Twin Scroll Turbos?? Monster Spool.. FACT or FICTION???
so did this discussion go from a twin scroll set up vs. a single scroll. to now comparing divided hotsides to open hotsides? or are we getting more indepth with something i missed a few pages back
*i was in the process of making a comparision but it would have gone on for too long.
*i was in the process of making a comparision but it would have gone on for too long.
A log showing response time on a shift or a lift throttle and mat it again in the same gear should show you how the boost characteristics or each turbo compare. The problem with all these partial throttle claims is that noone has an exact throttle stop to be sure the cars are at the exact same throttle opening. Also a big throttle bodied car will feel more "responsive" too because it allows more airflow for the same throttle input. 3rd gear pulls or 4th gear pulls at low engine speeds in vdr should show how the cars respond with the different turbos.
Another thing to consider is that unless is it the same car the results can be skewed. Some cars just make more power and respond better than others.
Another thing to consider is that unless is it the same car the results can be skewed. Some cars just make more power and respond better than others.
Just curious
When you're not an engineer or when you don't want to believe the documented technical merits of twinscroll - you need to see it to believe it... There are many 'experts' in the EVO industry that don't rely on engneering when developing parts, it's mostly trial and error to find what works best... I'm not knocking them, just stating a simple observation...
+1 Maybe Mikes finding that the fastest roadrace cars run twinscroll, a lot of OEM and auto manufacturers are moving towards twinscroll and even the new F1 discussions regarding moving back to turbo engines all involve twinscroll plans...
When you're not an engineer or when you don't want to believe the documented technical merits of twinscroll - you need to see it to believe it... There are many 'experts' in the EVO industry that don't rely on engneering when developing parts, it's mostly trial and error to find what works best... I'm not knocking them, just stating a simple observation...
When you're not an engineer or when you don't want to believe the documented technical merits of twinscroll - you need to see it to believe it... There are many 'experts' in the EVO industry that don't rely on engneering when developing parts, it's mostly trial and error to find what works best... I'm not knocking them, just stating a simple observation...
I may have said this before, but if one is just interested in putting his foot down and holding it there for 8-9-10 seconds, true, TS probably isn't advantageous.
For those who anticipate episodes of longer duration, where the pedal goes up-down-up-down, etc., TS brings benefits that are immediately noticeable.
That's about as simple as it can be explained. Anyone who can make their way out of a paper bag should be able to understand.
For those who anticipate episodes of longer duration, where the pedal goes up-down-up-down, etc., TS brings benefits that are immediately noticeable.
That's about as simple as it can be explained. Anyone who can make their way out of a paper bag should be able to understand.
+1 Maybe Mikes finding that the fastest roadrace cars run twinscroll, a lot of OEM and auto manufacturers are moving towards twinscroll and even the new F1 discussions regarding moving back to turbo engines all involve twinscroll plans...
When you're not an engineer or when you don't want to believe the documented technical merits of twinscroll - you need to see it to believe it... There are many 'experts' in the EVO industry that don't rely on engneering when developing parts, it's mostly trial and error to find what works best... I'm not knocking them, just stating a simple observation...
When you're not an engineer or when you don't want to believe the documented technical merits of twinscroll - you need to see it to believe it... There are many 'experts' in the EVO industry that don't rely on engneering when developing parts, it's mostly trial and error to find what works best... I'm not knocking them, just stating a simple observation...
It is like saying this turbo works wonders on Honda's, it should be great on an Evo. If I recall correctly, about 3 years ago someone went to DB with this same thought and it failed miserably.
If I remember correctly also, I think someone on this board is going to test an open scroll stock framed turbo to see if the difference is noticable also. That is the only test and result that matters.
BTW, I am not saying TS doesn't work. I am going by what a few people that would "run it if it worked." They look for the best thing and use it no matter what. So the people, one in particular, that don't use it and have tested more combinations that almost anyone on here, and has built faster cars than almost anyone on here, haven't seen a difference to switch to TS then who to believe? The people that only sell TS for a lot more than open scroll or the people that could sell either one and have determined that TS isn't worth the time or money? Now this might change depending on what a few people find and the more testing that goes on but this is where we are at this moment in time.
Last edited by jasnm21; May 2, 2010 at 01:18 PM.
I think comparing different cars with different setups is a waste of time and energy. If you want to prove definitively, somebody has to step up and do fair testing. The test has to fairly represent the respective market where this is even a concern. IMO, drag racing is not the market where this is meaningful. I think the single scroll benefits of reduced weight offset any gains that may be had from improved spool. After all, proper launch control pretty much negates spool provided it's a clean run.
If somebody is truly interested in testing this out, it is my opinion they need to design and build a manifold specifically for this purpose. It wouldn't even be all that hard. They also need to focus on a test that shows more street oriented boost levels. Not MAX HP on MAX boost, but a reasonable boost level somebody would be comfortable with on a true street car. No fear of a bad tank of gas causing catastrophic failure. This whole TS debate is about STREET CARS yet everybody wants to compare them for peak HP on 40+ PSI. How many here drive their cars like that???
Manifold Construction
1. Build a slip-fit collector based manifold.
This means the runners and turbo position remains the same for each setup and swaping the turbo involves just swapping the collector and not a complete manifold change. You simply build the manifold so it pairs the correct cylinders and build the collectors similar enough that they can be swapped around. Most likely, this turbo manifold isn't really going to be usable on a car so just build it out of mild steel to save cost.
2. Build a turbo mount that holds the turbo in a given place
3. Build 2 collectors, 1 4-1 with a single gate, 1 4-2 divided collector with either 2 gates or a properly divided single gate
4. Dump the wastegates to atmosphere for simplicity sakes and to keep the downpipe and remaining exhaust exactly the same.
Dyno Test:
1. Do your conventional 2000 RPM lug start test to compare "powerband"
2. Vary start RPM and chart RPM vs. time to compare transient response (this means using an inertial based dyno and not something that ramps RPM vs. time by adjusting the load)
Use "neutral throttle" as the starting condition. Allow the car to hold a given RPM with a constant throttle for several seconds then go WOT. Say 500 RPM increments from 2500-6000 RPM.
Turbo Selection:
1. When selecting housings you may need to slightly customize the housings to make it work. Adding a V-band outlet with extension to a T3 single scroll for example to make the downpipe match up to the twinscroll T4 setup (turbo centerline will be constant).
2. A/R selection - it would be best to have multiple A/R housings available for each. But as a rough rule of thumb, multiple the TS A/R by 0.6 to get a rough idea of the comparable A/R in the single scroll. Thus, a 1.06 A/R T3 TS is roughly equivalent to a 0.64 A/R single scroll T3.
3. Wheel dimensions - It has been pretty well proven different size compressor to turbine matches work better with the different setups. TS stuff likes to have turbine to compressor ratios closer to unity (1.08 - 1.12). Single scroll stuff can often get away with more compressor for a given turbine size (1.16-1.20). A standard GT3582R comes in at 1.20 which favors the single scroll setup to start with. I'd suggest a turbo selection that falls in the middle around a 1.14 ratio so that it doesn't really benefit either turbo.
While not a favorite turbo, for the purpose of testing a large turbo street car setup, I'd consider using a T04Z. 1.14:1 Compressor to turbine ratio and readily available in a large number of T4 divided and undivided housings. It may not set an HP or response record, but it would provide a fair balance to each system and show each strengths and weakness.
Engine Selection:
Unfortunately, the TS stuff has proven to work well with extra overlap as the TS system exhibits the ability to properly scavenge on the exhaust pulse. My only suggestion is anticipate potentially using fairly aggressive cams and be willing to adjust the cam timing to benefit each system. Single scroll with large turbos and moderate boost levels seem to work well with aggressive cams too, so it's mostly just an issue of overlap.
That said, I have driven and ridden in several large turbo cars running single scroll and twin scroll setups. I have no doubt that a twin scroll setup can make a large turbo much more street friendly. For example, a 2.6L Subaru running a GT4294R on a single scroll setup barely even got the turbo going in 1st gear and it wasn't until 3rd gear that the car was getting full boost. On the twinscroll setup with a larger GT4202 turbo and a smaller 2.55L motor, it was ripping to full boost in first gear with almost zero lag in second. Full boost IN SECOND by 5000 RPM from a 4000 RPM start. Even for a drag setup, despite the lower weight of the single scroll setup, I feel the TS has the added benefit of boost recover for when things don't go right (misshift, bogged launch, etc.). Import drag racing still isn't to the point where the field is determined by fractions of a second and being able to recover from a mistake may still give you a win light.
If somebody is truly interested in testing this out, it is my opinion they need to design and build a manifold specifically for this purpose. It wouldn't even be all that hard. They also need to focus on a test that shows more street oriented boost levels. Not MAX HP on MAX boost, but a reasonable boost level somebody would be comfortable with on a true street car. No fear of a bad tank of gas causing catastrophic failure. This whole TS debate is about STREET CARS yet everybody wants to compare them for peak HP on 40+ PSI. How many here drive their cars like that???
Manifold Construction
1. Build a slip-fit collector based manifold.
This means the runners and turbo position remains the same for each setup and swaping the turbo involves just swapping the collector and not a complete manifold change. You simply build the manifold so it pairs the correct cylinders and build the collectors similar enough that they can be swapped around. Most likely, this turbo manifold isn't really going to be usable on a car so just build it out of mild steel to save cost.
2. Build a turbo mount that holds the turbo in a given place
3. Build 2 collectors, 1 4-1 with a single gate, 1 4-2 divided collector with either 2 gates or a properly divided single gate
4. Dump the wastegates to atmosphere for simplicity sakes and to keep the downpipe and remaining exhaust exactly the same.
Dyno Test:
1. Do your conventional 2000 RPM lug start test to compare "powerband"
2. Vary start RPM and chart RPM vs. time to compare transient response (this means using an inertial based dyno and not something that ramps RPM vs. time by adjusting the load)
Use "neutral throttle" as the starting condition. Allow the car to hold a given RPM with a constant throttle for several seconds then go WOT. Say 500 RPM increments from 2500-6000 RPM.
Turbo Selection:
1. When selecting housings you may need to slightly customize the housings to make it work. Adding a V-band outlet with extension to a T3 single scroll for example to make the downpipe match up to the twinscroll T4 setup (turbo centerline will be constant).
2. A/R selection - it would be best to have multiple A/R housings available for each. But as a rough rule of thumb, multiple the TS A/R by 0.6 to get a rough idea of the comparable A/R in the single scroll. Thus, a 1.06 A/R T3 TS is roughly equivalent to a 0.64 A/R single scroll T3.
3. Wheel dimensions - It has been pretty well proven different size compressor to turbine matches work better with the different setups. TS stuff likes to have turbine to compressor ratios closer to unity (1.08 - 1.12). Single scroll stuff can often get away with more compressor for a given turbine size (1.16-1.20). A standard GT3582R comes in at 1.20 which favors the single scroll setup to start with. I'd suggest a turbo selection that falls in the middle around a 1.14 ratio so that it doesn't really benefit either turbo.
While not a favorite turbo, for the purpose of testing a large turbo street car setup, I'd consider using a T04Z. 1.14:1 Compressor to turbine ratio and readily available in a large number of T4 divided and undivided housings. It may not set an HP or response record, but it would provide a fair balance to each system and show each strengths and weakness.
Engine Selection:
Unfortunately, the TS stuff has proven to work well with extra overlap as the TS system exhibits the ability to properly scavenge on the exhaust pulse. My only suggestion is anticipate potentially using fairly aggressive cams and be willing to adjust the cam timing to benefit each system. Single scroll with large turbos and moderate boost levels seem to work well with aggressive cams too, so it's mostly just an issue of overlap.
That said, I have driven and ridden in several large turbo cars running single scroll and twin scroll setups. I have no doubt that a twin scroll setup can make a large turbo much more street friendly. For example, a 2.6L Subaru running a GT4294R on a single scroll setup barely even got the turbo going in 1st gear and it wasn't until 3rd gear that the car was getting full boost. On the twinscroll setup with a larger GT4202 turbo and a smaller 2.55L motor, it was ripping to full boost in first gear with almost zero lag in second. Full boost IN SECOND by 5000 RPM from a 4000 RPM start. Even for a drag setup, despite the lower weight of the single scroll setup, I feel the TS has the added benefit of boost recover for when things don't go right (misshift, bogged launch, etc.). Import drag racing still isn't to the point where the field is determined by fractions of a second and being able to recover from a mistake may still give you a win light.
Last edited by 03whitegsr; May 2, 2010 at 01:30 PM.
If somebody is truly interested in testing this out, it is my opinion they need to design and build a manifold specifically for this purpose. It wouldn't even be all that hard. They also need to focus on a test that shows more street oriented boost levels. Not MAX HP on MAX boost, but a reasonable boost level somebody would be comfortable with on a true street car. No fear of a bad tank of gas causing catastrophic failure. This whole TS debate is about STREET CARS yet everybody wants to compare them for peak HP on 40+ PSI. How many here drive their cars like that???
Manifold Construction
1. Build a slip-fit collector based manifold.
This means the runners and turbo position remains the same for each setup and swaping the turbo involves just swapping the collector and not a complete manifold change. You simply build the manifold so it pairs the correct cylinders and build the collectors similar enough that they can be swapped around. Most likely, this turbo manifold isn't really going to be usable on a car so just build it out of mild steel to save cost.
2. Build a turbo mount that holds the turbo in a given place
3. Build 2 collectors, 1 4-1 with a single gate, 1 4-2 divided collector with either 2 gates or a properly divided single gate
4. Dump the wastegates to atmosphere for simplicity sakes and to keep the downpipe and remaining exhaust exactly the same.
Dyno Test:
1. Do your conventional 2000 RPM lug start test to compare "powerband"
2. Vary start RPM and chart RPM vs. time to compare transient response
Use "neutral throttle" as the starting condition. Allow the car to hold a given RPM with a constant throttle for several seconds then go WOT. Say 500 RPM increments from 2500-6000 RPM.
Turbo Selection:
1. When selecting housings you may need to slightly customize the housings to make it work. Adding a V-band outlet with extension to a T3 single scroll for example to make the downpipe match up to the twinscroll T4 setup (turbo centerline will be constant).
2. A/R selection - it would be best to have multiple A/R housings available for each. But as a rough rule of thumb, multiple the TS A/R by 0.6 to get a rough idea of the comparable A/R in the single scroll. Thus, a 1.06 A/R T3 TS is roughly equivalent to a 0.64 A/R single scroll T3.
3. Wheel dimensions - It has been pretty well proven different size compressor to turbine matches work better with the different setups. TS stuff likes to have turbine to compressor ratios closer to unity (1.08 - 1.12). Single scroll stuff can often get away with more compressor for a given turbine size (1.16-1.20). A standard GT3582R comes in at 1.20 which favors the single scroll setup to start with. I'd suggest a turbo selection that falls in the middle around a 1.14 ratio so that it doesn't really benefit either turbo.
While not a favorite turbo, for the purpose of testing a large turbo street car setup, I'd consider using a T04Z. 1.14:1 Compressor to turbine ratio and readily available in a large number of T4 divided and undivided housings. It may not set an HP or response record, but it would provide a fair balance to each system and show each strengths and weakness.
That said, I have driven and ridden in several large turbo cars running single scroll and twin scroll setups. I have no doubt that a twin scroll setup can make a large turbo much more street friendly. For example, a 2.6L Subaru running a GT4294R on a single scroll setup barely even got the turbo going in 1st gear and it wasn't until 3rd gear that the car was getting full boost. On the twinscroll setup with a larger GT4202 turbo and a smaller 2.55L motor, it was ripping to full boost in first gear with almost zero lag in second. Full boost IN SECOND by 5000 RPM from a 4000 RPM start.
Manifold Construction
1. Build a slip-fit collector based manifold.
This means the runners and turbo position remains the same for each setup and swaping the turbo involves just swapping the collector and not a complete manifold change. You simply build the manifold so it pairs the correct cylinders and build the collectors similar enough that they can be swapped around. Most likely, this turbo manifold isn't really going to be usable on a car so just build it out of mild steel to save cost.
2. Build a turbo mount that holds the turbo in a given place
3. Build 2 collectors, 1 4-1 with a single gate, 1 4-2 divided collector with either 2 gates or a properly divided single gate
4. Dump the wastegates to atmosphere for simplicity sakes and to keep the downpipe and remaining exhaust exactly the same.
Dyno Test:
1. Do your conventional 2000 RPM lug start test to compare "powerband"
2. Vary start RPM and chart RPM vs. time to compare transient response
Use "neutral throttle" as the starting condition. Allow the car to hold a given RPM with a constant throttle for several seconds then go WOT. Say 500 RPM increments from 2500-6000 RPM.
Turbo Selection:
1. When selecting housings you may need to slightly customize the housings to make it work. Adding a V-band outlet with extension to a T3 single scroll for example to make the downpipe match up to the twinscroll T4 setup (turbo centerline will be constant).
2. A/R selection - it would be best to have multiple A/R housings available for each. But as a rough rule of thumb, multiple the TS A/R by 0.6 to get a rough idea of the comparable A/R in the single scroll. Thus, a 1.06 A/R T3 TS is roughly equivalent to a 0.64 A/R single scroll T3.
3. Wheel dimensions - It has been pretty well proven different size compressor to turbine matches work better with the different setups. TS stuff likes to have turbine to compressor ratios closer to unity (1.08 - 1.12). Single scroll stuff can often get away with more compressor for a given turbine size (1.16-1.20). A standard GT3582R comes in at 1.20 which favors the single scroll setup to start with. I'd suggest a turbo selection that falls in the middle around a 1.14 ratio so that it doesn't really benefit either turbo.
While not a favorite turbo, for the purpose of testing a large turbo street car setup, I'd consider using a T04Z. 1.14:1 Compressor to turbine ratio and readily available in a large number of T4 divided and undivided housings. It may not set an HP or response record, but it would provide a fair balance to each system and show each strengths and weakness.
That said, I have driven and ridden in several large turbo cars running single scroll and twin scroll setups. I have no doubt that a twin scroll setup can make a large turbo much more street friendly. For example, a 2.6L Subaru running a GT4294R on a single scroll setup barely even got the turbo going in 1st gear and it wasn't until 3rd gear that the car was getting full boost. On the twinscroll setup with a larger GT4202 turbo and a smaller 2.55L motor, it was ripping to full boost in first gear with almost zero lag in second. Full boost IN SECOND by 5000 RPM from a 4000 RPM start.
I am very interested in a TS because I live at altitude and I want something responsive but can give high HP. But I also think that someone that is building a car for high HP isn't using it as a DD, I know I won't be that is why I have another car.
Evolving Member
Joined: Apr 2004
Posts: 344
Likes: 0
From: san antonio
that 30r was for the guy who asked. not you and 700+hp. no it will never do that. if it did i think the compressor wheel would spin right out through the front of the car the 1st time it came close. lol i know its old tech but even again right there your comparing new stuff to old.
find a single scroll hta86 VS twin scroll hta86. with similar mods,boost,fuel
find a single scroll hta86 VS twin scroll hta86. with similar mods,boost,fuel
Based on 03White's post it becomes obvious to me that I really would want a housing larger than the 1.06 T3. Will it be satisfy the majority if it got dyno'd as a T3 then a T4 or does it need to be a T4 on both? 03 has it outlined as acceptable for his engineering tastes.
Then the test-
I dont mind doing a 4 gear pull (not 4th...FOUR) on the dyno to show what it does on the street. I could then do partial throttle stuff and transient response tests and just log it in EvoScan in a controlled enviroment.
It all boils down to ETS having the stuff for me to borrow to be clear. I am willing to use my car to test but I am not going to buy the pieces simply to test.
Me personally, I think it would be best to maintain frame size as it then gives a better respective size to allow direct comparison between the A/R sizes. Also, going T4 gives you more choices on A/R for the twin scroll stuff and would likely allow a size closer to an undivided 0.82 A/R T3 since the largest divided T3 is about the same as the 0.64 A/R undivided T3 which has been proven to be too small on a GT35R.
But I also recognize that realistically, that's not how you would run them so it's not really fair to the single scroll stuff. Mike has proven the T3 frame can go fast and there is no reason to go T4 on single scroll stuff.
If you do not keep frame size constant, to me it seems most relevant to then match up based on powerband of the 2000 RPM started dyno curve. The boost curves are likely to be different, so maybe just set your comparison point on a given torque value. Say where the pull crosses 300 ft-lbs and match up those points between the sizes.
Also, I think it's imperative to keep the externals constant. You are interested in TS vs. SS, so the ONLY thing changed must be directly related to those items. Same turbo placement, same downpipe placement, same manifold runners. Only the collector, turbine housing and wastegate should be changed.
I personally think there is too much vested interest that's been displayed for anything short of this type of test to be relevant. Each side seems to want to focus on different aspects. The SS guys focus on the fact that for a given A/R, the TS chokes and that's why it spools better. The TS guys focus on a difficult to quantify "drive ability" standard. I think the method I have proposed covers both parties as equally as possible.
I also realize only 1 of 2 types of companies are going to test this.
1. the company that wants to develop the best system. They will approach it as something they are going to have to spend money to get the true impact. For them, it's got to have a potential for making a quantifiable better system. If they invest the money to figure it out, it's going to be difficult for them to hand over all the data as a competitor will then be able to use their efforts.
2. The company that has something to gain by showing their competition doesn't perform as well. These type will skew the data as far as they can. They will make unfair comparisons and any" data" gathered (no matter how biased) will be sold as truth.
But I also recognize that realistically, that's not how you would run them so it's not really fair to the single scroll stuff. Mike has proven the T3 frame can go fast and there is no reason to go T4 on single scroll stuff.
If you do not keep frame size constant, to me it seems most relevant to then match up based on powerband of the 2000 RPM started dyno curve. The boost curves are likely to be different, so maybe just set your comparison point on a given torque value. Say where the pull crosses 300 ft-lbs and match up those points between the sizes.
Also, I think it's imperative to keep the externals constant. You are interested in TS vs. SS, so the ONLY thing changed must be directly related to those items. Same turbo placement, same downpipe placement, same manifold runners. Only the collector, turbine housing and wastegate should be changed.
I personally think there is too much vested interest that's been displayed for anything short of this type of test to be relevant. Each side seems to want to focus on different aspects. The SS guys focus on the fact that for a given A/R, the TS chokes and that's why it spools better. The TS guys focus on a difficult to quantify "drive ability" standard. I think the method I have proposed covers both parties as equally as possible.
I also realize only 1 of 2 types of companies are going to test this.
1. the company that wants to develop the best system. They will approach it as something they are going to have to spend money to get the true impact. For them, it's got to have a potential for making a quantifiable better system. If they invest the money to figure it out, it's going to be difficult for them to hand over all the data as a competitor will then be able to use their efforts.
2. The company that has something to gain by showing their competition doesn't perform as well. These type will skew the data as far as they can. They will make unfair comparisons and any" data" gathered (no matter how biased) will be sold as truth.