market is really interested in .

If they keep the same frame size and have housings similar to the G25s , in for example a G28/G30 series , then we can have efficient responsive turbos . Keeping them compact and having reliable hot sides is where they could have it over the EFRs .  

The funny thing is that G25-660 already has a hotside which is comparable to the biggest hotside available for the EFR7670, and compressor wise flows slightly better at all the points one would be likely to use on an RB25/26 if they were looking for a fast road car.  Need more real world results to get a gauge of what spool is actually like, but I can't see any reason why despite seeming so small - a G25-660 is actually big enough to cover the needs of your typical punter building a quick road car

Edited by Lithium
  • Like 1
  • 4 months later...
And because people like this kind of thing, a loose translation of what kind of area these flow figures may convert to in dyno numbers if everything else is up to the task.  
 
Bare in mind, these numbers are not what I'd say are "on kill", but where the compressor is starting to really earn it's keep - they should have more to give, but the things start working harder to get there and you start moving more into "how optimised is this setup" territory.  These numbers aren't a guarantee of what setups WILL hit using these turbos, it's what the turbos should be able to achieve working reasonably hard - but not too hard... and of course me making a bunch of assumptions lol.  
 
Turbo 20psi 25psi 30psi
GTX3071R 339 352.6 359.3
GTX3071R Gen2 345.78 352.6 359.3
EFR7670 359.34 379.7 386.5
S257SX-E 366.12 386.5 406.8
G25 660  372.9 406.8 400
GTX3076R 400.02 413.6 420.4
Gen2 GTX3076R 400.02 413.6 420.4
GTW3476R 406.8 433.9 454.3
EFR8374 440.7 461 467.8
GTX3582R 474.6 488.2 501.7
Gen2 GTX3582R 474.6 508.5 522.1
EFR9180 501.72 535.6 569.5
GTX3584RS 508.5 542.4 569.5
My real world result is quite close to your theoretical numbers, Lith.
The GTX3584RS with 542 at 25psi and 569 at 30psi vs mine with 550 at 29psi.
2 hours ago, K_arlstrom said:

I actually updated my list with a bunch more turbos (including the black series EFRs) months ago but this thread had very little interest so didn't think to put the updated version in here.

No photo description available.

 

7 hours ago, whatsisname said:

My real world result is quite close to your theoretical numbers, Lith.
The GTX3584RS with 542 at 25psi and 569 at 30psi vs mine with 550 at 29psi.

They were very much thumb sucks so that's nice :) What did yours make at lower boost levels?  

  • Like 1
2 hours ago, Lithium said:

I actually updated my list with a bunch more turbos (including the black series EFRs) months ago but this thread had very little interest so didn't think to put the updated version in here.

Thats what we get for just reading and not typing anything :whistling: 8374 might still be the better "scary fast street RB"-turbo

Edited by K_arlstrom
They were very much thumb sucks so that's nice [emoji4] What did yours make at lower boost levels?  

Well they were quality thumb sucks [emoji4][emoji106]

 

It made 533kW at 26-25psi. Not sure what it makes at 20psi. I remember it making 421rwkw at 19psi during early tuning a year or so back (posted info in the RB30 results thread) but it was far from optimal and it was prior to properly setting up the boost control.

 

Shaun's Dynotech DD dyno is a bit stingy too, having ran both my 33 (345rwkw at the time) and Stage 2+ Golf Pirelli GTI (193fwkw) on Shaun's and then a Mainline roller dyno across town, where they both picked up 21rwkw (366rwkw) and 19fwkw (212fwkw) respectively.

 

I think another 1psi and a slightly more optimistic roller dyno would put my 33 right in line with your theoretical 569 figure.

 

22 minutes ago, whatsisname said:

Well they were quality thumb sucks emoji4.pngemoji106.png

 

It made 533kW at 26-25psi. Not sure what it makes at 20psi. I remember it making 421rwkw at 19psi during early tuning a year or so back (posted info in the RB30 results thread) but it was far from optimal and it was prior to properly setting up the boost control.

 

Shaun's Dynotech DD dyno is a bit stingy too, having ran both my 33 (345rwkw at the time) and Stage 2+ Golf Pirelli GTI (193fwkw) on Shaun's and then a Mainline roller dyno across town, where they both picked up 21rwkw (366rwkw) and 19fwkw (212fwkw) respectively.

 

I think another 1psi and a slightly more optimistic roller dyno would put my 33 right in line with your theoretical 569 figure.

 

What tyres are on the back? it does sound noisy (implies they are sticky tyres)

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



  • Similar Content

  • Latest Posts

    • Well, after the full circus this week (new gearbag, 14 psi actuator on, injectors and AFM upgraded, and.....turbo repair) the diagnosis on the wastegate is in. It was broken. It was broken in a really strange way. The weld that holds the lever arm onto the wastegate flapper shaft broke. Broke completely, but broke in such a way that it could go back together in the "correct" position, or it could rearrange itself somewhere else along the fracture plane and sit with the flapper not parallel to the lever. So, who knows how and when exactly what happened? No-one will ever know. Was it broken like this the first time it spat the circlip and wedged itself deep into the dump? Or was it only broken when I tried to pry it back into place? (I didn't try that hard, but who knows?). Or did it break first? Or did it break between the first and second event of wierdness? Meh. It doesn't matter now. It is welded back together. And it is now held closed by a 14 psi actuator, so...the car has been tuned with the supporting mods (and the order of operations there is that the supporting mods and dyno needed to be able to be done first before adding boost, because it was pinging on <<14 psi with the new turbo with only a 6 psi actuator). And then tuned up a bit, and with the boost controller turned off throughout that process. So it was only running WG pressure and so only hit about 15-16 psi. The turbo is still ever so slightly lazier than might be preferred - like it is still a bit on the big side for the engine. I haven't tested it on the road properly in any way - just driven it around in traffic for a half hour or so. But it is like chalk and cheese compared to what it was. Between dyno numbers and driving feedback: It makes 100 kW at 3k rpm, which is OK, could be better. That's stock 2JZ territory, or RB20 with G series 550. It actually starts building boost from 2k, which is certainly better than it did recently (with all the WG flapper bullshit). Although it's hard to remember what it was like prior to all that - it certainly seems much, much better. And that makes sense, given the WG was probably starting to blow open at anything above about 3 psi anyway (with the 6 psi actuator). It doesn't really get to "full boost" (say 16 psi) until >>4k rpm. I am hopeful that this is a feature of the lack of boost controller keeping boost pressure off the actuator, because it was turned off for the dyno and off for the drives afterward. There's more to be found here, I'm sure. It made 230 rwkW at not a lot more than 6k and held it to over 7k, so there seems to be plenty of potential to get it up to 250-260rwkW with 18 psi or so, which would be a decent effort, considering the stock sized turbo inlet pipework and AFM, and the return flow cooler. According to Tao, those things should definitely put a bit of a limit on it by that sort of number. I must stress that I have not opened the throttle 100% on the road yet - well, at least not 100% and allowed it to wind all the way up. It'll have to wait until some reasonable opportunity. I'm quite looking forward to that - it feels massively better than it has in a loooong time. It's back to its old self, plus about 20% extra powers over the best it ever did before. I'm going to get the boost controller set up to maximise spool and settle at no more than ~17 psi (for now) and then go back on the dyno to see what we can squeeze out of it. There is other interesting news too. I put together a replacement tube to fit the R35 AFM in the stock location. This is the first time the tuner has worked with one, because anyone else he has tuned for has gone from Z32 territory to aftermarket ECU. No-one has ever wanted to stay Nistuned and do what I've done. Anyway, his feedback is that the R35 AFM is super super super responsive. Tiny little changes in throttle position or load turn up immediately as a cell change on the maps. Way, way more responsive than any of the old skool AFMs. Makes it quite diffifult to tune as you have to stay right on top of that so you don't wander off the cell you wanted to tune. But it certainly seems to help with real world throttle response. That's hard to separate from all the other things that changed, but the "pedal feel" is certainly crisp.
    • I'm a bit confused by this post, so I'll address the bit I understand lol.  Use an air compressor and blow away the guide coat sanding residue. All the better if you have a moisture trap for your compressor. You'd want to do this a few times as you sand the area, you wouldn't for example sand the entire area till you think its perfect and then 'confirm' that is it by blowing away the guide coat residue.  Sand the area, blow away the guide coat residue, inspect the panel, back to sanding... rinse and repeat. 
    • The detail level is about right for the money they charge for the full kit... AU$21.00 each issue, 110 issues for a total of $2,300 (I mentioned $2.2K in the first post when the exchange rate was better). $20/week is doable... 😐
    • If planning on joining us for the day(s) please indicate by filling in this form. https://forms.gle/Ma8Nn4DzYVA8uDHg7
    • You put the driver's seat on the wrong side! Incredible detail on all of this. It looks like you could learn a lot about the car just from assembling the kit.
×
×
  • Create New...