Kevin's Compound Sequential Twin Turbo Design

I've been thinking of doing a sequential turbo setup for a while now as a means of getting more boost at lower rpm, so that I can run a lower stall in my TH400 converter. A lower stall will allow me to have a converter that holds more power and has less converter slippage, which results in more power to the rear wheels.

Having looked at every compound, sequential, and twin turbo setup that I could find on the internet, there wasn't one of them that I was happy to install in my own car due to various fitment, cost and complexity reasons. Therefore I decided to design my own sequential turbo system.

They say a picture is worth a thousand words, so here's a diagram of how it works.

This sequential setup is basically a compound turbo setup on the exhaust side (but with identical size turbos) and a sequential setup on the intake side. This means you get all the advantages of a compound turbo setup for spooling the turbos, but with none of the disadvantages of the compound setup such as the boost being too high for a gas engine. Also both turbos are exactly the same size, so you don't get any issues with different size turbos fighting each other in the sequential setup and having the large one potentially overpowering the smaller one.

By spooling the second turbo with the gas exiting from the first turbo, the second turbo will spool much quicker then if it was just receiving some exhaust gas diverted from the first turbo, as most other sequential turbo systems do. This means that instead of having a big dip in the torque curve when the second turbo comes on-line, there should be little to no dip in the torque, resulting in a torque curve more like a large single turbo.

This setup is super simple and only requires 3 devices to be controlled, which is simpler then any other sequential turbo setup I've seen. There are two wastegates, one used for each turbo and they are set to exactly the same boost level. These wastegates need to be controlled separately as the two turbos will require different duty cycles on the wastegate to maintain the same boost level. This can be done via two boost controllers or via one boost controller and one aftermarket ECU. There is also one combined reed/air intake valve (with an attached actuator) required that needs be opened when the second turbos boost catches up to the first turbos boost pressure. This could either be done using a Hobbs switch set to the same pressure as both wastegates are or by an aftermarket ECU controlling the flow of boost pressure to the air intake valve's actuator.

My design uses a proper high flowing manifold to the first turbo, eliminating the ugly and often poor flowing log manifold seen in many sequential designs. Here's a drawing I did to show how my sequential setup will be implemented in practice on my 2JZ-GTE Supra engine.

My design uses a 3.5" down pipe between the #1 and #2 turbo, a 4.0" down pipe for the #2 turbo and the 60mm wastegates have 3.0" pipes to maximize the flow. This is the 2JZ-GTE intake air control valve I will be using. The curved section will be removed to make it straight through.

I am planning on changing my single turbo 2JZ-GTE Supra to this setup over the next winter season and I will update my website with the results from this setup then. My setup will use two Turbonetics 60mm wastegates, the Supra OEM reed/air intake control valve (50mm) and two MasterPower race series R6164 (61mm) turbos. These turbos are good for 700hp each on gas when run around 30-35psi boost, making for a potential of 1400hp from both turbos on gas or 1500+ hp when using E85. I use E85 and my goal is 1000whp through the TH400 gearbox which is around 1300hp. This will hopefully be achievable with around 35psi boost.

As this is a different design to what anyone else has done, its possible that it may require some tweaking to work correctly. After discussing this design with a friend, there are two areas that I think may cause issues. One is that the 60mm wastegate from the #1 turbo may not allow enough exhaust flow around the #1 turbo resulting in excessive back pressure or over boosting. The other is that the #2 turbo may cause too much back pressure in the #1 turbo's down pipe, which will reduce how quickly the #1 turbo spools. Both these issues can be resolved with some design changes if they do occur.

All constructive comments and suggestions are welcome.

Seems like a good idea, good luck with it mate

Love your threads mate!

Question, wouldn't the heat be an issue? With a 700hp turbo exhaust housing flowing over 700+ isn't that going to cause a backup of pressure and potentially an extremely hot housing?

I always thought that was the main concern with turbo's in series and not in parallel such as most big twin setups run these days.

Love your threads mate!

 

Question, wouldn't the heat be an issue? With a 700hp turbo exhaust housing flowing over 700+ isn't that going to cause a backup of pressure and potentially an extremely hot housing?

 

I always thought that was the main concern with turbo's in series and not in parallel such as most big twin setups run these days.

Thanks.

The #1 wastegate prior to the #1 turbo is to bypass all unneeded exhaust gas around the #1 turbo, so minimizing back pressure in the manifold and #1 turbine. It has crossed my mind that I may need to bypass more exhaust flow then one 60mm wastegate can handle, and if so I'll try and squeeze in another wastegate there as well, perhaps one on each side of the manifold.

As you say heat is definitely an issue with so many pipes and turbos. To try and counter this, I will be wrapping the manifold, both the down pipes and both wastegate pipes to try and minimize the engine bay heat.

Wastegate 2 would need to be large enough to cope with bypassing both turbos.

I'm wondering if the exhaust gas will have enough oomph left to do a proper job on the second turbo. You lose a few hundred degrees through the turbo so not so much exhaust energy is left to run the second turbo. The wastegate would feed the second turbo with proper hot gas, of course, which might compensate for it, but there might be some sort of odd boost 'shelf' where the second turbos doesn't really start spooling until the first turbo is properly on boost, so you stay at one boost level or even it drops off some as the revs rise.

If I was going to make a compound system I'd nick the BMW design, but I think you should have a go at this anyway as without trying, we'd get nowhere

Works very well on diesels, I know the US guys have done a setup on a supra but didn't seem too popular.

Interesting to see how it does though.

Wastegate 2 would need to be large enough to cope with bypassing both turbos.

Yes it does. Its a 60mm wastegate which should do the trick.

I'm wondering if the exhaust gas will have enough oomph left to do a proper job on the second turbo. You lose a few hundred degrees through the turbo so not so much exhaust energy is left to run the second turbo. The wastegate would feed the second turbo with proper hot gas, of course, which might compensate for it, but there might be some sort of odd boost 'shelf' where the second turbos doesn't really start spooling until the first turbo is properly on boost, so you stay at one boost level or even it drops off some as the revs rise.

I will definitely be doing this as I'm pretty sure it will work and testing will tell how well. My unique divided manifold quick spool valve is working very nicely and everyone (in the USA) thinks it can't be done. LOL.

If I was going to make a compound system I'd nick the BMW design, but I think you should have a go at this anyway as without trying, we'd get nowhere

I'm not familiar with their system, but I'll Google it. The one I'm most familiar with is the Boost Logic design.

Works very well on diesels, I know the US guys have done a setup on a supra but didn't seem too popular.

 

Interesting to see how it does though.

That's the Boost Logic system. It has some similarities with how mine works on the exhaust side, but it uses different size turbos and controls the wastegates differently.

I love making this kinda thing

Only problem you may encounter is the first turbo overspooling unable to bypass enough thru the wg into a pressurised pipe that's feeding the second turbo

A 60mm may be big enough

Why don't you just do your own sequential tubular manifold setup?

I love making this kinda thing

Only problem you may encounter is the first turbo overspooling unable to bypass enough thru the wg into a pressurised pipe that's feeding the second turbo

A 60mm may be big enough

A 60mm was big enough for the Boost Logic compound system but I may need more flow then they did. If so, two 50mm or 60mm wastegates could be used if I can squeeze them in there.

Why don't you just do your own sequential tubular manifold setup?

I have thought of doing something like this with a big pipe joining the two manifolds at the same place the wastegate pipe would fit. It would work very well, but I think my setup will out spool a standard sequential design. E.g. Boost Logic got 20psi by 2800rpm from their compound setup.

This is the kinda route I'm going down

[ATTACH=CONFIG]176866[/ATTACH][ATTACH=CONFIG]176867[/ATTACH]

 

This is the kinda route I'm going down

That is not what I call a well designed sequential system. Check out this link.http://www.supraforums.com/forum/showthread.php?512779-Another-SP-Sequential-Build

And here is why. The log manifold in that design results in poor flow during the spool up phase which is why their setup doesn't spool that quickly. It also had back pressure problems with the #1 turbo and a huge dip in the torque curve during the transition when the #2 turbo comes on line.

I'm not familiar with their system, but I'll Google it. The one I'm most familiar with is the Boost Logic design.

We discussed it here:

http://www.mkivsupra.net/vbb/showthread.php?296645-New-Developments-in-Supra-Turbo-applications

(Page 5 onwards gets a bit more in-depth)

It has compact packaging, elegant design, it's really simple to plumb in and control, and I think would be streets ahead of anything else. It compounds the intake charge rather than the exhaust system so won't suffer from any backpressure issues trying to run a second turbo off the used-up output from the first. You use a small first turbo to deliver rapid response and low down power, compound charge from it and a bigger turbo for the transition, then just the big turbo up top.

If you've got the resources to have a go at it, I think it'd really be worth the time. Someone with the facilities and experience to fab up what you're planning to do should be able to prototype this setup in a relatively short time

We discussed it here:

http://www.mkivsupra.net/vbb/showthread.php?296645-New-Developments-in-Supra-Turbo-applications

 

(Page 5 onwards gets a bit more in-depth)

 

It has compact packaging, elegant design, it's really simple to plumb in and control, and I think would be streets ahead of anything else. It compounds the intake charge rather than the exhaust system so won't suffer from any backpressure issues trying to run a second turbo off the used-up output from the first. You use a small first turbo to deliver rapid response and low down power, compound charge from it and a bigger turbo for the transition, then just the big turbo up top.

 

If you've got the resources to have a go at it, I think it'd really be worth the time. Someone with the facilities and experience to fab up what you're planning to do should be able to prototype this setup in a relatively short time

Thanks for the link. It was a good read.

Believe it or not, the BMW setup is EXACTLY the same as the Boost Logic compound system ended up (on the compressor side) in its final form. During testing Boost Logic found that the #1 turbo (~50mm) was causing too much back pressure at high boost from the #2 turbo, so they introduced a bypass from the #2 turbo at high boost directly to the intercooler to solve that issue.

The pics I've looked at of the BL kit don't have that bypass valve which is on the BMW setup, so yeah I can see the #1 turbo being a serious bottleneck at high revs high boost They went for some pretty hefty turbos and 3" intake piping I see, I would have thought sacrificing ultimate flow for the packaging and plumbing simplicity of smaller diameter pipework would have been worth it.

In my mind I was thinking of an almost-stock-size first turbo and then something of a T67 nature, and 2.5" tubing max. Instanto low boost, smaller footprint and cost, and a very useful 600bhp on the street.

20psi of boost under 3000rpms is mad, though

Edit - actually there is a big difference - the BMW has separate exhaust feeds for the two turbos, the exhaust side isn't ever compound. The BL kit is from what I can see and have heard. The BMW just shuts off the exhaust feed to the first turbo at some point in the rev range so it all gets diverted to #2. That eliminates any backpressure issues, but then the castings are so tight it probably causes just as much anyway. It's for a different market though, getting turbodiesels to be OEM powerful rather than loony road cars.

The pics I've looked at of the BL kit don't have that bypass valve which is on the BMW setup, so yeah I can see the #1 turbo being a serious bottleneck at high revs high boost They went for some pretty hefty turbos and 3" intake piping I see, I would have thought sacrificing ultimate flow for the packaging and plumbing simplicity of smaller diameter pipework would have been worth it.

 

In my mind I was thinking of an almost-stock-size first turbo and then something of a T67 nature, and 2.5" tubing max. Instanto low boost, smaller footprint and cost, and a very useful 600bhp on the street.

 

20psi of boost under 3000rpms is mad, though

I haven't seen pics of it either, but on one of the forum discussions, BL mentioned that they had to do this. Here is a quote from that discussion:

"The max power issue has been resolved, a third wastegate was added to the charge pipe between the large turbo and the small turbo that will vent the charge directly to the intercooler pipe (bypass compressor housing of small turbo)."

Edit: see post #70: http://www.supraforums.com/forum/showthread.php?548363-Compound-Turbo-Kit/page2

Edit - actually there is a big difference - the BMW has separate exhaust feeds for the two turbos, the exhaust side isn't ever compound. The BL kit is from what I can see and have heard. The BMW just shuts off the exhaust feed to the first turbo at some point in the rev range so it all gets diverted to #2. That eliminates any backpressure issues, but then the castings are so tight it probably causes just as much anyway. It's for a different market though, getting turbo diesels to be OEM powerful rather than loony road cars.

Yep, that's why I had already edited my original post to add "(on the compressor side)".

The BL setup requires very extensive tuning to get it working correctly, also according to BL. Basically I cannot see it working correctly without using two 3D maps, one for each wastegate and then another 3D map for the intake air control valve (IACV). That's all fine for BMW to do as they are doing it once for thousands of identical engines, but is very complex for tuning a one-off system. I believe this is why BL never offered it as a kit.

Mine will need a 3D map for the intake air control valve (IACV), but the wastegate controls in my design are very simple in comparison, only requiring the normal 2D map.

I can't imagine this setup working very well with 2 similar sized turbos working with different pressure ratios over their turbines. Once you open the valve in the intake pipe the boost will fluctuate like crazy.

I can't imagine this setup working very well with 2 similar sized turbos working with different pressure ratios over their turbines. Once you open the valve in the intake pipe the boost will fluctuate like crazy.

The purpose of the two wastegates is to control the boost from each turbo seperately. Why do you think the wastegates will not be able to do that?

When you open the first wastegate to slow the first turbo down you will gain alot of energy to the second turbo which will spin faster and raise boost. Then when the second wastegate opens you get lower backpressure in in the pipe between the turbines and that will make the first turbo spin faster. So at the same boost pressure you will need different pulsewidth controlling the first wastegate depending on backpressure in the pipe between the turbos. The only way I see it possible to get any control over boost is to base it on turbo speed. And the cost of that setup for 2 turbos would easily be 1000£

You make some very good points regarding the effects of the turbos on each other. I think it will come down to how quickly and how finely the ECU can control the two wastegates. If not very well then there will be boost cycling but if they can both be controlled very well the boost cycling will be very minimal. Anything less than 1/2 psi would be fine. Only testing will tell for sure.

I bet those pwm-tables will be a real PITA to get right. The best would be closed loop vs turbospeed but that´s alot of money. But I really like that you think outside the box. Just how I like to do things myself

I bet those pwm-tables will be a real PITA to get right. The best would be closed loop vs turbospeed but that´s alot of money. But I really like that you think outside the box. Just how I like to do things myself

Thanks. I'm planning on doing closed loop for each wastegate based on the boost pressure feed directly from each turbo's compressor housing. I guess if that fails, I could try setting it up to use turbine speed.

I'm thinking the IACV control (which probably needs to use 3D tables) will be even more difficult to get right for all driving conditions.