Why Twin Turb?
 

I'm sure this question has been asked before, but let me ask it again: Why did Mazda decide to use a two-rotor twin-turbo?

Even before I owned one of these things (and before getting a taste of turbo troubleshooting and the rat's nest) I remember thinking that an NA 3-rotor would be SO MUCH simpler and more reliable.

One theory would be "Twin Turbo" marketing.
Another would be tunability (doubt it).

What do you think?

I definitely wouldn't rule out a marketing scheme, but it definitely has to go deeper than that. Turbo technology back in the 90s wasn't as good as todays stuff so I'm sure they wanted a good powerband with low end torque to help around town and the sequential twins do a great job at that. Compound set ups back then weren't that good and a single turbo solution probably wasn't feasible with technology back then for making power big enough to keep up with other modern advances. An NA 3 rotor would have been a blessing, but I'm sure they have their reasons. Personally, I think they did the right thing even with how complicated the system was and it also gives the car a ton of character because of it.

i don't know, but you could ask Koby at sevenstock..

here is what we do know, we do know that the FC was making 215hp, and they wanted another 50. the goal was 10lbs/hp. the FC turbo could not do that, and they could have gone with a larger turbo and made the HP, but then it would have been too laggy. you have to remember that every road test of a rotary points out low little low end torque the rotary has, its like how every road test says Italian cars have personality, or german cars are well engineered, its just something that is in every road test.

so they did the sequential turbos. which have both, and its cool, and its unique (the JC Cosmo in 1989 had nearly the 1st sequential setup, and satellite navigation, and steering wheel controls for the HVAC, and phone)

from a marketing angle the turbo race cars were also twin turbo, although not sequential....

Most of the reasons have been noted already. The 993 Porsche turbo used the same sequential setup a few years later and for the same reasons. They ought to know given the on-off character of the 930 turbo. I owned one and they were scary.

I've talked to Koby and others who were involved. Especially as to why they didn't go with the NA 3 rotor. The responses I've gotten over the years don't work. They essentially said that the 3 rotor wouldn't fit and would change the balance of the car in terms of weight distribution. I was the first person to install an NA 3 rotor and the motor weighs less than the twin turbo and fits fine and in no way changes the handling. They were just wrong and their blindness to what they probably should have done may have been because they were engineers and they wanted something very sophisticated. An engineer with a budget is like a hammer. Everything looks like a nail.

I don't know if I agree with you there, gmonsen. While an NA 3 rotor is a very cool rig, and I respect what you've built, it really isn't where they were taking the RX7 and rotary platforms since the inception of the turbo 1st gen or what the public (Japanese) were salivating for. Boosted JDM supercars.

That said, was the sequential aspect of the recipe incorrect/overly complicated? Yes I think so. However what we also tend to forget is that when development started and was probably rubber-stamped turbo technology and tooling for such a low volume of weirdly sized turbos was still stuck back in the 60s (that may be an exaggeration ). It's not like they had access to the EFRs of today at a volume or price point that an OEM needs.

So I guess yeah, the engineers got carried away with their complexity and of course I agree with the saying 'An engineer with a budget...' however in my estimation the decision to produce a turbo rotary sled was not, on the whole, driven by engineers, the execution was.

How comparable would a Stock-ish port n/a three rotor and a stock REW be in the drivability, fuel economy and cost to build departments? I'm guessing the three rotor would have been difficult to get the required output and torque curves out of while keeping noise and emissions resonable at the time. Also, it was the 90's everything had to be turbo. And the Japanese manufacturers had that "gentleman's agreement" at the time where they had a self imposed 276ps power limit which basically meant they were all going for nice torque output.

I was always led to believe the twin turbo was for, first, low-end torque, and, second, for weight balance and packaging. These were the days of "money is no object" one-upsmanship in the Japanese auto manufacturing industry. A complicated, and emissions-compliant, twin turbo system on an engine with notorious low-end torque deficit and already poor emissions performance (and relatively poor fuel economy to boot)? The Japanese were all about engineering their way around it!
So, to say that the engineers got carried-away, when they were given carte blance by management to go ham, isn't really fair. I imagine that a 2-rotor with some form of a turbo system was always the plan. Likely, turbo technology for the time was found to be sufficient and within the cost envelope and the rest is history.
I like the idea of a 3-rotor as much as the next person, but I don't think that adding rotors is always the answer. Fully-funded auto manufacturers have a staggering amount of resources at their disposal, and I would fully believe if someone in the program told me that the 3-rotor engine changed the dynamics in a less-favorable way from the 2-rotor.

Again... We all know that the twin turbo was the cool thing at the time. I'm not saying the engineers "got carried way", but that they were out to build a very fast car that showcased their engineering skills and resulted in one the most iconic sports cars of the past several decades. They could have produced an NA 20b-powered car for a lot less effort and cost and the result would have been a better car. It is no sweat getting 325 whp out of an NA 20b compared to 220-230 whp from the 13bTT. (Mine is making about 380 whp and has made 397, call it "400", whp.) The NA 20b weighs the same or slightly less and fits. It would have cost them a lot less to build, as its just a Cosmo motor without the turbos. I have no idea what the miles per gallon it would have gotten, but it would not have been unreasonable. I knew at some point, but forgot. Didn't care. As importantly, an NA 20b motor would have been dead reliable and all problems the Rx7 has had with its image, it's "lack of reliability", over the years would not have been an issue. Further, they could have brought out a turbo 20b version at some point to offer even more performance. So, they just made a bad decision.

I think if they put the same effort building an NA 3 rotor engine as they did building / engineering the twin turbo 2-rotor set up in the FD, the 20b rotor would be superior in reliability and power delivery. Emission and torquet output are the only things I can think of they a twin turbo set up would have an advantage.

The emissions compliant 3 rotor wouldn't have made much power. 50% more than the s5 13b give or take. Kindof apples and oranges to compare the 13brew to an aggressively ported 3 rotor.

Well, lets just thank God that Mazda did decide to make the sequential twin turbo 13B-REW FD instead of an NA 3 rotor.

The tuning potential was huge and allowed the FD to make 350rwhp+ with bolt on mods and become the go-to Rwd tuning platform for racing in Japan.

Look at what a dud the RX-8 is in the tuning industry despite its superior chassis to the FD (and most any other performance car). Just much harder and less streetable to try to make power out of an NA rotary.

Back in the sunset of the FC RX-7 Mazda brought an NA 3 rotor FC for the journalists to play with- Mazda undoubtedly learned a lot from that experience.

At one of the Sevenstock events when the RX8 was relatively new (possibly Sevenstock8 or Sevenstock9) I got a short moment to chat with Mr. Kobayakawa who was responsible to many of the decisions on the FD. I explained that I have an FD, and that I enjoy it a lot, and I asked if there was anything he might have done differently given the 10-15 years hindsight he had at the time. He said he might have installed a 2-rotor non-turbo if he could have made similar power as the Renesis engine. It's been a while, but I think he suggested that FD RX7's could have been more popular if the cars were cheaper and less complicated.

Interesting question and lots of interesting answers.

First, I think the Mazda Japan engineers have always liked a turbo rotary. Also, Japan in the 80's and 90's was turbo obsessed - they put turbos on everything, big TURBO stickers, etc. The rotary engine responds very well to a turbo, the big exhaust ports make a LOT of energy to spin up a turbo.

As far as why 2 turbos, I think this goes back to the FC. The 87-88 FC turbos had an actuator in the exhaust manifold to block off flow to one of the turbo exhaust paths to increase spool up. Then, the 89-91 turbos had "dual scroll technology" where each rotor had it's own path through the turbo with its own wastegate port to increase spool. Going twin turbo - one turbo per rotor - is a logical next step in the evolution.

I think the marketing angle is a LOT of it - all the manufacturers in the 90's were doing twin turbos. Also, I've always felt that the fact you can feel the transition on a stock car - the 10-8-10 boost pattern - is pure marketing. A PowerFC can easily be set up to dial out that transition dip so there's no technical reason for it to be there, having a smooth powerband makes more sense. But, when you go on the test drive and "feel" that 2nd turbo kick in, that will sell you on the twin turbo concept. I think Honda's VTEC having a push in power was also a selling point, you can change the VTEC RPM to have a smoother powerband but it's not good marketing on the test drive.

As far as a 3 rotor I think the reduced fuel economy made it a no-go for the US. History again, the FC was almost hit with a gas guzzler tax and Mazda went to a lot of lengths, including reducing a lot of weight, to avoid that. It's impossible to say what the EPA test numbers on a 3-rotor would have been but it does use more fuel for sure and would probably have had highway numbers in the teens instead of the 20's. US emissions would have very likely been a lot tougher to make it pass, and these are all things manufacturers need to think about - you can engineer an amazing car, but if the EPA says it doesn't pass emissions you can't sell it, and that's the end of that.

3 rotor NA's are SUPER cool, but we are seeing it from a "remove this engine and put this engine in" perspective. If the car came stock with a 255 hp NA 3-rotor, that would be a lot of fun, but the modding potential as it sits would be very limited. Intake and exhaust, make another 10-15 hp, and you're done with the easy stuff. As stated this has been a big issue with the RX-8, mod wise it's really strung out from the factory and you can gain very little with bolt ons. Mazda may not have considered aftermarket modding as much when designing the car but we are fortunate it's a turbo car.

Dale

I think Dale makes so many good points. I remember going on a test drive in 1994. I was awestruck by the acceleration and the smoothness of it. Didn't really feel the 2nd turbo kick in, but was overwhelmed by the pull all the way to 8,000 rpm. Porsche and BMW (and the Corvair) had brought out turbo cars and they were considered the leading edge of engineering prowess. Mazda being an engineer-driven company had to have and market what was leading edge.

[I just want to say that, while an NA 3 rotor might have presented issues with emissions, I am so sorry no one here has driven an NA 3 rotor (other than myself). People can theorize about the specifications, but have no visceral sense of the motor. If Mazda had brought our an NA 3 rotor making 250-300 hp, I don't really see why the performance potential of modifying it wouldn't be there. Mine just has large (some say monster) street porting and ITB's. People would also take the NA 20b and stick turbos on it.]

While trying to find how to spell Mr Koby's actual name, I came across this reprint of a 1992 Motor Trend article about the FD.

https://www.motortrend.com/news/1992-mazda-rx-7/

It seems like the magazine writer was impressed by the broad powerband, which might have been difficult to achieve with a non-turbo 20B. It also seems Mr Koby and others on the team were mostly concerned with suspension and weight, so they might have thought it preferable that the weight from twin turbos is closer to the rear of the car compared to the extra rotor and housings of a 20B. The team went to pretty extreme measures to reduce weight, as evidenced by the barely-sufficient swaybar mounts. The team certainly wasn't perfect either, in hindsight I think the wiring harness has some odd

Personally I'd love to drive a non-turbo 3 rotor even at just 250 whp, I think the improved engine sound and response would be worth it. Around 2005 or 2006 I drove my stock twin-turbo FD and an LS1-swapped FD at the same autocross event, the the instant torque of the V8 RX7 was fun and the handling wasn't noticeably worse. GIven today's technology, I suspect a 3-rotor with aluminum housings should equal or beat a 1994-spec 13BREW in terms of power vs weight. You might even be able to make it run cleaner given modern EFI and catalytic converter technology. If you play the modern-tech thought experiment, an aluminum-housings 13B with a single modern turbo might make even better power-to-weight and keep that weight further to the rear, plus smaller displacement should have an even easier time with emissions.

It’s not clear a 3 rotor would have been positioned in a way that would meet crash standards and all other engineering targets.

You are right about the instant torque and also the fact that you don't double the hp in 1 second. The power is far more controllable in a corner. You never want to floor a big turbo FD before getting pretty much out of the corner.

Also, the NA 20b as installed in the car weighs about 75 pounds less than the 13btt. You add a rotor, a housing, and a plate, but you get rid of the turbos and downpipe and intercooler.

Would have been a cool option to have a "driver's car" version of the FD with an NA 3 rotor and some short gearing making it engaging to drive.

Luckily we can make that now and dont have to follow emissions or noise laws or even a budget- so it can turn out a lot faster, lighter and look cleaner than Mazda could have made it.

I'm sure there would be more 3-rotor swaps if it was cheaper & easier to do. I haven't looked into it lately, but last I checked the big cost items were the engine itself (and likely rebuild), new subframe or Defined Autoworks mods to fit in the OEM subframe, standalone ECU and wiring harness, and new exhaust manifold. I don't remember if the 20B will mate directly to the RX-7 transmission. I already have the standalone ECU and could handle the wiring myself, but still can't justify the cost and labor that would be needed for better sound and throttle response.

It was my understanding that a factory 20B is approx. 100 lbs heavier than a 13B. Now you can build one with alloy parts, tubular header, etc so maybe that’s true, but wrt what you’d expect for a factory OE 20B I’m doubting it’d be lighter, not by a long shot.

Blue TII... Complete agree. I ran 433 gears for a while and it made the car extremely quick. Actually ended up going with 390 gears for more time in gear at expense of accel.

Scotty...I think you could cobble together a 3 rotor from a 13b's housings and rotors, plus another set, and different plates. E-Shaft is about $4500 out of Australia. Mates perfectly to stock tranny.

TeamRX8... If they are saying the stock 20b weighs more than the 13btt, then that is because the stock 20b was a twin turbo. Take the turbos off and the NA 20b weighs less than the 13btt. My motor down to the collector weighs about 75 pounds less than the 13btt.

Sure but I think that's a pretty disingenuous comparison when we're talking about what an OEM would have done had they implemented an NA 3 rotor option in 1993. You've removed pieces on both sides of the engine (manifolds) and replaced them with thin wall tubular pieces when historically the construction of those things to keep production costs in check were cast, and cast heavy. I believe a more reasonable comparison would be a modern manifolded single turbo setup, or even twin itty bitty EFRs for that matter, so long as its a tubular designed manifold.

dguy... So, instead of being 75 pounds lighter, the NA 20b from the factory might just weigh the same as a 13btt? The cast manifolds cannot weigh 75 pounds more than the thinwall manifolds. Again, the differences are that the 13btt has 2 turbos, necessary piping, and an intercooler with piping that you take off and then add a rotor and housing to get to a 3 rotor. There would probably be heavier cast manifolds, but their weight wouldn't add 75 pounds. Its all a bit irrelevant in any case. I only know of 2 other 3 rotor NA cars. Most who want 3 rotor motors are going the turbo route for more power. I remember Peter Farrell giving a talk at a dinner in 1996 where he had just finished the 3 rotor Phil Hamlin. He was saying they would get to 900 whp soon. A turbo 3 rotor was the thing. Still are damned cool.


I'm stuffing an old video of my NA 20b here. This is why I did it.

I'm saying that you're not comparing an apples to apples build in packaging or execution, and that I believe you're omitting some things that may have been removed and simplified for your build that would not fly if it were a factory backed product. The weight that you were able to achieve would not have been feasible given the budget, production, amentities, and emissions constraints that the dream 20b we're discussing would have been under. Further as you must know you're not adding 'just a rotor and housing' when going to a 3 rotor, the extra thick iron, multi piece e-shaft design, extra gear and bearing among other things add up to a not so insignificant amount of weight beyond what you describe. I say all these things with confidence as I've built a 3 rotor car (tube chassis) that has since gone from supercharged to turbocharged and I'm literally about to dyno this thing in the morning:


https://cimg1.ibsrv.net/gimg/www.rx7...f24e2ed275.jpg



And while its far from stock, I also have all of the gear we're speaking of and have been handling it as we've been having this discussion off and on (thin wall 3 rotor manifold, more than likely half the weight of the cast 2 rotor NA manifolds):

https://cimg6.ibsrv.net/gimg/www.rx7...cd9f9c4e34.jpg



Last but not least I'm not dissing your build in the slightest, this discussion is based on what Mazda could have produced as an OEM in 1993 with production locked down in ~1991 or earlier, not the awesome work that you've been able to take on and refine.

Agreed, that’s more or less what I was trying to convey too.

Back to a twin setup, packaging aside it’d be kind of cool to see a 13B with dual BW S252-SXE turbos (uniquely a 7070 sizing) that has the combined potential of 108 lb/hr at 30 psi. Heard through the grapevine that an aftermarket v-band inlet turbine housing with higher AR is coming for it. Currently you can only get that in 0.83 AR. Pretty much a rainy-day dreamers idea though.

dguy... I've seen your build. Going to be a great track car!

Stay with me a minute on the weight issue and I think you'll see my point in comparing the weight of a 13btt and an NA 20b. Let's "convert" a 13btt into a 3 rotor NA? You subtract the twin turbos, their piping, actuators, and the intercooler, and its associated piping. You add the weight of 2 more runners, 2 more coils, plus the weight of a 50% longer e-shaft and bearings , another rotor and its housing, and a thicker plate. Assuming a productionized version, the weight of the original stock 13btt and the converted 3 rotor are going to be just about equal. But I think that, given the engineers focus at the time on saving weight, they would have made these parts for a 3 rotor as light as possible. I would think they could easily have gotten the car to 2900 pounds or under. You are correct that I can't take the weight of my motor as a standard on the weight issue, if thinking in terms of supposedly heavier cast production parts, but mine weighs about 2925, which is negatively effected by all the sound deadening, full leather interior, and metal instead of plastic interior parts.

Thank you, it belongs to Wayne Graham and has since graduated to a real engine (turbocharged, I finally got my way even after I was coerced into speccing and converting the original NA 3 rotor to a supercharged powerplant :)

See, I just disagree with that. I don't consider myself an expert though I do have access to literally every single part we're discussing that was designed and implemented by Mazda as well as examples of more modern and one off construction. The engine would not have been lighter had they had to follow their design ethos. I do not disagree that it COULD have been lighter and CAN be lighter when designed today, I'm saying that it WOULD not have been given the way we know that Mazda designed and built things in that era.

I have a couple more 20Bs that are more stock as well as a few that are disassembled that I think I'll weigh sometime (maybe even tomorrow) I think that people will be amazed at the extra gravity in the multi piece e-shaft, thick center irons and other extra hardware thats there. Im also curious to see what the thin 20b irons weigh compared to REW irons.

almost off topic, but the JDM tuners went through a twin single turbo phase in the late 90's early 2000's, and its complex, but it works. the poster child are the KSP cars (FC and FD) full bridegport with HKS 3037's, they were in the 700rwhp range. FEED did a couple too, there is a sinister looking black FD that was like 600hp.

these days you can buy a single turbo that is a better sizing for that kind of hp, and its simpler to plumb it, and much cheaper (you don't buy two of everything), but its also less cool...

I remember there was a 20b triple-turbo car some tuner did. Plumbing nightmare.

Kevin Wyum back in the big list days was trying to do a dual aftermarket turbo setup, I don't think it was sequential but just 2 smaller turbos. I want to say someone on the forum got the car and posted about it at some point, supposedly the setup worked pretty darn well. He was thinking about making it as a kit before his untimely passing. He's the guy that did the big ASP intercoolers.

Dale

that is ReTed's favorite car! i forget the name of the shop, they built some other badass cars
https://cimg7.ibsrv.net/gimg/www.rx7...7622dbd006.jpg

and just for funsies
https://cimg9.ibsrv.net/gimg/www.rx7...1af60d9cb9.jpg
https://cimg4.ibsrv.net/gimg/www.rx7...87b92d97bf.jpg

Things to remember on parallel twin turbo builds.

You add the AR of each exhaust side to get the full turbo AR. So the 0.83 plus 0.83 is a 1.66AR.

Learned that from the Yamaguchi FD book as well.

Another thing to remeber is you add the surge lines of each turbo compressor together.

Combine those two factors and you can see why parallel twins have garbage low end power. Even if you go smalm on each tuebine AR you are surging the compressors when they try to spool on the low end.

But EFR 7163 has some compressors with really good surge lines. Even added together the surge line is better tha efr 9180 singe and can flow more.

The mixed axial and radial flow exhaust wheel is the wildcard with the rotary though. Mazda testing showed the rturbo strongly prefers radial flow on the rotary. The "impact" turbine wheels for rotaries are heavily skewed toward radial flow.

and this is why it was done.. in 1999, when his was more common, they were either maxing out the T04R or living with the big laggy TA45/T88 turbos.

also all most of these setups are on 550-700hp cars, so it made sense. things have changed though, and not only do turbos come in more sizes, but people are building milder engines to do more street/mountain driving, so you don't see these twin setups anymore.

they are cool though

Yup, and what was old can be new again.

The dual 7163 vs 9180 is especially relative because both those turbos came directly from Indycar.

Honda went with the big single 9180 while the other manufacturers used twin 6758 turbos for their Indycar engine programs.

Indycar/Borg Warner developed the 7163 as the replacement 6758 and disallowed the big single having appeased Honda with the now improved twins performance.

So what I want to know is...
if you twin 7163 on a big bridge to spool it, run the exhausts parallel but the compressors serial for an insanely good surgeline and no exhaist choking-
what hapens when you max the serial compressor flow at ~400rwhp and then whack the compressors into parallel flow for more cfm potential?

Does power stay the same and then rise since the cfm output jumps when psi at the compressor drops (so stabilizes)?

What about a twin charging design like Volvo uses: supercharger/turbo?

https://www.wardsauto.com/engines/vo...-racing-engine

Obviously there are packaging and tuning concerns, but they somehow managed to do it on warranteed cars. Some kind of electric supercharger / hybrid would reduce the parasitic load at low RPM.

Also, have you guys seen this? I wonder if Mazda has already gone down this road of inquiry:

The move toward lower hp, but better spooling and more reliable motors has been great to see. Always wondered why there hasn't been as much experimentation with superchargers. I don't remember what happened to Colin's turbocharger/supercharger setup. Talked to Logan Carswell about doing a 3 rotor supercharged and think he may have built one or may still be working on one.

That video is a guy who doesnt understand how a rotary works explaining how to " improve" it to other people who dont understand how a rotary works.

The trailing sparkplug hole is already a small hole covered by the apex seal so there is minimal air/fuel loss as he proposes with the slot. Just the volume of the gap from sparkplug face to apex seal blows back into the intake stroke.

The larger leading sparkplug hole is carefully positioned so that the exhaust stroke pressure is the same as the compression pressure when the apex seal passes it, so there is no/minimal leakage.
The relationship between pressure of the compression stroke (amount of air) and the pressure of the exhaust stroke is constant because combustion pressure is set by the amount of air (compression pressure).

Ive posted the diagrams for this before from the book on the development of the Wankel engine.

"The Wankel Engine" Jan Norbye 1971

The ridiculous depression on the tips of the rotor instead of the middle. Again, he findamentally doesnt understand how the engime works.
The rotor tub is a transfer port, it has to be there or there wont be any air/gas at the leading tip of the rotor to push it around. Where the combustion force pushes on the rotor is a result of ignition timing and the air/fuel rushing through the transfer port in the rotor propogating the flame front.

Mazda has played with the shape to tailor when peak combustion pressure occurs (leading deep recess, trailing deep recess and what we have now middle deep recess).

I thought it sounded too good to be true... the title had a clickbait feel to it and the model the inventor was using was too primitive.

The second iteration of the 3 rotor chassis I've been building was running a somewhat custom Vortech/V-1 framed centrifugal setup at 18 PSI and it was pretty 'meh' in my opinion. On paper it seems neat but driving it exacerbated all the negatives that get thrown at rotary power bands and didn't give access to the torque we were looking for on corner exit to keep up with big v8s and v10s. I'll admit though I never liked the idea and may not have given it a fair shot with transmission/gearing development to keep it on power the compressor is also less than optimal for what we were doing but it was the only thing I knew of that was readily available even with the step up mods we had done to the gearbox.

On a side note I realized I wanted to be a bit more thorough with weighing some parts so I didn't get around to the whole shebang though I did quickly throw an REW e-shaft on a scale and then a 20b. On this scale (I didn't get any of the good gear out, this is a shipping scale so exact weights may be off, I was more interested in the differential) the REW shaft weighed in at 14 lbs and the 20b weighed in at 26, the 20b was wrapped in cardboard so lets say 25. ~78% increase in weight in the crank shaft alone. I'll bust out some Longacre scales and weigh more 20b vs REW stuff as soon as I can.

Yeah, for purely supercharging if you want low rpm torque it would have to be a positive displacement supercharger for low rpm boost and if you wamt it to make any power it would have to be huuuge.
Might as well just strap a other rotary on top of your rotary to make the power.

Centrifugal supercharger has the same compressor surge line issue as a turbo, so you cant just gear it to spin fast at low engine rpm for boost- will surge.

Combined with a tirbo for top end boost a smaller (still big by piston standards) positive displacement supercharger or a small centrifugal geared for low rpm boost would work.

But you thought Mazda's sequential turbo system was complex? Now you've got all that and a mechanical drive for the SC to maintain as well.

Hell no, might as well go with custom larger sequential turbos (not stock hybrid) since the ecus/programming are there to control it and its well documented how to run and maintain a sequential turbo system.

Pretty sure that two turbines at 0.83 AR are not equivalent to 1.66 AR. Or at least that’s not making sense to me from a flow potential perspective, but perhaps I’m not looking at it clearly. I looked around and couldn’t find a clear answer.

I think that's just what Colin did. Turbo down low and supercharger up high. He had it running, but I am not sure where he went with it.("twincharged" is his handle here?)

The AR is a RATIO used as a unit of measure calculated from the Area of the cross section of a turbos exhaust passage divided by the Radius that cross section is from the center of the exhaust wheel.

So, its a constant ratio.
Twin turbos double the Area cross section while maintaining the same radius to the center of an exhaust wheel.

Double the Area per (same) Radius. Double the AR.

The 3rd gen Yamaguchi book spells this out as well.

Easy way to think of it.
Undivided 1.00 AR T4 exhaust housing.
Divided 1.00 AR T4 exhaust housing.

See how each volute of the divided housing is half the total area of the 1.00 AR undivided volute.
Divided 1.00 AR exhaust housing is actually a 0.50 AR plus 0.50 AR exhaust housing.

An even easier way to think of it-
Look at the Borg Warner exhaust side flow chart. Each exhaust housing can flow that much exhaust. Two of the same exhaust housings combined flow is twice as much as a single housing.

Stupid question; do both volutes of a divided housing flow the same?

Not a stupid question.
Both volutes of a divided turbo exhaust housing are designed to flow the same with the exhaust wheels they were designed around.
This means they usually arent perfectly symeyrical where the scroll slot meets the exhaust wheel.

Some divided (dual entry) housings are not divided all the way to the exhaust wheel, but instead collect right before it to avoid the difficulty of even flow.

Then you have dual AR turbos like S4 TII thay have one tighter AR meant to be open all the time and a looser AR meant to open at later rpms.
They are not meant to be run on a divided manifold.

Well I looked at some non-rotary single vs twin comparisons and they don’t work out that way. You have to remember that each turbine on a twin setup is smaller than a single divided turbine. So that’s not a direct correlation.

Try this perspective instead; you’re not going to get 375-400 whp out of each 63mm or 70mm turbine with a 0.50AR on a rotary application, lol. Because what we’re talking about here is a 2-rotor with twin turbos making 750-800 whp. I’m thinking a 1.06AR isn’t even big enough with a 63mm turbine. You can probably make it with the 70mm turbine as you directly experienced.

One other point, the EFR 7163 is not favorable to high boost at low rpm. If you lay the S252SXE (7070) map over it you’ll see that getting up to 30+psi boost needed to generate the needed power level, that the S252 is going to handle it a lot easier where the 7163 is going to be way out in surge early on. Which I was already evaluating both for another application and came to that realization early on. The largest AR for the 7163 is currently only 0.85 too. Which I already have a complete new EFR 7163 setup and was a bit bummed. When originally purchased I was only going for 20 psi, which keeps it out of low rpm surge.

However, after looking into those non-rotary comparisons it’s pretty clear that with modern technology today that a single big turbo is the way to go. It will be way more cost and hassle to built a twin setup without any advantage, just as other people already stated earlier.
.

Had a friend YEARS ago that had one of the very few twin charged 1st gen MR2's in the US.

The 1st gen MR2 was available with a supercharged 4-cylinder motor. HKS developed a kit that added a turbo onto it with a whole lot of plumbing. The supercharger had a magnetic clutch like an AC compressor so it could come on at low RPM and spin the turbo up and then the SC would disengage and let the turbo take over for high RPM. It actually worked pretty well except for the skinny tires on the MR2, it was just a smoke show.

It's obviously a plumbing mess and adding a lot of weight and complexity.

Dale

That's very interesting. Basically the same thing Colin was doing. Just too complex.

I drove a supercharged 1st gen MR2. It was great! (The later ones grew too big and too heavy for my tastes.)

Yeah. That's what my twincharged setup does. Turbo blows through the positive displacement supercharger. It has a clutch (like an AC compressor), so it turns off when the turbo overpowers it.

It's still on the car, never tuned as I needed to wire in a new ECU (old electromotive was ok for its time but firmware needed to be upgraded and I wasn't going to spend any more on it) and then life happened haha.

Even without much tuning, the supercharger worked awesome. I don't have dyno numbers but it gave it immediate low end grunt (better than stock sequential twins (I've driven a fairly stock FD with bolt ons making about 300-325whp).

I never finished the tune for high end power and ironed out the crossover of the supercharger disengaging. So no real data on how well the overall system works from beginning to end.

I think you are exactly right that trying to make 400rwhp on 7163 with 0.83AR is going to be like trying to make 800rwhp on two 7163s each with 0.83AR.

Could be done, but few people would lean on a set-up that hard for 400rwhp when there are much more reliable ways to get 400rwhp.
People do decide leaning on a set-up that hard is worth it for 800rwhp. Not many ways to get 800rwhp on a 2 rotor.

We know from Indycar that they considered twin 6758 equal to one 9180 in power. I believe they switched 0.60AR to 0.85AR depending on course layout on the twin 6758s.

Indy car considered twin 7163 with 0.85AR an improvement in power over twin 6758 and single 9180 with a slight reduction in response over the 6758s with the 0.60AR

That is why I like the idea of using both exhaust sides in parallel and using the compressors in series at low rpm for the great surgeline and then parallel for high rpm.

Application would be relatively high boost at low rpm while keeping each turbo operating at low boost where they are efficient (pressure ratio is multiplied through serial compressors) and then lower boost at high rpm when in parallel mode.

Bridge port to spin the turbos at low rpm.
Intercooler per turbo so 1st stage is intercooled before going through 2nd stage when in serial compressor mode.

I think it would have a broad usable powerband.

Shoot for over 400rwhp and under 800rwhp

Streetport and 8374 is probably a way simpler and better idea though if 500rwhp is enough. Especially if Turblown gets their variable geometey housing working where ot switches from one volute at 0.60AR to both at 1.20AR.