A new thought for twins - Compound Turbos? Maybe?
 

So here's the deal, Most of you that have read my sig know what I've accomplished with a set of stock REW turbos.... 405 RWHP with all the squential hardware intact. I'll be on the dyno in a few days time to dial in the sequentials and get a base dyno for what I want my next project to be.

I recently acquired an FD that I have yet to pick up actually (BIG Thanks to Rico and Tray for hooking me up BIG) and the wheels have already started turning. I plan on using basically the same engine that I have now. A ported, polished, portmatched 13B-RE with various RX8 internals, stat gears, e-shaft mainly and the cosmo 9:1 rotors.

What I want to do is simple, beat the twins that I have now. This is easy in top end performance, but really lacks in the low end. This is where I need the most help. This is a street/auto-x/track day car. I drive my FC every nice day that I can and it honestly is a blast to drive. Non-sequentially it lacks but when I run them sequentially, she's an entirely different car. God willing I'll have the dyno sheets Tuesday and I'll post them up for sure so we (NoDOCH) can start doing some math ;)

This setup will be in an FD and I would like to see 450-500whp and a fairly flat torque curve. Reasons for the goal, FD has better suspension, can fit wider tires, she's heavier, and frankly, I want to outdo myself.

I've been doing alot of research on the evil forum, when IB decides to let it work :banghead: & I've seen a few things that jump out at me. Howard Colemans twin setup, Jason @ RX7Stores setup, A guy named Marcel Burnett (that used way oversized turbo's) and that's really it. Howards setup is everything that it claims to be. It makes 507 RWHP on pump/AI, there isvirtually nill as far as EMAP (something the stocktwins cant boast+) the turbo's aren't working all that hard, and the entire system sees pretty reliable for what it is. What I don't like about Howards setup is the fact that if you look at the dyn chart, it looks like every other single turbo chart out there. This is not what I want to see.

Jasons setup, well, it was never really finished. I thought it had potential based on the ONLY dyno sheet that was posted. It seemed like a good idea, good turbo choice (same one I was looking at for the most part, there are a ton of variants in the GT28 family) but the packaging was just aweful in my eyes. It looked rushed to me. There was little thought put into the intake plumbing to the turbo's, or from the turbo's to the intercooler. The exhaust mani was not equal length, or even close to it for that matter. I give him credit for going against the grain and trying it out, I understand he is trying to run a business, but at the same time I believe that system left alot on the table. I'm not going to let that "failure" discourage me.

Then I saw something interesting in a diesel application - compound turbo's. Aurura Turbo has a few kits' out for the diesels that are rather interesting. Basically there are two turbos of different size. The smaller turbo pulls it's air from the compressor housing of the larger turbo. Think about that as one of my questions is directly related to it. The exhaust is first routed through the smaller turbo and then to the larger turbo. All the exhaust passes through the smaller turbo before it enters the larger turbo.

This seems like a great idea. The way I see it, the smaller turbo is compressing a small amount of vaccuumed air becuase it has to pull through the larger turbo until it spools ever so slightly. However, once that larger turbo is spooling and creating positive pressure.... is that smaller turbo going to create resistance? Or is it going to further compress the charge air?

Looking at the dyno results of this system is impressive and appealing. It is EXACTELY what I'm looking for. Torque builds fast, and stays constant. Now, is this a function of the turbo's? Or the unique characteristics of the diesel engine?

I would like input on this. I am dead set on upping the ante with my FD build in terms of power. I would like to see a sequential looking torque curve that builds quickly to 360ish and holds there throughout the RPM band.

There will be no factory style exhaust mani's. I believe that is the limitation of the FDs turbo's. It's really a mess in there and it doesn't matter if I'm making 10 lbs or 14+ lbs, the EMAP is pretty much the same at around 280-290 kpa at redline. There is a direct almost perfect 1:1 ratio as boost builds. Once boost target is met though, the EMAP continues to climb. There is just too much exhaust flow and it can't effeiciently be routed out. Anyone that has seen the stock mani will understand why. In Daves words, Yeah man, it's a fucking mess in there.

So ideas and thoughts, lets here them. This will take a year if not more to come to fruition, and I don't expect any hard data until 2011 @ the earliest but mark my words, something will happen.

Interesting thought... I guess I'm trying to picture the diesel turbo set up you are talking about.. I'm not quite picturing how it works. Maybe a diagram of some sort??

A couple pics, good call Phil. I should of put these up immediately

http://www.atsdiesel.com/ats_new/ima...2272202935.gif

http://www.atsdiesel.com/ats_new/ima...2272202942.gif

Very vague for a reason I suppose. I like the theory though.

Hmm.. Interesting..

I wish there were more real data on these things.. What I would like to think about is, would smaller turbo getting air from the larger one eventually hit limit of one or the other. Example, the smaller one will spool first, than exhaust goes the the second, larger turbo, but the air is being pulled from the second.. Would the air eventually hit a limit as rate of pull would be different?? Also, how about exhaust rate after going thru first than the second?

Also, how about air temp?? Air seems to be warmed up already and then warmed up again by second turbo? To me, it might be better if two merges vs. one shoots warm air to the next.. Also, I've heard air temp isn't as crucial to diesel vs. rotary.. Again, thats what I've heard and I have no real experience in diesel.

again, i'm not an expert in this.. just putting down what I'm thinking in my head right now.

i have similar suspicions

I just read the description a bit more..

So, it looks like you would have to 'balance' the two turbos (sizes). Which sounds like you need to know what kind of power/response you want and set one turbo to match the other. I don't know, it sounds interesting but I feel like you could achieve this by going with a right single turbo.

Spew all the thoughts that you want in here, that's the purpose of the thread.

Diesel AIT's vs Rotary AIT's - not all that critical in diesel applications as diesels compress the air to superheated status THEN inject the fuel causing combustion. Typical diesel comp ratio's are 20+:1 and boost is in the 40+ range....... yeah. We all know what a poor intercooler can do for a rotary.

As for the function of them..... This is what I'm wondering as well. My friend was thinking about going twin charged - and infact there are some Yanmar diesels (I'm a marine guy too) that use twin charging. The way the Yanmar system is setup to work is with a supercharger and a BIG turbocharger. The supercharger gives the instant response and is run off a clutch system. When the ECU detects mani pressure above what the supercharger can produce, the supercharger is shut down and the turbo does the rest. Interesting.

What I'm gathering from these compound turbo's is the smaller gives the instant reponse while the larger is spooling up. This of course hurts backpressure which is a major hindrance for a rotary. But how much? Dunno until it's tested. For YEARS internet myths surrounded the twins saying they were a backpressure nightmare. In reality they are no worse than MOST other turbo's. I stay with in the generally accepted 2:1 ratio but I believe I am right at the limit of the effeincy of the stock hitachis. I would love to see what the BNR's could do BUT it's not all that much of a challenge.

After my ramblings....... the compound system, it seems like the smaller provides the instant repsonse that I'm looking for while the larger is gaining momentum. But when the larger somes online is the positive pressure from the larger being further compressed by the smaller turbo? I would say..... yes, yes it is. So then air temps as Phil pointed out can come into play. Such to the point that a LARGE Air-to-Water intercooler are needed. OK, not a problem.

Maybe the thing to do is not plumb the outlets of the turbo's in series, but instead use some sort of boost activated exhaust cut-out instead? Size the turbo's such that as one is out of breath, the other is ramping up. But then will that throw the smaller into surge? Is that why they are plumbed the way they are in the first photo?

How about Variable Vane Turbos?? Has anyone use these things on Rotaries? Those I worry about reliability aspects (especially high exhaust temp of rotary).

I don't see it happening in single form unfortunately. I'll be calling Garrett soon enough but I want the response of a GT25/28 with the top end of a GT35/37. These are two disparate goals and I don't see one turbo accomplishing both. I'll have to get this sequential dyno chart up and compare it to the myriad of GT35R dynos that are floating around.

My main concern is the area under the torque curve under 4.5-5k rpms. This is where I want the difference to be. This is where the sequentials really shine for a street/auto-x car. I SO wish Dave and I had more time last saturday right before the last auto-x on sunday :(

the VVT have some a long way, its mostly just cost.

You're 110% right on the money Phil. The combination of high heat and sutty exhaust create havoc for those turbo's. It's a great idea for sure, but reliabilty is a factor. I do plan on talking to Garrett about those as well as I have seen them and actually driven a tank of a boat, a 25' bertram with twin 4cyl Volvo's. I was on the test drive with the mechanic, I was at the helm and he said punch it. I did. I was VERY impressed with the response and torque they generated. When we got back I talked to him about them a little, got to see it first hand, it was actually the turbo that got replaced :rofl: and we chatted a little about them. Great idea, but I'm unsure if they could tolerate the rotary's heat.

One properly sized VVT is probably not 2x the cost of any other GT family turbo though, although I could be wrong. Definately worth looking into though.

The point of a compound turbo system like that is that it provides very -very- high boost levels. Let's take an example:

The larger turbo runs at say a 1.5:1 pressure ratio, consuming something on the order of 1200 CFM of air, compressing it down to something on order of two thirds that (somewhat more than that, given heat pollution, but this is example math, so don't crucify me.). So we have 800 CFM at the outlet at 1.5 Bar pressure.

This feeds into the inlet of the smaller charger, which runs at say, 2.5:1 pressure ratio. The turbo doesn't care what pressure the inlet is at, it just compresses it further. So our charge gets compressed to ~320CFM at 3.75 bar, or around 55PSI absolute; think about 40 psi of boost.

It's not really a device to increase response like a twin setup, it's a device to get around the boost/airflow limits of turbochargers. The larger turbo flows enough air for the entire system at a reasonably low boost level, while the smaller turbo, instead of having to worry about massive airflow numbers (The numbers I used are probably par for 1/3 to 1/2 throttle in a big diesel application at that boost level) can compress a smaller charge further.

Heat buildup, however, is pretty intense, since unless you absolutely match the turbos with regard to efficiency ranges, they'll both be running outside of their efficiency ranges most of the time. Even if they are well matched, you're getting heat from two turbochargers in all of the intake air. Hence why they generally run massive intercoolers in diesels.

In short, if your goal is to run > 30psi or so of boost, this might be something to look into, and if you want to run more than about 45, it's almost a necessity, but otherwise, it doesn't do what you're looking for.

Ive seen a number of trucks in here with compound setups, and always had wet dreams about one on a rotary.

I know there's a kit for supra's using a compound plumbing setup, and its laying down RIDICULOUS spool and insane top end on the 2JZ. Search around the net, can't recall the name of the kit ATM.

Personally, I think taking the general idea and adjusting it to suit the rotary's strengths is the best approach. I would personally like to see the small turbo fed first, but reversing the direction of the intake air. Have the large turbo suck through the smaller one. Thoughts behind that being: the smaller turbo would have -essentially- vaccum on the compressor which is going to help it gain shaft speed very early. These babies like the air volume as we all know :) And once that snowball effect starts to roll on as the exhaust pumps out more and more pressure and volume - well we're off the races :)

I do feel though that cooling will become a larger issue with this setup. We know its a problem that you have manage well on these things already, and 2X the exhaust plumbing, 2X the compressors etc is going to create a larger easy bake oven then a single turbo. It may be worth considering a small cooler between the first and second turbo (which ever you mount first) like one of those tubular style babies. Just something to add some extra density and take some load off the main intercooler.

I love the idea, and think that with a well tohught out setup, we can see a true example of what Mazda may have designed if the intended the FD to be 550whp from the show room :)

If there was ever a man I would have hoped to tackle this task - its you Brian. If you head down this road, I know its going to be all it can be.

The big turbo sucking through a small turbo is the same as a big turbo sucking through an inch and a half inlet; the big turbo is going to pull vacuum on it as soon as it starts to spool, and then the smaller turbo is just a restriction.

What you'd want is a sequential system with a reasonably large turbo for the top end, and a small turbo to create boost down low; that's what I'm working on with my sequential controller. It uses electronic controlled valves to control the secondary turbo and the wastegate, and allows a small turbo to boost to say, 14 PSI almost immediately, and allows the big turbo to make as much boost as you care to let it when it spools. And since the wastegate is electronic, it doesn't even start to crack until the second turbo is at full boost.

Boost Logic has been working on a kit for the supras, you should search supraforums and see what they have done or give them a call. They probably have the most knowledge of a compound kit on a gasoline engine.

This right here is what I was unsure of, thank you for the clarification. It was my believe that if the smaller turbo is seeing say 2bar Absol, and running at a 2:1 ratio, that it would be producing 3 bar of BOOST. While it makes perfect sense in theory, I was unsure if the turbo would physically be able to do it. I'm sure it's not simple absolute math as ineffeciancies do exist.

In your opinion, is this something that COULD be configured to provide the response that I'm looking for? My main concern here is the response. It's pretty common to see a 450-500rwhp rotary. There are a number of basically bolt-on kits, turbo and mani, that allow for this. None of them however, address the response that I so dearly love in my FC. I understand this is an outside the box kind of thinking as the general rule is you have to give up response for power and vice versa, but Mazda figured it out almost 2 decades ago. Granted they have engineers and foundarys that I don't, but still, anything's possible.

Heat I can deal with. It's a matter of fabrication as far as I'm concerned. Not to sound arrogant but my FC on the dyno on a 65* day can run 4-5 4th gear dyno pulls in 20 minutes and the coolant will stay at or BELOW 185*, oild ~170* and AIT's will start ~60* and end BELOW 80*F. If need be a small inline Water-to-Air could be utilized to lower the charge temp of the smaller turbo's intake back to about ambient before the smaller turbo compresses it, reheating it, then it's just a normal intercooler after that.

My goal is 360-375 ish torque with it being as early and as flat as possible. A PSI goal hasn't entered my mind yet. I would obviously like it to be as low as possible as I would like to avoid AIT related problems and Meth or water is something that I don't want to run either. Although I think it might come down to two retardedly small for a rotary turbos (GT25ish size) and some meth to keep it under control, although that defeats the really plush easy daily driveably charateristics that I want to keep with this car. I love driving my FC around, but it has become a little raw.

Thanks Joe, I'll check it out.

I'd also love to see this scaled for a rotary as well. The dyno charts that I've seen for the diesels look exactely like what I'm after. Instant and constant torque.

What I'm wondering is, could we have the smaller turbo feeding the the larger turbo, both intake and exhaust. What would the larger turbo do when it sees compressed air and spinning at low shaft speeds? Would it pose a restriction? Or just let it pass? If the smaller turbo is running a 2:1 and the larger is just starting to spin, would it allow the 2:1 to pass through, wold it start to spool up with a positive charge thus evacuating more exhaust? Or would it just act like a plug?

Heat can be managed, I think size constraints in the engine bay are going to be the limitations

How sick would that be?

Thanks Joe! Something will be done, and it will be done in an unconventional way. I just don't know what it is yet.

I've done some research on VVT's as well. Garrett has a GT37 designed for diesels with this technology. I need to call them soon and get some information. There's also a company callled Aerocharger that's offering some promise as well. I was VERY skeptical about this until I saw all the military applications for it. Still seems "gimicky" to me though.

Thanks Anthony, I'll do a little searching around.

Would you mind sharing your idea for this? This is something else that I'm considering. Without doing any math, the sizes that I'm looking at are GT25/28 for the small, and a GT3071-R for the large. My thoughts for this type of system are

Exhaust - an equal length mani, each with a wastegate and an internally gated turbo just to give the turbo the chance to vent out as much exhaust energy as possible. The two individual gates will plumb back into the exhaust AFTER the second larger turbo. The larger turbo is also internally gated.

Intake - First, I'm undecided if the intercooler will be a dual inlet or not. For simplicity and to give the first turbo the least amount of space to pressurize, I believe the turbo's will be merged as close to the first turbo as possible and an electonic cut-out will be utilized. So, the first turbo is plumbed strait to a merge collector. The second turbo is plumbed to the merge collector as well but inline is an electric cut-out to keep the first turbo from back-spinning the second. There also needs to be a vent to allow the turbo to spool without creating a TREMENDOUS amount of exhaust backpressure.

Control.... well I'll be using a Motec M800 so the control's are virtually limitless. I'm thinking there needs to be a 3d map setup such that when MAP = the charge pipe for the seconday turbo, the valve opens and the vent closes. Easily accomplishable. Two different WG controls but the closed loop being on the larger turbos internal gate and the WG on the mani's being run off duty cylce at a preset limit to control "absolute boost"

Thoughts?

It's all so clear now
http://forums.hybridz.org/showthread...=145379&page=2

http://i522.photobucket.com/albums/w...d/SL380903.jpg
http://i522.photobucket.com/albums/w...d/SL380904.jpg
http://i522.photobucket.com/albums/w...d/SL380906.jpg
http://i522.photobucket.com/albums/w...d/SL380907.jpg
http://i522.photobucket.com/albums/w...d/SL380909.jpg
http://i522.photobucket.com/albums/w..._month_049.jpg
http://i522.photobucket.com/albums/w..._month_084.jpg
http://i522.photobucket.com/albums/w..._month_145.jpg

Well after seeing the above design, the exhaust side makes alot of sense, but I'm still having trouble understanding why the air from the larger turbo has to pass through the smaller one, wouldn't the smaller compressor become a bottleneck at a certain level? What would happen if you ran a manifold like the one above, but had the compressor sides set up like normal, eventually merging before the intercooler?

For the rotary I think that makes the most sense HOWEVER - if you're going to do that than you need compressors that flow the same. Essentially making a sequential twin turbo. If one turbo can out pressurize another, it will most likely backfeed and overun the smaller.

The point of having the larger one feed into the smaller is to further compress the intake charge. The smaller one will compress whatever it is being fed. Whether it be atmospheric pressure, or 15lbs of boost, it will further compress it.

I'm coming up with a few ideas for a sequential twin turbo idea. I'm hoping Rotary Prophet shares some of his as well.

Compound charging is one of those things that looks like it wouldn't work right, but apparently with all the success in diesel engines, and now Supras and other cars it apparently works quite well. The easiest method is to use at least an internal wastegate on the first turbo (smaller one). And for a rotary you'll want to make sure the wastegate and housing can flow enough so as not to become a bottleneck.

The EMAP is what really concerns me. So many things can and do go wrong when this is left unchecked.

A few thoughts from yesteryear:

Believe it or not, 20 years ago, a lot of people were going the staged route. This was before turbos could make more than about 2:1 pressure ratio. The trick to success on a staged setup was (and still is) dual intercoolers.

The boost from the first turbo is cooled before it passes through the second turbo, and then it is recooled after the second turbo. This allows for reasonable charge air temperatures, while minimizing lag.

On a diesel (as has been noted) the intake air temperature is not an issue for engine longevity (except for possibly the valves). It is an issue for charge density and thermodynamic efficiency. This (and new NOX emissions standards) is why almost all modern diesel engines have intercoolers.

I have seen many diesel engines with staged turbos and a gaseous-fueled engines with staged turbos (good detonation resistance).

Benefits of staged turbos:
Quicker spool of smaller turbo
Insane power potential from larger turbo
Cooler charge air with very little if any more boost lag due to intercooler volume.
Very high boost potential (120 psi + boost).

Drawbacks of staged turbos:
Underhood piping disaster
Increased EGT from engine
Increased EMAP from engine
Higher cost
More potential failure points.
Insane underhood heat problems

Here is the reason that the turbos should be connected in the standard staging configuration:
Turbo Flow Maps are measured in inlet volume (or mass) per time. (Typically CFM, lbs/min, etc.) This means that at 2:1 pressure ratio on the primary turbo, with a well cooled charge, the secondary turbo will flow twice as much as it is rated on the curve, as the air is twice as dense going into it. This means that the secondary turbo needs to be 1/2 the flow rate of the large turbo (for this example).

Actually, the density ratio is what actually determines the turbocharger flow rate ratio. a 1.5:1 density ratio across the primary turbo means that the secondary turbo should be 2/3 the size of the primary turbo.

On the exhaust side, the air is very hot coming out of the engine. It is also under very high pressure. The exhaust turbine on the turbo charger is rated in pressure ratio. This pressure ratio across the turbine varies with air flow rate.

High pressure air from the engine does not occupy as much volume as lower pressure air coming out of the secondary turbo, thus the flow rate through the large turbine is a factor of the exhaust gas density ratio across the secondary turbine (the temperature will decrease some). Thus the output flow rate of the secondary turbine is significantly larger than the input flow rate (in volumetric units, obviously the mass flow rate is constant, due to the conservation of Mass limitations in our non-nuclear internal combustion engine). The larger turbine requires a higher flow rate, and must therefore come after the secondary turbine on the exhaust stream.

If the large turbo was first, it would spool very slowly, while the small turbine would quickly overspeed.

Certainly. My initial design is something like this (and keep in mind, this is a budget buildup to even see if it's plausible, so the parts combination is all whatever I could find in a junkyard; I'm currently using a T3 turbo off of some 4 cylinder, and a Masterpower 70mm T4.

The exhaust manifold is based on a standard, short, fat runner design that merges into a T4 divided flange, in about the same position as a megan racing 13b manifold. About halfway between the engine and the T4 flange, we run a second, somewhat smaller pipe off of each runner, and route it to a T3 flange, in a position to make mounting as easy as is possible in a setup like this. At roughly the same position that the T3 splits off from the main runners, we split on the other side, route the pipes out toward the exhaust, and mount the electronic wastegate. I'll get back to that.

So basically, each runner of the exhaust goes three places; a wastegate, a small T3 turbo, and to a T4 turbo, which you'll note has exhaust priority. In my setup, the wastegate pipes come off of the top of the runners, between the exhaust and the intake, and run down towards the downpipes. The T3 runners go from underneath, and are routed up towards the front of the car, to mount the turbo somewhat in front of the T4, and higher up. It's also worth noting that neither turbocharger is internally wastegated.


The only thing of note in the intake system is a pair of valves on the turbocharger outlets. Basically, you use a reed valve on each one to prevent back flowing through either turbo. This could happen through the T4 when the T3 is doing it's initial spool up, or from the T3 if you choose to run more boost from the T4 than the T3 is capable of pushing.

As for the control, I'm an embedded systems programmer by trade, so I chose to build my own microchip driven control box. In it's current form, it's dead simple, with just two stepper motor outputs to control the two exhaust valves (we'll get back to those), and two MAP sensors, one connected to the outlet of each turbocharger. The capacity is there to input RPM, throttle position, gear, or whatever else, and to map the boost just like a full fledged boost controller.

In the down pipe for the T4 turbocharger, there's an electronic valve. It's basically a throttle body that's driven by a stepper motor, and designed to deal with the kind of heat that we're going to see in a rotary exhaust. The part I'm working with now was designed to route jet engine exhaust, actually. This valve, when closed, prevents any exhaust from flowing through the T4's exhaust turbine by blocking the outlet. This is preferable to blocking the inlet, as the temperatures are not nearly so extreme as in the manifold.

There's another valve located where the wastegate would otherwise be, and functioning as it.

To help you visualize how all this comes together, here's a play by play.

You floor the engine at low RPM. The control box reads vacuum on both turbo outlets, and closes both the turbocharger select valve, and the wastegate valve fully. This routes all exhaust through the smaller turbo. Once the primary turbo shows that it's reached it's pre-set full boost limit, the system begins opening the turbocharger select valve. The system will gradually open this, working much like a wastegate, releasing any exhaust pressure not required to run the T3 at capacity through the T4 turbocharger, which gradually brings it up to speed.

Once the MAP sensor for the T4 charger reads that -it- is at full boost (some time before the turbo select valve is fully open, generally), the system begins opening the wastegate valve in the same way, metering it exactly to keep the boost solidly where you want it. It also fully opens the turbocharger select valve, if it wasn't fully open yet.

Anyone familiar with how a wastegate works knows that as a control device, they do a terrible job, as they waste energy by opening before the turbocharger reaches full boost; by using an electronic valve for a wastegate (which has been done before, but not very often on road cars, usually due to cost), we keep the wastegate closed until it's absolutely necessary to open it to prevent over boosting. By doing this, we pick up significant torque during the engine spooling phase.

With two properly sized turbochargers, there's no reason you couldn't have the rock solid straight-line torque numbers you're looking for with a system like this. A small T3 can spool before 2000 RPM on a 13b, but I would investigate a slightly larger one that spools by, say 2500 RPM if I were planning to run high boost numbers with a relatively large secondary turbo. This would help keep the transition between the small and large turbos smooth and unnoticeable from the driver's perspective.

In a way, the smaller turbo selection is more important, because the goal of the entire system is to run the small turbo into it's peak efficiency range, and then keep it there by gradually moving work to the larger turbo. This will help the low end torque numbers by keeping heat out of the incoming air charge. Then you just need to select a secondary turbo that, when coupled with the primary, will leave both of them in their peak efficiency at about 75-85% maximum revs.

I know I've already got a book here, but it's worth looking at pros and cons versus a single turbo setup.

As far as pros go, you've got the very quick spool up of a small single turbo, the high end efficiency and power of a large single turbo, and a very smooth power band and seemless transition. Operationally, the sequential setup takes it every single time.

The list of cons is entirely logistical. It requires two turbochargers, a custom manifold (although, I'm hoping I can start producing these once I get it up on the engine dyno and start tuning, and can put out some numbers of my kit vs a single turbo vs stock twins.), some not inexpensive valve hardware, an electronic control box, and let's face it: a f*cking plumbing nightmare under the hood.

Sounds like a VERY intreresting project! I like the thought that went into it. Unfortuneately for me, I only hav the option of running one stepper motor with the Motec that I'll be using. I was going to use it for the OMP but I like the throttle body butterfly for the intake idea more. I suppose I could use an internal style wastegate actuator connected to a butterfly valve as well.

My main question would be, how would the turbo's handle working against each other? The larger T4 turbo in your example would be moving more CFM at the same boost pressure as the T3 style. Pressure is pressure and if held constant between the turbos, would the T4 over power the T3 in anyway?

The setup I had in mind is similar. Here's the setup that I'm contemplating

Two equally sized turbo's as far as compressor goes. Possibly changing the A/R on the secondary turbo to reduce the inherent backpressure that would be created between the two turbo's. Basically what I'm contemplating is an equal length mani with a T3 flange and the largest WG possible right under the flange. The primary turbo would get the full brunt of the rotary exhaust. This primary turbo MAY be gated on the snail, and if so, that WG will run to the DP, I don't think the will be necessary though. Routing from the turbines exit of the first turbo is plumbed directly to the turbine inlet of the secondary turbo. The first WG is also plumbed back into that same exhaust stream allowing the exhaust energy to be split. There is a second WG below the turbine of the secondary turbo as well. This WG is vented directly to the actual DP. That's the exhaust side of things.

The intake side is rather simple as well. THe primary turbo is plumbed directly to the intake track. THe secondary turbo will have a vent valve placed between the compressor discharge and a TB style butterfly valve to keep the two turbo's seperate.

Play by play -

Theory being that when you floor the go pedal, the smaller turbo gets the full power and flow of the exhaust and such will spool VERY quickly. The WG associated with that turbo keeps the primary turbo;s boost in check. The secondary WG will COULD be open at this point to reduce the backpressure between the two turbo's to allow for a bit faster spool in the lower RPM's. (If I need to build seperate vac and boost storage tanks to aid in this, so be it. Easy enough to do, They're on my setup now actually) As RPM's and EMAP increase prior to the first turbo, the primary WG will open to keep the primary's boost in check, when the prim WG is fully open, the secondary will be in charge of controlling boost in the entire system. Once the secondary turbo begins creating boost, the seconday WG will most likely begin to close to aid in the secondary creating boost. Once boost is realized in the secondary turbo, the vent will close, the butterfly valve will open and both will be running in parrallel. The point of the second WG on the snail of the first turbo would be to vent the exhaust energy from the first turbo if the difference in flow of the exhaust system was such that the primary was receiving more energy than the first.

In my Datalogs of the stock sequential system, the rear turbo will always run hotter by ~75-100*. I've switched signals, repeated runs and gotten the same result so I know that it is not a function of different sensors, wiring, or signal amps. My theory is that the exhaust flapper in the rear of the housing creates enough resistance to increase the backpressure there, keeping the gas hotter. There is a direct correlation with the seperation of EGT's and EMAP.

The system that I am proposing is very similar to the stock sequential system. In my eyes, the drawback to the stock system is two fold. The inability of the hitachi turbo's to tolerate more than 15-16 lb's of boost reliably. More important though is the exhaus mani itself is a convoluted mess. A tubular mani in my eyes would yeild significant gains.

The plumbing would be a bit of a chore. However, I believe that after scouring the goodridge catalog, they offer enough fittings to allow for a somewhat neat packaging. I would run one -3 line to the top of the turbos and hope that I could get the CHRA's in a banjo fitting and just use an on-the-run type fitting to feed the second turbo. Drains are pretty simply if you're using an RE or an REW block, there is a drain for and aft. I would use the outlet/inlet of the WP housing to cool the first turbo. Perhaps use the nipple on the rear housing to cool the rear turbo and just weld a second -6 bung on the waterpump inlet. Although I do plan on using an REW WP housing, I plan to take the T-stat housing and outlet off and build my own to suit my needs. In that case the AST would be contained there and I would just have a ton of fittings welded on to it. I think the exhaust plumbing would be much more of a headache than oil/coolant plumbing it thats what you guys are reffering to with the plumbing nightmare comments. Hardlines would be awesome to have here as well and are very easily made, even when using banjo fittings.

I do love this snail though

http://www.atpturbo.com/mm5/graphics...cial%20t25.jpg

could also do a twincharge setup with a roots blower and a large turbo. The roots blower would add power down low, any extra power downlow = more exhaust pulse and faster spool of the larger turbo. The large turbo would allow the engine more power potential.

The beauty of my setup is that the control box for the turbos is a separate box, and doesn't rely on the stock ECU; you just input your max pressures and allow it to go about it's business.

No, as long as the pressures are equal, both turbo outlets will flow freely; if one produces more pressure than the other, the reed valve on the outlet will close, preventing backflow.

If you were using a system with more secondary turbo boost than the primary, it would probably be worth using another valve to redirect air from the primary turbo outlet into the turbo inlets, basically making a loop so that the turbo keeps spinning instead of stalling out against a closed valve.

Plumbing the turbos like that isn't likely to gain you anything. The thing to remember is that a turbo is a pressure differentially operated pump. In short, it relies on the difference in pressure between the inlet and the outlet on the hot side to spin the turbine.

So, you have to run your small turbocharger first, or else the exhaust, which can freely flow through the larger turbo, gets backed up against the smaller one, which means there's no pressure differential on the larger one. Because the smaller turbo must come first, the net effect is that once the exhaust runs through the small turbo, you have the difference in pressure between the primary turbo outlet and the secondary turbo outlet to extract useful energy from; far less than there exists in the manifold. Certainly there is energy there, but you'll find your larger turbo takes much longer to spool than it would in a single turbo configuration, which necessitates a larger primary turbo, which raises your boost threshold.

Your concept would certainly work, and I believe your system of wastegates would work perfectly, too, I just don't know what the performance would be like. I would be curious to see it in action, though.

Now, running sequential twins is an option, but it requires two reasonably small turbos. I feel a small and then a somewhat larger turbo will give better top end performance, but again, I'd be very interested to run the dyno numbers on a twin sequential setup with properly sized turbos and a good manifold.

I don't know where you're at in the country, but if you're anywhere near Cincinnati, I've got my engine dyno at the shop setup to test and tune rotaries, and I'm just a bit of fabrication work away from testing my setup. I'll be testing on a stock RE motor, with stock intake and upgraded injectors, and a water to air intercooler for consistency. That way I can test several turbo setups and map them against each other in an apples to apples comparison. It'd always help to have another brain when the time comes to do it.

Very nice little system. I lack the electronic knowledge to be able to do this. I wouldn't be using the stock ECU though, I would be using a Motec which has enough inputs/outputs and software to allow me to do this.

I see. Pressure not flow. Even though one turbo is flowing more air, becuase the pressures are equal at the outlet, they work together and not against each other. Mazda did this with the 20B turbo's actually. One was a Hitachi HT-10 while the other was a Hitachi HT-15.

That's interesting. That's more of a compound sequential system though if I'm reading it properly. That would give very quick spool fom the smaller turbo, but then when the valve re-directs the smaller turbo's boost from the the intack to the larger turbo, boost would rize VERY quickly. Again, this is assuming we are vizualizing the same thing.

I think we're looking for two different goals which is why our systems are slightly different. I'm looking for a total of about 450RWHP and 360ish torque. Not all that much from two turbo's when singles are getting that fairly regularily now. I'm more interested in creating the flattest torque curve possible and extending it as far throughout the rpm's as possible. I BELIEVE you are looking for more power which is why you're using not only bigger turbo's, but slightly different sizes as well. Please correct me if I'm wrong.

Regarding my manifold setup - One of the things that concerns me, is as you put it, the smaller turbo acting as a plug in the system. Leaving the second turbo with much less energy to spool up to the same level as the first. I'm worried that the extra distance traveled, the less heat available, is going to lead to a secondary turbo that just cannot keep up with the first. The main reason for thinking about not only the very large traditional WG on the primary turbo's mani, but also an internal gate as well.

I agree that a different sized setup like the one you are building will net larger gains in the top end. With auto-x and track days and street driving being the primary role of this car, I'm more concerned with the low and midrange performance of the system. I believe that twin GT28's of some trim size will net 450 RWHP if not slightly higher. Jason's setup did 450 or there abouts, and the BNR turbo's are regularily getting close to this mark as well. It's my belief that the restriction in the BNR's lies in the factory manifold. As far as I know, I'm the only one on stock twins that can actually tell you what the EMAP is. Most people just look at it and say it's a nightmare. In reality it's not horrible. It's certainly not on par with a tubular mani and a large A/R turbine housing, but that's not an apples to apples comparision. I believe if we can get the EMAP down in the higher RPM's power and effiecency will increase. Negating the need for a larger pair of turbo's or a larger secondary turbo to get to say 450 rwhp. MOST people would argue that this is a waste of time for such a small power gain, but in terms of NET hp and the intended use, I certainly think it is.

I live in the SW corner of CT, about 1/2 hr from the city. Honestly though, I would love to come out, see the setup, see the results, and lend my brain to the project. It's probably about a 10 hr drive out which I'd be more than willing to make. I'll check some airfares as well. Please keep me in mind when the time comes, I'd love to see it in action and be a part of it.

In regards to the exhaust being choked on the smaller turbo...

What if you approach the manifold design from a EMAP standpoint, forgetting about spool at all for a second.

Example: Twinsturbo FD. Sure you've seen the crazy long, multi bend labyrinth manifold. Responsive no, but 1:1 pressure ratio - yes. Wastegate is a straight shot from the runners (although they curve all over the place) while the turbo is actually 90* to them. Aaron's manifold is similar, wastegates are a straight shot out of the port, while the turbo is getting its feed on after them..........again, low backpressure.

Perhaps the best layout for the exhaust would be to favour flow to largest then smallest "vents" as possible. Maybe wastegate(s), large turbo, then small turbo. Once everything's closed up at low RPM and you hammer the throttle, the physics of the smaller turbo vs. larger should itself spool one before the other, and hopefully plumbing in the order of largest to smallest will allow enough flow for the power needed up top without making EMAP unreal.

One would imagine, given the rotaries exhaust power, that the simple physics of the two turbos in the system would create decent spool for their size even if flow was directed entirely to the gates. As long as the gates are closed and some of the pressure building is pushing on the turbos, they'll spool while the flow favouring the gates should keep pressure as low as possible.

????

Actually, I was hoping to start producing the manifolds, packaged with the control box and the valves. Just pick your turbos, and have a shop make a downpipe.

I don't think we are. In the situation where that valve would open, the larger turbo is already creating -more- boost than the smaller; the only point here is to allow the smaller turbo to keep spinning and flowing air to somewhere, so that it doesn't stall and stop spinning; it's basically useful to keep the turbo spinning during gear changes and things like that, when you drop back down to the smaller turbo for a short period. Same concept as a blow-off valve; by letting it vent, the turbo doesn't slow down.


No, I think we're looking for basically the same thing. My goal would be to imitate the FD's stock boost curve, except with higher boost levels, and eliminating that transition dip. By bringing the secondary turbo online slowly instead of all at once, I think that's possible.

The theory of putting one turbine behind another, and having every bit of exhaust flow through both is a sound one... in theory. Once the exhaust leaves the turbo, there's still energy to be extracted, but less so. But here's what I see happening:

To control boost and turbine speed on the primary, you want to put a big wastegate on the manifold, and maybe internally gate the turbo, as well. The problem is that by venting all that pressure past the turbo into the inlet of the secondary, then at higher RPMs, where there's a lot of exhaust, you've eliminated the pressure differential between the inlet and the outlet of the primary; the pressure coming in is the same as the pressure going out, thanks to the big wastegate opening a valve between the two in an effort to keep boost under control. The wastegate will stay open instead of closing, though, because it's linked to system boost, not individual turbo boost. Eventually, because there's very little pressure differential to run it, inlet pressure from the second turbo will begin flowing out of the primary's inlet.


You can get a flat boost curve with a pair of smallish twins, or with a small turbo and a moderately big turbo; either way should work about as well. The small turbo should flow enough air to make up the difference while a bigger turbo spools, it's just a matter of it working alone for longer. As long as the turbos aren't too far apart, it doesn't matter.

The question is, why leave performance on the table when you can get the same results, get the same outstanding low end response and torque when you can -also- get a higher top end? My system would work either way, but it seems silly to leave power when it's there. At very least, you could use a larger turbo, tune for high boost, and use the control box to bring it down when it's unwanted.

And here's something else to consider; in a setup like this, the larger turbo spools much -much- faster, due to the simple fact that the engine is already running under boost while spooling it. A 13b under 15psi of boost is exhaling as much exhaust as a 5.0 liter NA boinger, but with a much more favorable exhaust arrangement, in terms of exhaust pulses, and manifold setup. The primary turbo, being spooled already, is only using at small chunk of that exhaust energy, and the rest is generally wasted via a properly named wastegate.


I'll let you know, and we'll try some stuff and see what happens.

That would make it VERY easy to sell, I like that idea. I generally don't like piggy-back systems, mainly for aux injection type situations it scares me. Your idea though controls everything related to ONE part of the system which I don't see a problem with.

I agree, especially in a situation where we're looking for the best response from a turbo, keeping as much energy in the system is critical. I also believe that the EMAP plays a huge role in the torque curve as well. So in a situation where the EMAP is on the rise, the torque will fall off. Eliminating a spike in EMAP by keeping the turbo spinning would further increase low end torque.

I think with standard valves and what not, it'll be difficult. That's the problem that we ran into on the dyno on Tuesday. We would either see a dip, or a surge. We were logging the boost that the secondary was creating in the space between the compressor outlet and the Charge Control Valve so we knew what the turbo was producing. I adjusted the RPM resoltion down to 50 rpm increments around the transition and it was still difficult to get a smooth transition. I think your REED valve idea is going to be the only way to bring the secondary online smoothly. Looking at the datalogs we were able to get the transition dip down to .15s. Its VERY slight, but it's still there. I'm still working on the transition on the street and it's definately getting better.

That's the reason that I'm thinking that two turbos of close to the same size would be a better choice than one small and one large unless the larger one is run first - to take advantage of the pressure differential. It's my belief that the smaller turbo would need less of a pressure differential to create some CFM to add to the larger turbo.

What I would like to try is a pair of very similarly sized turbo's collecting both runners exhaust right at the point where they would split off in a Y - one going to the primary turbo, the other to a LARGE WG. I would like to take advantage of the energy post WG-before being sucked up by a turbine. This would be routed directly to the bottom of the flange on the secondary.
Going back to your point of the lack of a pressure differential on the primary, I'm not sure if the exhaust of the primary should be routed around the secondaries turbine, or to it. I think there is response lost if it's routed around it. However, maybe the thing to do is use a slightly larger A/R on the primary to let it breath a little better.

I agree it's silly to leave performance when it's there. But I am more focused on a flat torque/boost curve than higher HP. This is just based on my experience with my FC. Over 300ft/lbs, the suspension/tires really can't handle it. I may feel differently when I get a set of A6's, or get on a track where my RA1's can get up to temp. But right now, the higher end performance is wasted becuase I simply can't put it down. Now with the ability to put a 10.5" wheel and bigger tires behind an FD, and given the better suspension geometry associated with the FD, I may feel differently. That's why my goals are 450RWHP, and 360ish ft/lbs.

Agreed, which is why I'm torn between routing the exhaust of the primary to the bumper, or to the secondary turbo

Please do. I'd love to get a first hand look at what's going on, what works and what doesn't

Just remember that the more restrictive turbine needs to come -first- in the exhaust stream; if you put the less restrictive one first, exhaust will back up against the secondary turbine, and you'll lose your pressure differential, and the first turbo won't spool.

That's why in the diesel compound systems, the smaller turbo always goes first, then the exhaust runs out to the secondary. I'm not sure exactly what would happen if you ran a pair of identical twins in a compound setup.

I don't know for sure if that's what you were saying or not, but I figured I'd clarify, just in case.

No, you're right, and it makes perfect sense.

I'm not thinking about a pair of twins in a compound setup, but more of a sequential.

Do you happen to have any diagrams or pictures of the setup that you're working on now that you'd like to share?

Not really. The pictures of what I've built so far aren't very good, and my actual real camera refuses to turn on. Suppose I'm going to need to replace it.

But I'll see what I can come up with as far as a basic diagram, and send it your way.

Great, thanks so much.

As for Camera's, I picked up a Nikon Coolpix L20 specifically for shop detail. @ $130, it really can't be beat. I even use it to take in-car track video's.

maybe I missed it but was it mentioned that in diesels with compound turbocharging a positive boost ratio, meaning more boost than backpressure, across the operating range is possible?

if the turbos are sized correctly, could this be done on a rotary?

I have something to contribute to this thread, but I don't want to come off as an obsolete fool/youngin'. Let me go back and read this novel of a thread before I post it.