The Split Pulse Turbocharger Thread

[Dale FZ1]

Well there was plenty of discussion on Mafia’s 300rwkW on-a-budget thread, regarding split pulse turbos. Discopotato03 has been urging for the use of a split pulse housing in this higher power range, which roughly equates to 475-500 crank hp. The idea of starting a tech discussion focused on split pulse application for RB series engines was suggested, in order to keep that thread on track. No-one else took the challenge, so here it is. I’m very interested to keep this thread on track too.

Firstly, my intention is to investigate and discuss more of the theory behind split pulse and what makes it work. I don’t mind if we get technical here – I want to learn more than I currently know about this subject.

Secondly, I’d like some effort from you blokes (and girls) to see if we can find out what actual options there are to economically configure a split pulse turbo for an RB over a certain range of streetable power outputs; let’s say between 200rwkW and 300rwkW. I DO mind if we get over the top with cost here – we want some viable, cost-competitive alternatives to the various single scroll units currently fitted to RB engines to get in that range. We’d previously focused on the Garrett range, but I believe there are other brands out there that should be checked out for suitable spec.

We are after a unit that will give as high and wide a torque spread as possible, and within the standard rev range. Early/quick spool, and free breathing high rpm power? Sounds like the “average power†concept that Sydneykid proposed quite a long time back. I’ll take it a step forward, and say that it is about maximum area under the curve. Let’s not get hung up about torque and power; we know that power is merely a function of torque and rpm. My personal preference is to look at the torque curve, but that counts for little. What does count is getting the curve up high, early in the rpm range, and keeping it there for as long as possible.

So who’s up for it?

Edited May 26, 2007 by Dale FZ1

[Devils_Advocate21]

Id go there if i had the funds to as im highly interested in results, The Manifold to support this can easily be done with the 6boost design as its split pulse so 3 collectors each side into a t4 flange split type, and could easily be accomodated to be modded into a dual waste true twin scroll manifold, The housing options as far as i know mainly in the high end scale of the turbos these aquired are like T04z split pulse but can easily import the smaller turbo housing choices for aka gt3076 or gt30 series turbos only by import as far as i know you cant get the housings over here? Correct me if im wrong

[Dale FZ1]

Not entirely sure of availability within Aust. But will be interested to find out. As I said initially, we had focused on Garrett product, but there are other options. I would say the T04Z range would be pretty well outside the 300rwkW cap too.

It’s pretty clear that a true “tuned” split pulse setup relies on correctly designed and built manifolds that use the same principles as N/A extractors. I found this link that helps to explain with pictures:

http://www.awe-tuning.com/media/pdf/2.7T_h..._comparison.pdf

After viewing the pics there, I’d have to say the stock RB20/25 manifold is configured rather like a pair of joined log manifolds. It does seem that Nissan found benefits in keeping the flow from 123 separate from 456 until the turbine inlet though.

If it was a tuned manifold, both the diameter and length of the runners have to be important, and would affect both the mass-flow capacity, response time, and engine rpm where it all works most efficiently (just like extractors). This design approach is about getting efficiency by correlating the timing and strength of exhaust pulses at the collector point, using pressure waves to assist “drafting” the waves emerging from neighbouring exhaust runners. Here’s another link on manifolds and exhaust systems generally, that might be useful: http://www.nsxprime.com/FAQ/Miscellaneous/exhausttheory.htm

Pretty much what I gained from that is ANY form of manifolding has compromises, and there will generally be a leaning either to the bottom end, mid range, or top end performance. Sounds like the rest of the engineering in your RB, and makes it clear to me that we should be looking to complement the rest of the engine with manifold and turbo.

[Dale FZ1]

It’s interesting to check out the market and see what machines are generally fitted with split pulse turbos. We’ve seen evidence of Toyota having a go with MR2 running 3SGTE and (unconfirmed) early Supra fitted with 7MGTE; but generally otherwise in the Japanese market it’s mostly been the domain of the 4wd rally rockets which require (and can use) maximum torque, early and rapidly. The other common fitment is for various diesels, which as Adrian correctly pointed out in Mafia’s thread have a narrow effective operating engine speed range, so also NEED to develop maximum torque early and rapidly.

Common theme? I’d say yes, particularly when the issue of using full throttle on the rally rockets is factored in. Full throttle = maximum mass air flow for the petrol engine, and diesels do not have a throttle – they get the full whack of flow 100% of the time.

Is it possible that the 2wd Skyline with a split pulse turbo might suffer traction and driveability problems if either

1. this turbo configuration requires full flow (or lots of throttle) to get the pulsing effect and respond (ie. not progressive in delivery)?

2. this turbo configuration responds so vigourously that boost/airflow/torque rises very rapidly (ie. not progressive in delivery)?

OR – do we follow the compromise through, by running a larger A/R housing, with the technical advantage of split pulse partly off-setting the dulled turbine response, with a gain up high because of increased flow + lower turbine inlet pressure? ie. on balance lose little down low, and gain something up top?

It sounds a bit of an airy-fairy question, but one that is reasonable (I think) because of the existing typical application. Hopefully it’s obvious that I’m zeroing-in on whether it’s possible to achieve good transient response and linear delivery at less than 100% WOT, as well as the targeted beefy torque curve on the dyno sheet which is at WOT.

[paulr33]

the common theme

proper setup manifold with split pulse setup

exhaust housing with split pulse

external gate

that will give you the best of everything, but it wont come cheap and it wont look factory OEM either

[Dale FZ1]

With a bit of research, I've found that MHI is running a technology and product sharing arrangement with Holset - similar in some respects to Garrett/HKS. I think their main business focus is on diesel engines.

The main point of interest here is that both MHI and Holset market split pulse turbo units, possibly (hopefully) within the mass-flow and efficiency ranges that could work for the petrol RB.

Available specs seem limited, but two things caught my eye:

1. they are available both free floating (non gated) and with IW, in a range of A/R

2. they produce units with titanium aluminide impellers (resistant to extreme heat)

There is evidently a HUGE range of options, and it would take a fair bit of work to find what could be built, but there could be something in it. Perhaps we might be able to unearth some specs and flow maps?

[Dale FZ1]

the common theme

proper setup manifold with split pulse setup

exhaust housing with split pulse

external gate

that will give you the best of everything, but it wont come cheap and it wont look factory OEM either

Not necessarily Paul, that's why I'm looking around.

From a maximum system efficiency viewpoint, the properly engineered manifold is desirable - but in the power range I've nominated we don't know if it's ESSENTIAL.

Depending if an IW split pulse unit is available economically, there may be good gains without having the big shift away from stock appearance.

FWIW, the Holset range are evidently not expensive. The model families of relevance seem to be HX30, HY30, HX35, and HY35.

More tomorrow.

[AndrewD]

I would also like to add a reply, however it may be slightly off topic.

Have you considered trying the VNT turbos from Garrett? Even though their complexity is higher, they will look standard from the outside and should integrate even easier than a twin scroll saving cost. Performance wise they are more than capable of your response and power/torque spread goals.

Also regarding those AWE tuning headers... what were they thinking! They are tuned runners by even any stretch of the imagination! Maybe the division name says it all; AWE FABRICATION! Heres why

- Runners are not equal length

- Collector design lacks... steep merge angle, 3-1 cross section is not integrated properly.

- CNC cut flanges have left lots of sharp edges for thermal stress concentrations

- My guess is they have ceramic coated it to cover up dodgy manufacturing methods.

I'll leave my rant there.

[Dale FZ1]

A few good points made.

VNT is somewhat off-track, but also something I think we'll see more of in due course. Early adopters will pay through the nose, as with all new technologies. The issue has evidently been getting the system to withstand the peak temps associated with petrol engine exhaust gases.

At the moment I'd just like to see if there is a reasonably priced split pulse setup that could be configured to suit an RB.

I don't really have an opinion on the AWE headers, but their pictures and diagrams helped with explaining exhaust gas dynamics and the how/why certain designs work better than others. FWIW, I don't believe any ceramic coating will hide shoddy workmanship or poor design. Its main objective is to work as a thermal barrier. One thing I did notice was the sandwich construction used by Audi in the OEM manifolding though.

[paulr33]

i dont think you will find a reasonably priced solution for split pulse

most guys just replace the stock turbocharger with a high rated unit (hiflow, 2530, gtrs, gt28) and get it tuned

i think the costs to sort out a proper split pulse system are out of reach of the amid fan who wants 250rwkw

[mokompri]

Here’s another link on manifolds and exhaust systems generally, that might be useful: http://www.nsxprime.com/FAQ/Miscellaneous/exhausttheory.htm

i dont agree with everything thats said on that link, specifically in regards to the 'piping size' section.

one of the major downsides of twin scroll turbine housings to me is the need for running an external gate and the complex piping thats required for an optimal setup. i spotted discopotato posting on another forum about the possibility of garrett releasing twin scroll housings if there was enough market,but what caught my interest was that he mentioned them releasing these with internal wastegates, running some sort of exhaust flap if i remember correctly. anyone have more info on these ? sounds promising...

[Devils_Advocate21]

With the IW style they run similar to the current ones at the moment although the difference being that the waste gate flap will cover 2 not 1 hole/s at the rear housing (as far as i know) disco may be able to help out here, you may have seen these before as i have seen a pic somewhere if i find one i will post is if someone doesn't beat me to it

Cheers

A

[Adriano]

Or it could have two flaps(aka serV rx-7) I would be very intersted in a comparison in the future when i look at replacing my HKS t45s with some newer technology. btw the t45s is split pulse but laggy as, i hope due to the old school technology, although its only a .90 ex housing. The HKS manifold that i (and im assuming disco)am using is cast, and not equal length i believe, although the two "groups" of cylinders seem to be close(im going off memory here, i haven't seen it in 2 years), would that significantly reduce the effect? Also do you think there would be a possibility of keeping the exhaust housing and fitting some newer technology in the housing? Im hoping for 450rwkw with an rb31 and 272deg 10.2mm lift cams with custom plenum, ported head and maybe WM injection.

[Dale FZ1]

i dont think you will find a reasonably priced solution for split pulse

most guys just replace the stock turbocharger with a high rated unit (hiflow, 2530, gtrs, gt28) and get it tuned

i think the costs to sort out a proper split pulse system are out of reach of the amid fan who wants 250rwkw

Don't get me wrong Paul. I know there are some pretty well sorted upgrades already out there. I have gone the Garrett-cored high flow myself, and satisfaction is about 95%, given the application of my Skyline. The thing is, we don't realistically know if something with the stated performance parameters CAN be found for a comparable price to the products that are popularly used. Adrian agitated for finding something that could fit up to an RB, so it's worth a look at least.

i dont agree with everything thats said on that link, specifically in regards to the 'piping size' section.

one of the major downsides of twin scroll turbine housings to me is the need for running an external gate and the complex piping thats required for an optimal setup. i spotted discopotato posting on another forum about the possibility of garrett releasing twin scroll housings if there was enough market,but what caught my interest was that he mentioned them releasing these with internal wastegates, running some sort of exhaust flap if i remember correctly. anyone have more info on these ? sounds promising...

I'm not arguing the toss on piping size. I just found the site useful to explain the dynamics of exhaust pulsing, and the impact on optimal manifold configuration.

Yes, I've seen some posts by discopotato03 on other forums canvassing internally gated split pulse interest. I'd say it's equally worthwhile seeing what is out in the market already, and trying to establish whether it is worth a close look.

The HKS manifold that i (and im assuming disco)am using is cast, and not equal length i believe, although the two "groups" of cylinders seem to be close(im going off memory here, i haven't seen it in 2 years), would that significantly reduce the effect?

Thanks for the info on those HKS cast manifolds, I've not seen one before. So they are basically an enhanced version of the factory manifold design?

I don't think we know to what extent unequal runner lengths in each branch would upset the pulsing apple cart. Hopefully someone can unearth a bit more literature, or better still some pics of what a Mitsubishi Evo manifold/turbo setup looks like, or even that of a truck. Seeing what they do might be enlightening.

[discopotato03]

Oh er something like these ?

[Adriano]

My 6.7l garret t05 truck has a split pulse manifold, and turbo, but the runners arent equal length by any stretch of the imagination, its just like a log manifold., and as far as i know the new trucks aren't any better, they just use wastegates.

[Devils_Advocate21]

Yep they be the ones disco i was thinking of, just couldnt remember where i saw them

Oh er something like these ?

[discopotato03]

They are either IHI VF 37 or 38 turbos using I think what they call the P25 turbine housing .

Somewhere out there in cyber space there is an MHI (Mitsubishi Heavy Industries) graph of a single and twin scroll turbine housing on possibly an Evo turbocharger . I didn't save that one because I don't need convincing that is has positive function .

The reason I'm pushing for Garrett to produce twin scroll divided wastegate turbine housings is that they would make the GT30/35 based BB turbos more cheaply and easily available and in a more effective format to more people .

The way I see it integral gates (and correct turbo matching has a big say in this) are not the bady many make them out to be , ask Mafia what he thinks of his .

The Japanese are famous for "bolt on" upgrades and with good reason , all the one off fabrication and customisation COSTS . When you bolt a better turbo to your factory fitted manifold and the waste gate issue is taken care of its another biggie solved . If the thing pulls great torque numbers over a wide range you lose the need to rev piss and pick handles out of the thing to get motorvating . I think if the shove in the back everyone loves starts at lower revs (and road speed) more can get their jollies without going at warp 9 . Its the pulling telephone number revs through not too many gears that puts us in gaol or worse .

I don't think warmed over road engines have to have that nothing nothing nothing WHAM power delivery but getting it right means having all its dogs barking and they don't all have to be bull mastiffs . Wanting to do it on a budgets fine but just remember your searching for 5+ litre V8 performance so it was never going to be a cake walk .

Just on the VATN turbos I don't think they're a free ride either , Variable Area Turbine Nozzle I think equates to variable turbine inlet pressure turbine housing as well so its not without its compromises .

Cheers A .

[StockyMcStock]

lol, talk about making a mountain out of a molehill. you guys have taken the massive leap of believing that split pulse turbine housings make significantly more power, or are significantly more efficient than an open housing. there certainly is a difference, but it is by no means significant.

my opinion is that split pulse will always be better (all else being equal) as you shield the interference pulses from each other a bit more effectively than on an open collector.

the downside is the angles involved in the merging gas streams are usually a bit more acute, which is not good. but shielding the pulses so that any given exhaust valve never sees a huge spike in back pressure from another valve is PROBABLY going to outweigh the gains given by a slightly more obtuse entry angle. can't say for certain.

what i can say is that the difference between two optimised designs of split pulse and open collector (all else being perfectly equal) will not be hugely significant. by simply porting the head, or using a custom manifold or larger exhaust size, you will easily outstrip the power and efficiency gains to be had from using an optimised split pulse exhaust side.

essentially if you're looking for a bolt-on turbo for an RB25 manifold, you are totally negating the benefits of using a split pulse turbo anway, since you are using the shitty stock manifold to bolt it to. the gains would not be measurable, and certainly not as cost effective as fitting a better intercooler or exhaust manifold.

[Dale FZ1]

Oh er something like these ?

Not a bad start, but what I REALLY want to see is the turbo/manifold combination. The idea was to see what Mitsubishi actually does with the manifold runner length etc, as there has been discussion about the whole pulsing / tuned length aspect.

I haven't actually seen one, or a picture of one.

Correct me if I'm wrong Adriano, but from what I've seen of typical truck manifolds with split pulse they tend to run the 2 x 3 cylinder branch approach, similar to the RB20/25 setup? Looks very industrial, but with their mass-flow requirements it must work well.

The MHI fitments to the Fuso range of trucks is something I feel is worth investigating Adrian, particularly because they are more likely than many to be offered with split pulse + compressor combinations that are not too far out of the ballpark - and availability and technical information should be accessible to some extent. I'd say it is likely a TD05 and TD06 would be very likely fitments in the smaller 4-6 litre engines, and the split pulse housing a fairly common option.

FWIW, the HX30 Holset model seems to be a reasonable starting point, IF that technical info can be obtained. Compressor flow range 45-50lb/hr, high efficiency up to PR 3.1, and turbine options available (IW, various A/R, and those titanium impellers). AND they have a T3 flange.