Lithium

Oops! Sorry. Yes, that. Though I do like the idea of two solid sized turbos, a huge meaty power delivery AND revving to the moon which is how I got sucked into the off topic bit. Would love to know more detail on what @Butterswas talking about re: dyno plot, I know the exact detail / plot itself may not be available... I'm guessing it epitomises the description you gave there? @khezz, when you're "keeping it on the boil" etc it can be surprising how shit a power delivery can be used to make a car go fairly quick. One of the cars I tune runs an over-cammed (imho) 4EFTE head in a Toyota Starlet running a stock Toyota W58 gearbox. The engine doesn't start to really trap useful air until around 4500rpm, but the block/bottom end setup is not safe to really run much over 8000rpm so basically ANY turbo you put on it is going to be laggy as f**k but you don;t even get to rev it to the moon to make up for it. When I first got involved with it the thing was running a Turbonetics turbo which wasn't full boost (~23psi) until around 6000rpm (later after gearshifts on the strip) and still managed to run a 10.7 @ 130mph, driving the absolute shit out of it - but that's basically not making full boost until somewhere between 6000 and 7000rpm on the strip and only revving to 8k with a 1.5litre engine. He ended up putting a G25 550 in there to make it a bit easier, and we ended up with full boost at 5300rpm (remember the engine doesn't really start even driving the turbo until after 4000rpm, so that is a VERTICAL dyno plot lol) and way better transient response... brought it down to 10.5 @ 133mph at the same boost/similar power, and most of the improvement is just because of response, and that's him being less brutal on the gearbox. Seems off topic, but the moral of it is you can go surprisingly quick at the strip with a rubbish power delivery even without flatshift or a sequential transmission - that Starlet has a woeful power delivery and still does solid 10s without a huge amount of power. Clearly improving the delivery helps, and it would go much quicker with a better transmission, but still be a dog for anything else. The point is 1/4 mile times are pretty much the worst measure for how nice a car is to drive.

Firstly, I was responding to your conclusion "extra brownie points for doing it in series", I had no issue with the concept in sequential. Second, you wouldn't run a Lysholm as the first stage - unless I'm missing something you cannot move more through the compressor than it can flow by itself. You'd need a significant bypass to allow a 90+lb/min turbo to be able to move all the airmass it needs to achieve ~1000hp, and if you're doing that kind of bypass then it's a sequential setup. With a multi-stage compressor setup you run the biggest compressor first and use the smaller ones as pressure ratio multiplier, you've increased the density of the air going into them so they don't need to move a larger volume of air than they are able to. If the Lysholm was able to support 90+lb/min then there is no point running the turbo at this point. Now for my other comment, I have to admit I stuffed up my math - more because of rushing it and mis-juggling the numbers than anything else. To give an idea of why I thought that... I assumed 65% adiabatic efficiency for a twin-screw (I think this is VERY fair) at pressure ratio 1.5 with 24c pre-compressed air temp then you end up with 84c outlet air temperature. If you WERE going to compress this by a PR of 2.0 using a typical centrifugal compressor without intercooling it first then you'd end up with a second stage temperature of over 200c. Where I messed up my calculations was that I had a brain fart with the "final density" calculation and ended up effectively treating it like you needed to cool 207c using 84c ambient temperature - not 24c, so that ended up with a MUCH lower air density in the intake manifold... 88lb/min mass flow "conventionally turbocharged at 30psi" became 73lb/min at the same "boost level" and made it look like a fail, but actually it's not that bad at all. It would actually "only" fall down to 85lb/min.... so still less dense than single stage but you're right, non-intercooled between the two stages wouldn't be a game changer. That aside, I prefer the idea of a regulated 2-stage turbo setup for spark ignited small displacement/"moderate boost level" engine than a compound supercharger/turbocharger setup.

That's more like I would have expected It is a pretty cool idea but you do really need to intercool between each stage otherwise the air density becomes rubbish and as such the second stage becomes a lot less efficient and you don't necessarily end up with the results you could while also having mechanical drag added to it... But the general concept is awesome

Very interested to see where it ends up That'd be a good result!

Ahh, yeah sometimes hard to tell with tone Meaningful boost by 5500rpm isn't inconceivable, though full boost at 4000-4500rpm - I'll have something to learn if your setup manages that, unless you've missed out mention of nitrous or that the turbo tweaks include changing to much smaller wheels etc 😮 In saying that, I've been involved in brain storming for a twin setup for >1000hp @ hubs hoping for full boost around 4000rpm on a 3litre and we have hopes that it's achievable but that's getting pretty cunning with the overall setup - from what you've said you're not getting up to the kind of mischief we're planning

I wasn't trying to talk you out of them, just making sure you did appreciate that it would be super laggy (I never said they won't spool either) - not just a bit lazy. I'd call that realistic, not negative. Fwiw, something measured using no-lift-shifting and generally being used at full throttle/max possible rpm isn't the best metric for what it's going to drive like but yeah. Anyway, you sound like you are pretty happy with the compromises - as I said at the start, I love the old school JDM drag "style" and look forward to seeing results (you don't ACTUALLY get to see what this kind of combo dynos like etc very often at all) and videos... miss this kind of build, and keen to hear it.

Sure, good torque from 4000rpm on a VCT 2.5 litre engine with a SINGLE one of these. I'm hoping I'm missing something and you aren't assuming that two T67s will spool similar to one? It REALLY doesn't work that way. As I said above, in terms of spool/response - you shouldn't be expecting it to spool like you have a single T67 on a 30% bigger engines, you should be expecting it to behave like you have two engines which are 56% smaller than an RB25 each driving one of these turbos (so a 1.6litre engine), but without the spool benefits of VCT, and I doubt very much the turbine wheel will be any different in design or material over the Trust units and that will be the biggest influence on transient response etc. They are going to be way way laggier than a single one on an RB25, absolutely zero doubt about that. It will drive OK off boost because of sheer torque etc, but it's going to be laggy compared to a 76mm Precision turbo.

I for one fully approve of "doing what is going to make you love the car afterwards", I discovered JDM drag racing in the late 90s with bootleg VHS etc of Japanese cars with highmount twins having loose as f**k 1/4 mile runs making amazing noises and looking like they're rolling coal so I definitely feel some this. I've definitely yarned with mates about it, and approve it if you're understanding and ok with the fact it's not going to work as well as a nicely matched-for-purpose single turbo - I know I personally wouldn't feel like I've built "my car" if I had just another "I'll have the 7685 RB32 auto package deal everyone knows exactly how its going to go and sound deal thanks"... but it's still probably worth having a real solid think about what you should expect and how much it's going to be seem worth it after you've got it going. I could be completely wrong, but when you say "Even if they don't spool until 5000rpm" it sounds like you're talking about that as a worst case scenario... while I'd guess that to be a best case scenario, or outright optimistic. A single T67 25G is making solid boost on an RB26 meaningfully after 4000rpm, with twins you'll effectively have two RB16s running these things - but they are feeding a common intake so may not perfectly balance so not even work at the same efficiency of the single version of the same thing. I'd not be shocked if you're waiting until 6000rpm or more for things to start getting spicy on a loaded dyno run, let alone on the street where the big ol' journal bearing turbos do their best to demonstrate why transient response matters so much when it comes to having an enjoyable car to drive. The billet comp wheel upgrade is going to do nothing for spool/response. I suppose the trick here is you're probably looking for "big power" to a degree, and the kit will use Trust's 3-bolt flanges - otherwise there could always be the option of throwing something less garbage on there. When we'd talked about this kind of thing hypothetically we'd discussed raiding the Kinugawa/Mamba parts catalogue for a Greddy 3-bolt TD05 turbine housing and TD05-16G6 (or generally something more appropriate than a TD06H-25G) CHRA with 3" inlet comp housing so you have something still capable of up in the 900+hp @ crank territory, still look like a legit high mount twin TD06 kit but actually have a usable power delivery. These days a custom kit with the likes of a pair of G25-550s or XR4951Ss on small a/r vband housings would be hard to argue with for a 1000hp high mount setup

Dang. Forgot to attach my S361SXE compressor map with the RB25 (green) and RB30 (red) lines on it and SAU immediately blocked me from editing the post, so here is the previously promised map

Yeah. I probably should have spent more time on this but it did feel like there would be a LOT to cover to get on the same page. I guess this simplest way to try and illustrate it - and PLEASE people don't take this as gospel, this is again - more to illustrate how this can vary, and why I am confident that this would be fine on an RB25. What I've done here is taken the power we pushed the S362SXE and assumed we're using about 66lb/min of air to achieve that using a high ethanol blend on a hub dyno, and drawn a line (in red) which represents that max airflow requirements of the RB30 at different pressure ratios until it hits the choke line of the compressor graph. I've then drawn a green line which is simply scaling the RB30 airflow needs by 83% (from 3litre to 2.5litres). This is coarse AF, realistically this would depend on the head and bolton mods etc - given @Looney_Headhas upgraded cams and a ported head etc and the RB25 in this story is going to be limited by the stock cams, it'd be a pretty conservative estimate. You can safely assume that an RB25 would sit somewhere between the two lines at any given boost level, and in this particular case the fact that it needs more boost to make power means that an RB25 would be able to make MORE power than the RB30 with this turbo. Sure. The turbos weren't available before 2020 that I know of, I know of no failures with them, we've had no issues with them (have used a few now), the performance has if anything exceeded our expectations so far. The only "question mark" things I can think of, which are really not-big-deal-things, is needing to do things like make sure that the speed sensor plug is tight if you aren't using a speed sensor (we lost one and developed a big boost leak haha), and general "make sure everything is tight" checks are well recommended. At least in my experience so far, turbos which are going to be a problem due to quality control etc will show their issues in the first 1000km/first "hard push" which is why things often get a bad name quite quickly. As I mentioned above, I've STILL not heard of a failure with them - which is part of why we started giving them a go in the first place. Obviously can't guarantee anything, but just sharing my experience so far.... I'll be the first to raise a red flag if I have reason to think a product is going to give issues.

so with my RB25 being as a street car it would be living low / mid boost most of the time. to make the same power the rb30 does with the 362 the rb25 would need to be at a higher rpm/ at a higher boost level so for the rb25 to make somewhat similar power at same low/mid boost i would need a bigger compressor to push more air no? Are you saying you think the turbo needs to push more air to make the same power on a smaller engine? If so, no - it doesn't. The amount of power you make has a very close correlation to how much air the turbo moves. The only reason you may need a different sized turbo is if the compressor (or turbine) will be inefficient in the operating area it would work with on a different engine. The RB25 and RB25/30 are similar enough that this will not be a thing. All going to a bigger compressor (or bigger compressor and turbine) is going to do is increase lag - ESPECIALLY if you have stock cams. 'It has been proven to support 900whp (US dynos)/ near 600kw (Oz dynos) on the right setup. People are often oversizing the hotsides on Borg Warner Airwerks turbos and living with more lag than they need, imho. This comment was purely about the hotside. The 61mm compressor wheel will not support that by a long shot. The 64mm one won't either - on an RB the 64mm has bugger all advantage over the 61.4mm actually, especially on an engine which doesn't run >30psi. but i am looking for some comparable turbos between the 3076 and 3582 to as the 3076 would be jsut to small for my goals and i have read a few showing that the 3582 is to big for a stock rb25 and not street friendly This is pretty much why I suggested the G30 770. It's a slightly smaller turbo than the GTX3076R (same size turbine, smaller compressor) but flows similar to a GT3582R. It basically means you end up with something with a similar boost threshold to a GTX3076R, better transient response and able to make GT3582R power numbers.

There is some pretty confused stuff in this so a bit to unpack, I'll try and be as helpful as possible on this as it feels like you have quite a bit to learn and could easily end up choosing something that REALLY doesn't suit what you want. Don't rush out and buy something until you've done a bit more research/thinking about it, but I'll try my best to get us on the same page etc. So to address your points: RB30/25 build with Cams for SXE360 on e85 he made ~650whp This is @Looney_Head's build, I just tuned it. He started with the S360SX on E85 and that tapped out at 600hp. The S362SXE (61.4mm inducer) we've seen over 650hp and could potentially stretch to 700hp on E85 depending on the setup and the dyno. FFP , T3 Turbo Manifold w/ EWG The Borg Warner turbos are best used with T4 twin scroll manifolds, I wouldn't use one if you are going T3 manifold. More on this soon BP 93 (Only 2 e85 stations in my STATE (not city, State) and not consistent ‘e85’) So you're using flex fuel? Hope you know that you'll have to drop your power targets for when it's on 93? Looking for FUN Street Car. Great Drivability , Great Midrange , fine with Good/ok Top end (actually being realistic here) This is all open to interpretation of what you find driveable, but the S362SXE would be quite doughy on an RB25 - like potentially mid 4000rpm before full boost. The S366 will be getting closer to 5000rpm, they are not things I'd call great for street use. This range is great for 3litres, but smaller it depends on how much you mind it being a bit lazy under 4000rpm. I think I may need a slightly bigger SXE364 / 366 to compensate for the rb30/25 additional mods the has It doesn't work this way, if anything you often need a bigger turbo for similar power on a bigger engine. Any power you can make on an S362SXE on an RB30, you'll make happily with an RB25. What I should add here is that you really should be looking at cams and definitely valve springs if you want to go much past 500whp. Which has been proven for 900whp the SXE362 or EFR? There is no way you're going to make 900whp with an S362SXE, why are we talking about 900whp? You said 600-700hp max earlier But people *have* tapped 900hp on hub and some roller dynos with an EFR8474 but really you'd be going an EFR9280 or an S369SX-E if you wanted that kind of power... and they WON'T be a fun street setup on an RB25 unless you don't mind waiting until 5000 before it gets fully spicy. but not sure If I should be thinking more ‘2020 era’ for something with better performance relatively close in budget 1-1.5k I'd consider a Garrett G-series copy from Pulsar turbos, they perform well if you don't mind the ethical grey area of running a turbo which is cheap because they've let someone else do the R&D. Something like this will get you over 600whp on E85 and over 500whp on 93 while still having reasonable response and a good price, and it conveniently comes with a T3 open housing option: https://www.pulsarturbo.com/product/pulsar-5855g-aka-g30-770-dual-ball-bearing-turbo/

Seeing as someone brought this discussion up, I thought I'd do a slight update. It turned out that the initial fuel pressure drop off was partly due to a power supply issue to the fuel pump, so we had another crack on the dyno since this. We were confident enough this time to run it at 20psi and at least put the extra timing that it was missing in at that power level, bearing in mind that we never actually tried putting it in last time due to the fuel delivery issues. This time we were running just over 85% ethanol and hit almost bang on the same ~466kw @ hubs that we did last time on 20psi, possibly slightly more - so everything seemed healthy and I went about optimising the timing. We ended up JUST cracking 500kw on 20psi with timing full put into it, but found that fuel pressure was STILL a bit of an issue with this ethanol content - more likely at this point that the fuel pump was actually just tapped out. This was a bit of a bummer, as last time we had run it up to 23psi of boost on lower ethanol percentage - so KNOW that the turbo still has more in it. Either way, we've proven it to be comfortably capable of passing 500kw @ hubs with a decent power delivery on an RB30 I'm not sure if it's going to get wound up any further or not as the thing is making good power as it is, and with RWD it's an absolute menace even in 4th gear hahaha. Here's a wee video of us playing silly buggers late last year when we hired the local drag strip, rolling 4th gear skids

If you want to get into specifics relating to the Borgs then put your questions in the Borg thread. Re: Power targets, make it specific here or in there as well. In this thread you've mentioned 600-700whp, we made 650+hp without maxing out the S362SXE. It's an easy 600+whp turbo, so any further posts in here relating to G-series or in the other thread relating to Borg try and clarify why you believe you may need bigger than x turbo or whatever. Mention what fuel you'll be using etc as well.

Maybe an unpopular opinion, but I prefer 20B sound 80% of the time. 26Bs sound great when NA and at full revs, but 20Bs sound good all the time

20B is a LOT cheaper to build and easier to install, and in some ways the sound may make people second guess what's in it as well. Hopefully with a decent matched turbo it should be peppy enough

To be fair that hasn't actually been an issue, it's been running for near a decade (granted, no daily haha) and I went for a hoon in a few months ago so I assume it's not THAT bad. The engine is likely to get pulled to put in a 20B for awesome as opposed to necessity

Haha, well the old engine was an RB20 so realistically there isn't much that is worse going to a 13B aside from reliability.

Can confirm, a mate has a 500whp bridgeport 13B R32 GTS-t and it's hilarious fun

I only disagree with you when you're wrong So in the spirit of that, I'm going to back up your sentiment here. I'd use the .61a/r hotside on a G30 660 being matched for an RB20. While we're agreeing - also relating to your early comments regarding turbine vs compressor matches etc, this is a thing I've long bashed my head in conversations with people online about - people see a specific A/R that works in an application and decide that is IT. Same with turbine vs compressor size, and generally assuming that they know how a turbo is going to behave from the turbine size. Even when I knew comparatively F*all about turbos (and I still have way too much to learn) it was pretty obvious to me that aero is going to influence this, as with the dynamic between the compressor and the turbine. As I've learned more I've only grown to further appreciate this. If someone chose a 1.01a/r v-band G35-1050 over a 1.05a/r EFR8474 because the G35-1050 (G8468 if you're going to refer to it with Borg nomenclature) expecting a significantly more responsive/spooly turbo that has a more restrictive hotside then they're going to end up with something a little different to what they anticipated. Broad rules of thumb based off previous experience can work, but then there are times they will lead you WELL astray - depending on the setups. Like even the same turbine housing/wheel combination can perform significantly different in terms of what it can sensibly support depending on the compressor it is matched to.

Does my head in, I have no idea why workshops do that kind of thing - feels like it's just getting customers to dump all the money with them as well as ending up with a big number result at the end which get other people's attention, but pushes the person who is actually wanting to build their "dream vehicle" past the point of no return and it's now more or less a dedicated straight line acceleration car. I guess if people believe it when they're given that speel they potentially are half trying to talk themselves into it anyway, in which case that's fine haha. The only way that would seems plausible is if the rest of the setup has significantly changes, like adding a solid amount of displacement and/or fixing issues that the setup previously had haha. Even then, with the huge amount of added "mass away from the axis" will cause a much much higher moment of inertia... if the threshold doesn't seem that much worse, the transient response will be rubbish by comparison. I would wait for a couple of results to come out, but if you're looking for something bigger than a 6870 that still drives like a reasonable street setup then there is a possibility than the G40 1150 *could* be a really solid option. I've already seen someone claim to go from a Precision 7275 to a G40 1150 and claim a significant improvement in driveability and not losing any power but no specific data yet. The 6870 is something of a beast in terms of how much power it makes versus how usable it is, so really anything that is capable of making more power is likely to add more lag as well. I suspect there may be interesting releases before the end of this year that could mix it up, but we'll see...

This was said ages ago but just for sake of accuracy - people have definitely been able to squeeze more out of the 6875 than they could with the 6875. The turbine definitely helps, even if the 68mm compressor is pretty much tapped it does help squeeze enough that if you're looking for absolute max power from a 68mm inducer it's worth doing.

Awesome build, and the 8474 is an excellent turbo - a made recently made 620kw @ hubs on his 2JZ with an insane power curve. If you aren't shy with it then you're going to completely slap the old T76 setup everywhere Always concerns me seeing Bosch 2000s, especially on builds like this. Hope you don't have any issues - look forward to the final results!

Sounds like we're mostly on the same page. "Bullshit" may be a slightly strong word, but historically Garrett (and other) manufacturers tended to set the hp claims as something that were pretty achieveable without relying on ethanol or pushing the turbo to the max. Garrett's rule of thumb was basically hp = massflow @ PR2 x 10, so a semi-conservative estimate of what power it was capable of at ~1bar of boost on premium gasoline. Now it appears to be if you run it potentially off the map on E85. Back in the day Garrett used to claim the GT3582R as a 600hp turbo, they've now refactored their claim for the same turbo to 675hp (which is still pretty realistic, the old claim was probably overly conservative for a guideline) but then you have the G25-660 which is advertising it as being suited for 15hp less than a GT3582R. I would *love* to see someone make an honest 660hp @ crank on an RB26 on pump gas with a single .72 G25-660. In fact I'd like to see one get anywhere near what a GT3582R is capable of. If it could be done it would be a very rewarding street setup, but. The native .72a/r G25 hotside flows less than a .63a/r GT30. I'd not feel super confident running that to 300ishrwkw on pump gas on an RB25.