How are you going to equalize exhaust flow between the turbos? At low boost exhaust from both rotors are going to the primary turbo. At a certain boost pressure a wastegate on the first turbo's runner dumps exhaust into the secondary turbo's pipe, which opens the butterfly valve. Now each rotor is flowing air into its own turbo, but the primary turbo's wastegate is still dumping exhaust back into the second runner. This means the second turbo will be subjected to much more exhaust flow.

 

If it has a large enough wastegate, that shouldn't matter. The wastegate on the second turbo should be able to bypass enough of the exhaust gas.

There is also always the option of external wastegates that can bybass more air.

Kenton

 

There's a flaw here everyone is missing. You need an RPM triggered solenoid because you need to know which turbo is in it's optimum flow.

example:
powering out of a corner from off-throttle at 5000rpm. The mechanical seq spins up the primary which can't flow anywhere near enough air so boost stumbles from not much to heaps when the 2ndary kicks in.

example2:
gear change: even at full throttle the primary will have to be spun up on every new gear.

You can run mech seq on an FD:
* rip off CRV - connect hose from where it was to pressure side of TCA
* rip off vac side of TCA
* take hose that connects to the press chamber and connect it to the press side of the CCV
* connect the vac side of the CCV to 4mm nipple next to the CRV nipple on the second turbo
* run precontrol it a t-piece bleed setup

I've run this on my car as a mechanical seq but it feels wierd. It's because it engages the primary from every off-boost situation which you dont want at high RPM, you just want the 2ndary.

Anyone know exactly how the supra system works ?

-pete

 

In the sequential setup that I am talking about, you are not going from one turbo to the other, you are going from one turbo to both.

The first one spools and builds boost quickly, and then the other one of equal size spools to help with airflow on the top end.

Kenton

 

I am probably going to make a sequential TT setup as soon as I get enough time or can justify pulling the perfectly strong 60-1 off of my car.

If any of you have ever driven a 3rd gen with working twin turbos then you have noticed how much bottom end torque a small turbo puts out (especially on 3rd gens with DP, ecu, etc). Disreguarding the switchover point, lag is almost non-existent on modded TT setups. According to my calculations you should be able to run 2 either mitsubishi Td05 14b's or 16g's (or mix of both) with good success (400hp) or even run 2 T3's off of something like a SAAB or SVO mustang. Anyways thats besides the point, you can run some PRETTY small turbos if any of you have seen the size of the 3rd gen setups and still have good response, I personally would choose the mitsu's cause they are CHEAP and plentiful and I have a 16g in my shed.

In talking about the structure of the seq TT system I agree that the 3rd gen setup should be copied. The setup running a 2nd turbo through an external wastegate wouldn't work because the flappers do'nt extend far enough to reduce the turbulence going through them, also it would KILL your exhaust pulses to the turbos thus increasing lag for the 2nd turbo. The reason the flapper between the 2 compressor discharges is needed is so the primary turbo can pressure up the system instead of blowing the charge straight out of the secondary turbo (path of least resistance). This flapper can be run straight from the pressure signal in between the valve and the compressor of the turbo so that when the secondary turbo reaches some amount (you can install MBC here to adjust) it will switch the second turbo online. The setup is correct in that you will need an adjustable summit rpm switch (59.00 I think) that can run a vaccuum switching valve (you have 4 on the stock spider those will work great) that can be run to a pressure actuated valve (wastegate diaphragm) to switch a flapper in the manifold to allow the second turbo to recieve exhaust and start spooling. If I wasn't at work right now I would make up a diagram much like the 2nd draft (correct BTW) with a more descriptive setup. The only thing really keeping from doing this is the time making the manifold and downpipes for the setup. Also I have pretty much eliminated water cooling to my turbos and would have to run all that piping again, not to mention oil piping geez now I remember why I didn't do this yet. We'll see guys.
post more later,
Patrick

 

I have another idea too that would make the setup a lot simpler. Instead of venting the wastegate from the first turbo into the second, why not just leave the exhaust control valve slightly open. This will start to spool up the second from idle on up, making for a smoother transfer.

This second idea will keep cost down because of less fabricated parts, and it would make for a cleaner engine bay. Some people like pipes everywhere though.

I also want it to be completely mechanically controlled. Instead of having RPM switches, just have the actuators controlled by vacuum/boost. This will also keep it cheaper and less complicated. I am only seeing two vacuum hoses that will be needed. One to give boost to the exhaust control valve, and one to give boost to the charge control valve.

Kenton

 

As some of you can tell, I am pretty interested in this subject.

A couple of these should suffice:



http://turbochargers.com/currentspec...H-16G-7CM2.htm

Kenton

 

Greg,
Its too much in my opinion. Im trying to work something out with border. Now that im on vacation
im going to go talk for 15%. Normal price is $6250 plus shipping. You can also get the turboless kit for $3750.
The turbos are Apexi AX turbines...

Jason
www.jt-imports.com

 

Nope kenton you're still missing the point. At high RPM the first turbo will spool quickly, yes, but it will overspin trying to make decent boost outside of it's flow range. It will not even open it's wastegate at high rpm because it just wont flow the boost.

...ask me how I know...

-pete

 

I see your point now. I must have just read your post wrong. Now I will just have to modify the idea a little.

I stated this once above too. If you take off the wastegate vent into the second turbo, since it wont matter anyway, and just leave the exhaust control valve open about 10 or 20%, this will let the second turbo start to spool with the first.

Once the first reaches boost, it will start to mechanicaly open the exhaust control valve more until it reaches 100%. Then the second one will be online with the first.

Does this sound better?

Kenton

 

hmmm 10 - 20% is equivalent to the mazda pre-control system Your progressive idea has another problem - are you going to open the secondary turbo outlet valve progressively as well? You'll just end up with backspin otherwise...

-pete

 

You know I am going to do it progressively. I was thinking about a flapper valve in the intake that would alow the second turbo to push against to open. It would be competeing with the pressure from the other turbo so that it wouldn't just fly open at the smallest hint of boost.

I have never read up on how the stock system works, so if what I am saying sounds familiar to it, it is not meant to. I guess great minds think alike...

Here is what I want the whole setup to be:

A custom manifold that is kind of like an X. It will merge the two exhaust streams then seperate for each turbo.

Where the X branches off for the two turbos, there will be a butterfly valvle in one of them that is controlled by an actuator with a progressive rate spring to open it as the boost on the first turbo builds. It will remain open 10 - 20% with no boost from the first turbo.

There will be two of the turbo's that I posted above. They seem about the right size I guess. If anyone has any better ideas let me know. I like them because they have internal wastegates, which makes for easier plumbing.

In the intake before the intercooler there will be a Y type merge collector with a flapper valve in it that will be closed automatically on one side. This will be pushed on from the second turbo as it builds boost, which will be progressive.

Or there could be another butterfly valve for the intake just like the exhaust, except it will remain 0% open until it has boost behind it.

From there it goes in as normal. I am sure this could be done rather cheaply. I can get some pipes for the manifold and DP and have them bent. The only real hard pieces to fab, would be the valves. But I think it should be easy enough.

What do you think?

Kenton

 

The CCV valve on 3rd gens opens either when the secondary boost exceeds the primary or when the ECU commands it to open whichever comes first.

So you're just doing a slightly simplified version of the Mazda system. The difference is that you rely on full pressure on the primary turbo to engage the secondary. This wont work when the primary is out of flow range.

Let me know when you give up and just change the Mazda system's compressor sizes

-pete

 

I'm not waiting on full pressure from the first to engage the second anymore. If you notice, the second will start spooling with the first because of the valve being opened a little from idle. Since there is a progressive rate spring in the actuator, it will open it more and more as the boost builds.

That way the turbo will be engaging long before the first reaches full boost, but wont be providing airflow until it matches the boost level of the first.

Kenton

 

IF you think about how it works he's right in saying that 2 equivalent sized turbos will work just fine. The first for spoolup on low end and then switch on a second one so both are online at the same time will not neccessarily run the first turbo any faster, but will use the second turbo for additional exhaust and air flow!
The stock system basically works just like that. I wouldn't bother with any delayed switching besides the rpm switch to tune where you want the flapper to switch the second turbo online.
The 3rd gen setup is fine, but just not clean enough for me. I would want 2 turbos facing forward and ducted likely.
Patrick

 

Thanks for the support Megaspeed, but he deffinately had some very valid points whether he got the 2 turbos will work idea or not.

I am trying to avoid RPM switches so that there is not the kick in power. I want a completely seemless power delivery for better predictabilty of traction. This is why I want a valve that progressively lets the airflow in from the second turbo.

I did think of something that might be a problem, but I am not sure. If I have the intake valve controlled by the boost of the second turbo, wont it build boost behind the valve too quick and try and open it before it has enough airflow to support an even stream from both turbos? If this happens, will there be turbo backspin?

I am going to have to figure this one out so that the valve not only opens against boost pressure, but against airflow too. This way it will open up just a little and let the pressure escape into the intake tract that is building up behind the valve, but also allow the turbo to get the speed behind it to open the valve all the way and get some real boost.

Kenton

 

OK now I get you on the engagement stuff. You still might look at letting the closed position be 0 not 10%. The primary needs all the exhaust it can get @ low RPM. But the progressive rate idea is good if it can be reliable.

There's a company in Australia that make an exhaust flapper valve that is controlled by an external pressure source. They are designed for dual exhaust setups (as in the cosmo) but they could work on a turbo too....

PM me if you want me to look them up.

As for your problem on the charge control - the FD system can be wired so it opens in proportion to the ratio of the two turbo's boosts. The door opens gradually when wired this way so I dont see a problem.

I'm starting to like this idea of yours....



-pete

 

but bear in mind you're still gonna shut the exhaust valve on gear changes which screws you. Let me know when you come on board with the RPM triggered solenoid idea. If only you could get electric exhaust flapper valves......

-pete

 

I knew it would catch on!

But I am trying to figure out what is the problem with the exhaust valve closing? Once you get on the gas again, it just starts the process over. If the turbos keep the RPM up, it will happen quite quickly too.

The only reason I am not liking the idea of RPM switching is because of non-linearity.

I think this will be easy once I get all the bugs worked out in my head. AutoCAD would be nice right now with some 3D modeling.

I would deffinately like to see the flapper valves that are already being manufactured, I will send you a PM at once!

Kenton

 

The RPM switch is not to decide the transition from the 1st to the 2nd it's to stop it from reverting to the 1st in off-boost high RPM situation.

Pressure lines to a flapper (via a check valve) engage the secondary. The solenoid just bleeds that pressure away below (say) 3250 rpm to disable the secondary when the revs drop down again.

I am going to do this for sure. I am just waiting for a vaccum hose to blow to give me an excuse ...

Get it ?

-pete

 

Ok I agree that using the check valve and selonoid as a dump would be better for high RPM response, and I have just gotten an idea for making the slow changeover.

This should work with some luck and testing but you could just put a TINY orrifice or valve inline with that pressure signal to the switchover diaphragm the changeover would only go as fast as the restriction would allow volume to fill the diaphragm with air; and this would apply in either way (increasing RPM, and shifting fast). It just depends on how slow you want the changeover to determine the how small of a restriction pill or even a tiny valve that could be adjusted. Any problems here? I got this idea from the secondary butterfly openening diaphragm on the TB of my TII. IT has a tiny orrifice to delay the opening, but if you run a pressure signal to it it will open faster under boost!

I also don't forsee a problem making a smooth transition and backspinning because when exhaust finds another turbine housing to go through its going to make that turbine start spinning reguardless. If you have problems it will be solved by delaying the switchover even higher up in the RPM range. Boy would I like to get this thing setup on my car!
Patrick

 

That sounds like it will do the same thing as the progressive spring. With a little boost it will only compress the spring a little becuase the resitance would start to increase on the spring. Deffinately a good idea though.

If only I had an FD to start experimenting on...sigh. I will have soon enough!

I think this is a much better setup than a single. It just has more of a real world application. Drag, on the other hand, would probably go more towards a single setup.

Kenton

 

What about this cheap cast iron one from JC whitney.
http://www.jcwhitney.com/product.jht...D=5199&BQ=jcw2

I think it will work PERFECTLY! and its cheap enough to build a prototype out of. I think I might talk to some of my professors about this being my senior project.

Patrick

 

That looks perfect actually. Thanks Megaspeed.

I now see the problem with the exhaust valve closing when you let off the throttle after a PM with rpm_pwr. Here is my fix...

The spring loaded actuator for the exhaust control valve can be filled with fluid on the spring side. There will be a tube connected to a resevoir that the fluid goes into when the valve opens. When it tries to close the valve, it will have to do it at a smooth rate because of the fluid going back into the actuator.

The speed at which the valve closes can be regulated by the thickness of the liquid. Say use motor oil. If you want it to close slower, up the weight.

Kenton

 

Don't forget this stuff has to be close to the turbos and manifold. It is possible that this fluid could boil or errupt into flames! Its almost the same principle that I said about the restrictor, except more messy.
Patrick