Compound charging

[Jesus-Ninja]

So, here's what's been kicking around my head recently. A bit of a ramble and doubtless I will re-read and edit this many times....

There's the ongoing debate about twin or single turbo upgrades. The question has been asked a million times, and the purists on both sides have made their opinions clear, but to me the arguments are as follows:

Bigger Twins - maintains the characteristic of the car as it was intended. Shorter pipework to the tubs - less lag

Big Single - Scoobs use big singles fine with longer pipework. No custom manifolds or issues with fitment. Easy to upgrade again. Only one turbo to buy.

Given the choice, I'd personally go big single. However, I've been mulling over another option. Compound charging. A bit of background.....

Turbos move volume and compress by ratios. So, a turbo that makes 1 bar of boost is operating at 1:2 (atmostpheric being 1 bar) ie it's taking 1 bar of atmospheric pressure and giving 2 X atmostpheric pressure - and absolute pressure of 2 bar, which is 1 bar of boost. If that turbo had an inlet pressure of 2 bar absolute, then at the same spool would still give 1:2, but the boost it would produce would be 3 bar - ie it takes the inlet of 2 bar absolute, double that which is 4 bar absolute, or 3 bar boost.

This all sounds too much to be true, and in a way it is, for a numbr of reasons, the main one being the heat produced at each stage of charge also increases exponentially. It's generally the preserve of big ass tractors and sled pulling diesels where manifold pressures of 100+PSI are needed. However I don't know that it's entirely without merit on a petrol engine. It has been done with varying degrees of success, using one turbo to feed another as well as using combinations of superchargers and turbo chargers.

It's worth noting that this is not the same as sequential charging. That requires various combinationos of controlled valves etc on the exhaust and inlet systems to remove one (big) turbo from the system whilst the smaller one does the work, and then before the small one runs out of puff, the big one is allowed to spool and ultimately takes over. But these do not typically blow through each other.

A compound setup, in it's simplest form, has a small turbo mounted traditionally on a manifold, feeding it's charge into the plenum (we'll ignore intercoolers for now). The inlet to the small turbo is fed from the charge from a bigger turbo. The bigger turbo receives exhaust from the small turbo.

So, where am I going with this?

If you look at the boost curve of our twins, they make about 18 or 19 PSI and then boost steadily tails off over the rev range. What if we could keep that early spool and and hold the boost to the read line. Maybe even increase the boost and have a steady increase?

What you'd need is a charger before the twins that would raise the pressure that they "see" and it wouldn't have to be by much - the main thing is that whatever is producing the initial charge needs to flow well, even if it's not providing silly boost. For this reason using another TD03 probably isn't going to help, as although it doesn't need to provide bags of boost, it does need to let enough volume through. One interesting notion, as an aside, was to compound charge the existing twins with each other - if the curve boosts to 18 PSI and the trails off to 13PSI as twins, then compounding one into the other would see boost hitting 26PSI before trailing off to 16PSI, but as I say, I'm pretty sure that flow would be the issue.

So what would you put infront of the twins - well one idea is a centrigual supercharger. These seem to be only selected in the event that tubos or a positive displacement supercharger won't fit, but have some merit in that they produce boost which rises linearly with rpm - exactly what we want. At lower rpm the multiplication of boost will be slight, maintaining the cars chracteristic, but it would then give more lift to the stock twins at higher RPM. Given the same curve that boosts to 18 PSI and the trails off to 13PSI, overlaying a centrifugal supercharger that, on it's own, only makes 5PSI at 7000RPM, the resulting boost curve would hit 23PSI and hold a steady 22PSI to the red line. The problem is that superchargers incur significant parasitic power losses to just spin them, compared with a turbo which is "almost" free. That said, centrifugal superchargers are the least parasitic of superchargers, as far as I can tell. The benefit though is that provided it physically fits somewhere, a centrifugal supercharger is as easy to plumb into the inlet as a turbo, but without the need to manage it's boost and no exhaust plumbing.

The other option of course is a third turbo running off the exhaust gasses from the twins. Boost would probably have to be managed independently, monitoring it across the big turbo. There are plumbing questions around whether to use just the wastegate gasses from the small turbos or all of it. I would suspect you'd want the lot, even though some of it has been "spent", as you'd potentially get into difficult situations where you may fall into a valley of death where the twins trail off and fall below their preset boost level (wastegates close) and so no gasses get to the big turbo which doesn't then spool.

Finally, there's an overall issue of how would you control boost on the twins? The whole system relies on the multiplying effect of twin charging, but because it's compounded the control of the waste gates on the twins isn't really based on the total boost above atmospheric, it needs to be based on the pressure on the outlet of the turbos compared with their inlet.

It might be possible to do it arse about face, and have the twins feed thair air into the single charger - their boost would be easy to control then, but they'd be expected to flow more as their outlet is being compressed by the next charger, and I suspect they'd be too restrictive.

This is all conjecture as well, of course. And there are very good reasons why people don't do this, but if you wanted to maintain the characteristics of the car but add top end power, then this might be of interest, and would be more bolt on (certainly the Centrifugal SC would be) than either big single or big twins.

I've been mulling it over so I thought I'd present it to the wide CLubVR4 to pick holes in it and shoot down my balloon before I go too high in it! :)

52665

[foxdie]

Using a combo of twin turbochargers and single supercharger has been done before on a 3000GT, including comments on how the setup performed compared to other solutions:


http://www.youtube.com/watch?v=bCocbmg34mQ

Although notably, they're using turbos to charge the supercharger, which charges the engine.

[Jesus-Ninja]

Interesting that they are using a roots blower, which are quite costly to run - I think they sap something like around 100HP to run, and so need to return a lot of power. EG A 300BHP engine which then gets a SC that draws 100HP to run, needs to make an extra 100HP just to break even, and then of course you want a decent power hike as well, so you're looking for an extra 200HP from the system to make a net gain of 100HP, hence lots of injectors! As you say, interesting that they're boosting the SC with the turbos, and not the other way round.

[Jesus-Ninja]

It's a good video though. Basically saying what I was trying to, although they explain it better! And saying "Don't bother", or certainly not with a roots blower anyway. The parasitic losses high up are the issue.

Still, maybe a third turbo? But then the plumbing isn't much different to just going single turbo.

[wintertidenz]

I saw a forum thread about this being done on an Eclipse - they used a big ass turbo to compress the air into the smaller one.
What they did was feed the gasses out of the small turbo into the larger one, so as the smaller turbo closed the wastegate and was at fully spooled, the larger turbo would start to spool up.

I've thought about this for a while on our cars - we could use something a bit larger, like a TD05, to feed the smaller turbos and still have some extra boost at the top end.
You could regulate this turbo off an EBC or even just wastegate spring pressure - remember once the large turbo is fully spooled, it will be putting around 10 PSI into the small turbos, which then compress that again to 20 PSI from each turbo (based on 10 PSI set on them).

The piping will be a pain in the ass, and even more so for auto owners as there is a big transmission in the way, but I think it would be worth it. Aside from the extra power you would gain, it is also unusual and quite unique - who expects to see three turbos under the bonnet?!

Here's the link: http://www.dsmtuners.com/forums/custom-fabrication/336541-my-compound-turbo-set-up.html

[scott.mohekey]

This is something I've been thinking about also. I've seen it done quite successfully on I6 engines which obviously have a bit more space in the engine bay to achieve it. The biggest problem would be the piping, but I think if you turn the stock turbos around, the exhaust outlet can then very easily be routed to just above the transmission. It's then a matter of routing the output from the larger turbo into the two smaller ones.

[Kenneth]

I have been thinking about compound turbo charging for a couple of years and plan to eventually implement it using 3 turbos. 1 off each bank and then one big one above the gearbox.

That video looks to me to be dubious. It appears they have the wrong supercharger for what they are trying to do and so are over working it which is causing issues. Indications are the air temp problems they had (at Idle) and the fact they had lots of belt trouble.

[Jesus-Ninja]

If you used the stock turbos in the system you'd certainly need them to be closest to the engine in terms of charge. They would struggle to flow enough volume if they were charging another turbo, but a big lazy turbo that makes a bar of boost could double the boosr that the twins can pump.

Ken, I agree with you on their use of SC. A roots or twinscrew draws too much power, hence their losses. A centri or tub might still have mileage though. I'm struggling to find any clues as to how much more efficient centris are than PD, although tubs are the most efficient.

Sent from my GT-I9100 using Tapatalk 2

[Jesus-Ninja]

Scott - the piping should be no harder than a single big tub. The only minor niggle is where to draw oil lines from.

Sent from my GT-I9100 using Tapatalk 2

[Jesus-Ninja]

Wintertidenz - it doesn't quite work that way. If the first and second chargers can both make 10psi, it doesn't equate to a compoumd charge of 20psi. Both are operating at 1:1.66, so you get 1:2.75 which is 26psi. That's the beauty of it!

I've seen that eclipse before though, and he seemed to get some good results.

Sent from my GT-I9100 using Tapatalk 2

[foxdie]

I suppose the next question is, does someone have a centrifugal supercharger lying about they can temporarily bolt onto a 6A13TT and dyno it? :D

[scott.mohekey]

I have been thinking about compound turbo charging for a couple of years and plan to eventually implement it using 3 turbos. 1 off each bank and then one big one above the gearbox.

That's pretty much exactly what I was thinking. Do you have any concerns about the balance of the compressor outlet from the large turbo to the compressor inlets of the two smaller turbos?

[scott.mohekey]

I suppose the next question is, does someone have a centrifugal supercharger lying about they can temporarily bolt onto a 6A13TT and dyno it? :D

Funny you should say that. I think my flatmate has one from a toyota of some sort that he was going to put on his honda.

[Kenneth]

No. Is pretty much the same concept as our current turbos except that the inlet to the 2 smaller turbos will be above atmospheric pressure when the large turbo is producing boost.


That's pretty much exactly what I was thinking. Do you have any concerns about the balance of the compressor outlet from the large turbo to the compressor inlets of the two smaller turbos?

[scott.mohekey]

Ah of course. Were you going to use the stock turbos off each bank?

[Kenneth]

No, though they should be adequate really.

[scott.mohekey]

Ok, I've decided to do this. I'm going to build up a forged block, and use the ported heads with the uprated cams from adam's group buy. I'll be using the stock turbos flipped so their exhaust outlets point towards the gearbox with a larger turbo (undecided what yet) sitting above the gearbox with its exhaust outlet pointing towards the rear of the engine bay. The outlet from the larger turbo's compressor will be split above the gearbox and route infront of and behind the engine to the inlets of the stock turbos. The stock turbo's compressor outputs will point down and be joined below and to the drivers side of the front one, feeding a large FMIC. The outlet of the FMIC will then come up over the gearbox close to the passenger's side rail, before entering the throttle body.

The manifolds will be custom made from mild steel, with external waste gates bypassing the stock turbos to the larger turbo so as not to push the smaller turbos too far. The larger turbo will also have a wastegate to manage overall boost.

One thing I'm not sure about just yet, is how to manage the waste gates. I've seen other people have the waste gates on the primary(s) set to an arbitrary total boost level, and the waste gate for the secondary (larger turbo) controlled with manual/electronic boost control. But I'm wondering if it might be possible to get more finely tuned results by having both sets of waste gates electronically controlled independently.

[Kenneth]

The challenge with wastegating is going to be keeping a pressure ratio rather than absolute boost on the small turbos.

As such, you really need to be able to pressure each side of the wastegate diaphragm (as you can with an external wastegate) so that inlet pressure forces the wastegate shut against the manifold pressure trying to open it.

Electronic wastegates would be ideal, but you need to have another MAP sensor in the inlet for the smaller turbos so that you can calculate pressure ratio.

You are also going to want to think about how much exhaust gas you need to bypass on the small turbos. The internal wastegate might not be sufficiently big enough to stop boost creep.

[scott.mohekey]

I was thinking of welding the internal waste gates on the stock turbos, and using larger external ones for them. If I use mechanical waste gates and take my pressure measure from after each of the turbos, would that suffice? (I have it in my head that the pressure after the primary would control the waste gate for the primary, but I'm not sure if using the pressure after the secondary to control the waste gate of the secondary is right or not).

[SEAN-NZ]

in regards to pressure regulation, how about a something similar to what we use, but before the small turbos, to control the big turbo, then just another at the manifold as we have it now for the little turbos, to me seems fairly easy in theory, although would likely prove problematic

[scott.mohekey]

I'm not sure I follow?

[Jesus-Ninja]

then just another at the manifold as we have it now for the little turbos, to me seems fairly easy in theory, although would likely prove problematic

Wouldn't work. To keep the smaller tubs working in their efficiency range, you need to control their waste gates by calculating the ratio of absolute pressures across the compressors, not as relative boost above atmospheric.

When the bigger single is producing no boost, then you'd want the little tubs' wastegates to open at a manifold boost of, say, 1 bar. But when the big single presents the little tubs with one bar then you want the little turbos' wastegates to open at a manifold boost of 3 bar.

[scott.mohekey]

Seems to me the best idea would be to have waste gates that will open at the lowest pressure needed, and electronically control them so you can have them opening at higher pressures.

[Jesus-Ninja]

Funny you should say that. I think my flatmate has one from a toyota of some sort that he was going to put on his honda.

Doubt it's centrifugal, unless it was fitted aftermarket to the toyota. AFAIK centrifugals aren't generally fitted to production vehicles for a number a reasons, the main one being that if you were starting from scratch, it's not a good design. Turbos and PD are better methods of forced induction.

They're popular with V-Techs, because their linear boost increase with rpm matches nicely with the power delivery of the honda lump. V-Tech engines are gutless below about 5000RPM and so would be ruined with, say, a roots blower. However, they come on song later in the rev range, just as the centri starts to sap power.

That said, I understand that centris are generally more efficient than PD in terms of parasitic loss.

Centris are also easy to install which makes them popular for aftermarket fitment, but because none are fitted by manufacturers, it means there isn't a glut of them lying around scrap yards.

[scott.mohekey]

I'll have to ask him what it is when he gets back on Monday.

[Jesus-Ninja]

No, though they should be adequate really.

They'd be as adequate as the turbo that's feeding them :) They flow volume so, in principle, if they can comfortably compress atmospheric (1 bar) air to 2 bar (ie 1 bar boost), then if you fed them 3 bar instead of 1 bar, they'd compress that 3 bar to 6 bar (ie 5 bar boost) - however heat management appears to become the issue, so either a massive intercooler after everything, or possibly even an intercooler between the single and the twins and then another after the twins. Packaging becomes complicated. Air>Water>Air charge coolers?

[scott.mohekey]

Something like this maybe, if you don't want to go down the route of water cooling:

52667

[Jesus-Ninja]

Definitely something like that :) Provided that the two separate coolers are adequate.

I'm thinking that for a mild setup using stock twins, that you want a single to feed them that flows really well but doesn't make mental boost.

Here's what would, in theory, happen if you ran the twins to produce about a bar of boost on their own (peaking at just over a bar and then trailing off, so not pushing them hard), and then added a bigger single that only made 10psi

52668

[Jesus-Ninja]

Funny you should say that. I think my flatmate has one from a toyota of some sort that he was going to put on his honda.Doubt it's centrifugal

I tell a lie - seems they used one on the Scion. http://www.motortrend.com/features/performance/112_0808_new_toyota_trd_superchargers/photo_05.html#photo

[scott.mohekey]

What if we were to use a differential pressure sensor between the inlet and outlet of the smaller turbo's compressor. This would give us the pressure ratio currently being generated and could be used to control the smaller turbo's waste gate.

[Jesus-Ninja]

What if we were to use a differential pressure sensor between the inlet and outlet of the smaller turbo's compressor. This would give us the pressure ratio currently being generated and could be used to control the smaller turbo's waste gate.

It would give you the difference, but you want the factor. Eg with no boost from the big one, the difference for a bar of boost on the little one(s) would be 1 bar, but at the same ratio with the big one producing one bar, the difference would be 2 bar. You'd need something clever to electronically work out when the ratio is 1:2, unless there is some other mechanical way of doing it that my brain can't compute :)

Sent from my GT-I9100 using Tapatalk 2

[scott.mohekey]

Ok, so three pressure sensors. One before the turbos (there's actually already one in the MAF we could use), one between the turbos, and one after the turbos.

The difference between the first and second gives us the pressure ratio of the larger turbo, the difference between the second and third the pressure ratio of the smaller turbos, and the difference between the first and third, the overall pressure ratio.

[foxdie]

Ok, so three pressure sensors. One before the turbos (there's actually already one in the MAF we could use), one between the turbos, and one after the turbos.

The difference between the first and second gives us the pressure ratio of the larger turbo, the difference between the second and third the pressure ratio of the smaller turbos, and the difference between the first and third, the overall pressure ratio.

I'm thinking... an Arduino could be used to monitor the 3 MAP sensors via AnalogIn, control the duty cycle on multiple wastegates via PWMOut through a Transistor / Darlington Array, and then perform some calculations and either output a MAF emulation signal (PWMOut again) or a MAP sensor signal (AnalogOut) to connect to an ECU ;)

[scott.mohekey]

I'm thinking... an Arduino could be used to monitor the 3 MAP sensors via AnalogIn, control the duty cycle on multiple wastegates via PWMOut through a Transistor / Darlington Array, and then perform some calculations and either output a MAF emulation signal (PWMOut again) or a MAP sensor signal (AnalogOut) to connect to an ECU ;)

Kinda what I was thinking.

[foxdie]

Kinda what I was thinking.

Stop stealing my thoughts!

http://i619.photobucket.com/albums/tt271/neuralstate/interwebs/get_out_of_my_head_charles.jpg

/haz

[scott.mohekey]

/haz

[wintertidenz]

Okay, assuming you get it all working with separate boost controllers, or use wastegate spring pressure... will the internals of our engines hold up to 30 psi?

I know you are going for forged bits Scott, but there is still that factor of potentially scattering engine parts all over the place once the big turbo kicks in.

[scott.mohekey]

The big turbo doesn't kick in.. it gradually spools up as it normally would. You just have the advantage of already having boost from the smaller one. If you size the turbos right you get a nice smooth transition.

[foxdie]

Okay, assuming you get it all working with separate boost controllers, or use wastegate spring pressure... will the internals of our engines hold up to 30 psi?

I was always told above 1.5 bar (22 PSI) was Kenny Loggins (http://www.youtube.com/watch?v=siwpn14IE7E&feature=artist) (Danger Zone).

Why would we want that high boost anyway to start with? Surely just being able to hold 1.2 bar boost with relatively colder air through the rev range would be a far better for power top end?

[scott.mohekey]

Initially yes. But where's the fun in stopping there?

[foxdie]

Initially yes. But where's the fun in stopping there?

I didn't say stop there, I just think we should keep it realistic and set goals ;)

First goal is to see if charging turbos with a centrifugal supercharger at least breaks even :)

[wintertidenz]

To be honest I think that just going for the third turbocharger might just be the easier and quicker way to go.

[scott.mohekey]

To be honest I think that just going for the third turbocharger might just be the easier and quicker way to go.

Yep, no way am I bothering with a supercharger.

[thfelipeth]

im interested to see where this is going. ignore me

[Jesus-Ninja]

To be honest I think that just going for the third turbocharger might just be the easier and quicker way to go.

It depends. Given a choice of either, a tub is better, I think, in terms of efficiency. But a centrifugal charger can be fitted as easily as a turbo but without any changes to the exhaust system and it requires no boost control. The only thing it needs is a pulley.

[Jesus-Ninja]

The difference between the first and second gives us the pressure ratio of the larger turbo, the difference between the second and third the pressure ratio of the smaller turbos, and the difference between the first and third, the overall pressure ratio.

No.

The difference doesn't give you the ratio. It simply tells you the difference. If your plan is to run the two pressures across a dual port wastegate, then it won't work. As the two pressures operate against each other you'll simply get the difference. Let me explain:

To make the numbers easy, we'll say that our twins are most efficient at 1:2. That is to say that with an inlet pressure of 1 bat (ie atmospheric pressure, NOT 1 bar boost), they will produce 2 bar absolute pressure (ie 1 bar of boost above atmospheric which is 1 bar)

If the big turbo hasn't spooled yet and the twins are on song, then the difference will be 2 bar absolute - 1 bar (Atmospheric) = 1 bar.

Lets now say that our big feeder has run up to 1 bar boost (ie 2 bar absolute) which the twins are now seeing. For them to still be in their same efficiency we want them to still run at 1:2. So they'll be taking that 2 bar absolute from the single and making that into 4 bar absolute - this will be a difference of 2 bar across the twins compressors and a difference of 3 bar between the plenum and atmospheric.

So - what you need is some electrickery that will read the pressure before the twins and after as absolute values and divide one into the other and make a decision as to whether to open a solenoid that will bleed boost off to the wastegates.

IF it was the other way round and the twins fed a third charger, AND that was a supercharger, then life would be much easier: Boost on the twins would be easy to regulate, as the inlet is ALWAYS atomspheric and you could operate the wastegate on pre SC pressure, and a supercharger doesn't need regulating in the same way.

[Jesus-Ninja]

I'm thinking... an Arduino could be used to monitor the 3 MAP sensors via AnalogIn, control the duty cycle on multiple wastegates via PWMOut through a Transistor / Darlington Array, and then perform some calculations and either output a MAF emulation signal (PWMOut again) or a MAP sensor signal (AnalogOut) to connect to an ECU ;)

You'd definitely want an ecu to control it /yes

AKAIK the diesel setups only have a wastegate on one tub.

[scott.mohekey]

The only thing it needs is a pulley.

And a place to put it.

[scott.mohekey]

No.

The difference doesn't give you the ratio. It simply tells you the difference. If your plan is to run the two pressures across a dual port wastegate, then it won't work. As the two pressures operate against each other you'll simply get the difference. Let me explain:

Ok, so my terminology was wrong, but with the three sensors as I described, and an ecu of some sort doing the calculations you mentioned and controlling the waste gates, I think it will work.

[Jesus-Ninja]

And a place to put it.

Very true :)

In the place of the aircon pump?

[Jesus-Ninja]

with the three sensors as I described, and an ecu of some sort doing the calculations you mentioned and controlling the waste gates, I think it will work.

Yes :)

[Jesus-Ninja]

Interesting read: http://www.yellowbullet.com/forum/showthread.php?t=216811

[scott.mohekey]

That certainly makes the waste gate control sound a lot simpler.

[Kenneth]

There is no reason why you can't get pneumatic control of the wastegate and chances are it'll be a lot easier to manage. You can get fuel regulators which use a raising rate factor greater than 1, a wastegate is the same principle just slightly different in size and actuation.

alternatively you can use a standard external and control both sides of the diaphragm. While this will not exactly give pressure ratio control, you can still manage total boost ok. The thing with this setup is that as the big turbo spools up, it will start taking up more and more of the total compression requirements. Which, with the turbos available today, could be a good idea.

[roryfromnz]

This is an awesome topic of conversation. Im sure i had seen a Toyota 4agze 1.6l engine "Twin Charged" and running well over 300hp at the wheels.. The did have some tuning issues but perhaps the small toyota supercharger might be a fun place to start if going for the supercharger route. The were running the turbo into the charger im pretty sure..

[ILLEGAL]

single GT35R.. problem solved.. LOL grey hairs avoided.

[lancerevo3]

subscribed

[scott.mohekey]

So, I'm thinking one of the GT37 range of turbos from Garrett. I haven't really looked at the compressor maps etc yet to see if they're ideal or not, I'm mainly basing this on them being the smallest Garrett turbo with a twin scroll exhaust inlet, which I think would be ideal for the fact I'm going to be running the exhaust from the two smaller turbos into the larger turbo.

[Kenneth]

My understanding is that split pulse is to reduce cylinder interference (pulse interference from other cylinders) which reduces turbine efficiency, but usually comes at the cost of peak flow

This being the case, you would not need (or probably want) a split pulse exhaust housing.

[scott.mohekey]

Yeah, my thinking was to make it a bit easier to merge the two flows. I guess the back pressure caused by the turbine wheels on the smaller turbos is enough to remove the cylinder pulses?

[SteveH]

Not sure this is any use to you guys?? http://www.ebay.co.uk/itm/Blitz-Compressor-Supercharger-Kit-Toyota-Celica-ZZT-231-TRD-Vvtli-190-/130688812674?pt=UK_CarsParts_Vehicles_CarParts_SM&hash=item1e6da90682

[foxdie]

I'm possibly gonna be in the market for bigger power soon, can someone actually put theory to practice? ;)

[SEAN-NZ]

why dont you put it into practice? im guessing it hasnt been done on these engines, would make a fairly unique car, i would do it, too much other stuff to pay for though haha

[wintertidenz]

Scott is currently working on it. There's a lot of work involved though.

[UPPCOS]

Hey, this is CK from UPP, the owner of the "Triple Charged" VR4.

I've done some extensive testing with the compound setups, and I've been invited here to discuss what I know with you guys. After cruising through the posts a bit here I'll share with you a few tid-bits that I've learned.

#1, the largest pump MUST be first in the system. IE, it was asked why I fed the turbos to the SC, it's the same reason you use turbos to feed the engine. Because the turbos flow more than the engine can. Imagine if you used only one stock turbo to feed your engine. It couldn't do it, it would become a restriction, and inhibit the flow of the engine. This is because they have "choke" limit on them, and can only flow so many CUBEs of air through them. However, they don't care what the density of the "CUBEs" are that they flow. To better explain this, your engine is a 3.0L (183ci). So it can only flow 183ci worth of air through it. But if you compress the air before it enters to 14.7psi (1bar), flowing the same amount of "cubes", it will now process the volume of air equivalency of 366ci. Now with compounding the same principal applies. The smaller (SC in this case) doesn't care what the density of the air is that it's passing, it's only passing "cubes" as it sees it. However the system as a whole will never flow more than the first pump in the system. This is why your turbos die off after X-rpm, and you stop making power. This is why the SC was second in the system, after the turbos. (BTW they were TD05-16G's (for sale w/DNP manifolds), and only an Eaton M90).

#2 I used a Eaton PD SC because the provide the absolute best in low end boost. Stomp the gas and the boost is already there. This isn't the case with centrifugal SC's like everyone assumes it to be. Centri's provide a linear line of boost. Meaning if someone has says they're pushing 8psi on their Centri, they don't actually see 8psi until right at REDLINE! Let's assume that they're pushing 8psi on an 8K rpm engine. That's roughly 1psi per 1k rpm. So at 3K rpm's, it's showing only 3psi.....whoop-de-do! By 3K my turbos are starting to wake up. Therefor for compounding, centuri's are really pointless. The PD Eaton, worked well, it did exactly what it was supposed to. It would spike 10-12 initially, which would ramp the turbos up quickly, and 28+psi by 2k rpms. But like the video said, 28psi @2k rpms, really isn't as "fast" as it seems it should be. With the reversion on the cam profile, and the lack of RPMs, it's still doesn't produce a million hp (maybe with vtech it would!). It did produce more power down low, everything before 4K rpms, so for a diesel setup this would be ideal. You'd get the instant response of the SC with closer to the efficiency of the turbo. But as someone mentioned the heat and parasitic loss of the PD blowers, the efficiency lacks on the high end. The only way to properly do this is with a clutched SC pulley, with bypass valves (it's been tried and it's even more $$ and a pain in the @$$).

#3 This leaves compounding with 2x turbos. This is a sound technology and it does work. But the problem here is in the complexity of routing everything, and the proper sizing of the turbos. I'll go to extreme to demonstrate this, but imagine you're going to run 1x TD04 turbo for the bottom end, and a T76 for the top. The engine will spool the little TD04 instantly, which causes the WG to open (lets say 6psi), which dumps into the 2nd turbo (along with all the exhaust from the first turbo), but this still isn't enough to get the giant T76 spooling. So what happens is the car boost instantly to 6psi, then flat-lines until the revs are high enough to get the T76 actually spinning. So your car accelerates, flatlines, then takes off again. Now you could increase the initial W/G on the primary turbo, but this may make matters worse. You'll spool 14psi instantly, but if that's still not enough to jump-start the T76, the TD04 will soon run out of umph, and taper off, before the T76 has a chance to spool. Now your power goes up, then down, then back up. So the smooth pulling acceleration everyone is looking for, isn't there... Now at this point you have two options. Increase the size of the TD04, which takes away your low end and spool time, or decrease the T76, which takes away from your high end. Either way you're starting to be counter productive towards what you were initially working towards. Now don't get me wrong, it can/does work, but to what degree that everyone is expecting is questionable. And for the $ and complexity there's many other options that will do the same or better.

#4 The other options. First and foremost, Nitrous! In the video I overlayed the map with the roots blower (which no turbo will spool faster than) and the nitrous STILL beat it hands down, no contest. And the nitrous only really sprayed for about a second (just to spool the turbo), so my bottles last forever. I'm not a big nitrous junky, but for spooling a turbo....amazing! Aside from that there's anti-lag, and now we're working with rolling anti-lag. I can build full boost while sitting still at the stop light before I ever take off, so no need for a SC anymore. Just retards the piss outa the timing and generates stupid unuseable amounts of boost before we even launch. Or if you want to do a roll-on "highway" race. No, problem, works there too....Rolling anti-lag(tm), just get set going the speed you want to start from (ie 40 roll), press the button and smash the gas. The boost builds, but the car doesn't accelerate. When you're ready to run, just let go of the button, and Ape $#it!!! It goes crazy!

So in conclusion, it does work, but the efficiency of it, and how practical it is, is debatable. But my hats off to anyone who is attempting this in the pursuit of knowledge and finding out "what if". I would be more than happy to help and add in my .02 for what it's worth. If nothing else, it makes funny noises, and gets ALOT of attention!!

http://www.youtube.com/watch?v=4PriA0WT-08
(filmed in mexico)


CK

[UPPCOS]

I made a post in here yesterday, fairly lengthy, but it said it was "waiting on mod approval"? Can one of you mods check into this please as I really don't want to have to retype the whole thing....


CK

[foxdie]

I made a post in here yesterday, fairly lengthy, but it said it was "waiting on mod approval"? Can one of you mods check into this please as I really don't want to have to retype the whole thing....


CK

I'll vouch for CK as I invited him to comment on this post :)

/CC BraindG Subaru ETA

[BraindG]

Done.

edit: i have moved you to registered user, thus not impacted by the newbie posting requirements.

[foxdie]

Hey, this is CK from UPP, the owner of the "Triple Charged" VR4.

I've done some extensive testing with the compound setups, and I've been invited here to discuss what I know with you guys. After cruising through the posts a bit here I'll share with you a few tid-bits that I've learned.

Hi CK, thanks for accepting my invite. For the others reference, there are a couple of Youtube vids that CK has published regarding triple-charging;

http://youtu.be/wsUWt5x_6dc
http://youtu.be/bCocbmg34mQ

CK, after reading your post (thanks for taking the time to write that), I'd like to pose some more questions and clear up some issues on why our VR-4 application is a little different to yours, and why we're trying to ascertain the best course of action to try triple-charging a Galant / Legnum VR-4. These may come across as slightly inflammatory or provocative but that is not my intention :)


#1, the largest pump MUST be first in the system. IE, it was asked why I fed the turbos to the SC, it's the same reason you use turbos to feed the engine. Because the turbos flow more than the engine can. Imagine if you used only one stock turbo to feed your engine. It couldn't do it, it would become a restriction, and inhibit the flow of the engine. This is because they have "choke" limit on them, and can only flow so many CUBEs of air through them. However, they don't care what the density of the "CUBEs" are that they flow. To better explain this, your engine is a 3.0L (183ci). So it can only flow 183ci worth of air through it. But if you compress the air before it enters to 14.7psi (1bar), flowing the same amount of "cubes", it will now process the volume of air equivalency of 366ci. Now with compounding the same principal applies. The smaller (SC in this case) doesn't care what the density of the air is that it's passing, it's only passing "cubes" as it sees it. However the system as a whole will never flow more than the first pump in the system. This is why your turbos die off after X-rpm, and you stop making power. This is why the SC was second in the system, after the turbos. (BTW they were TD05-16G's (for sale w/DNP manifolds), and only an Eaton M90).

Okay, that makes sense, but whatever is next in the system, wouldn't that become a bottleneck? Both are pumps, both have to isolate the system. If the compressed output of a turbocharger (or two) is fed into the intake of a supercharger, whilst the supercharger isn't running in a zone of producing measurable power, is it not restricting the compressed airflow from the turbos?

Same situation in reverse I guess, supercharger's compressed airflow passing through slowly spinning turbos could be highly restrictive (does feeding compressed air into a turbo compressor actually spool it up such as feeding hot air into the turbine?).

Also, for ref, 99% of us here are using the 6A13TT (2498cc / 152.4ci) with a compression ratio of 8.5:1 and 2 x TD03s set up as with the GTO/3000GT twin turbo setup. What we're trying to achieve is triple-charging these TD03s as there isn't a lot of room in our engine bays for anything much bigger (plus there are no off-the-shelf manifold kits like with the 6G72), simply switching to the TD04-13Gs for example would cost in the region of $5000-6000 US, whereas fitting a supercharger could probably be done for a third to a half of that cost.

The flow profile for our TD03s is poor, sure we come on boost stupidly quickly but after 4500 RPM they start to run out of puff, by 5500 RPM they've dropped from pushing 1.25 bar / 18 PSI of boost to 0.8 bar / 11 PSI of boost. I am rinsing the crap out of my TD03 turbos at the moment and made 340 HP at the fly with 380 FtLbs of torque, yet our engines are good for up to 450-500 HP (and similar levels of torque) before internals need to be upgraded.

So am I right in guessing you recommend we do something like replace the intake manifold with one fitted with a supercharger, compressing the output of the TD03s above 4500 RPM or so? Or have I got it the wrong way around?


#2 I used a Eaton PD SC because the provide the absolute best in low end boost. Stomp the gas and the boost is already there. This isn't the case with centrifugal SC's like everyone assumes it to be.

Okay so we avoid centrifugal SCs. Got it.


The PD Eaton, worked well, it did exactly what it was supposed to. It would spike 10-12 initially, which would ramp the turbos up quickly, and 28+psi by 2k rpms.

This confuses me and makes it sound like the SC is first in the system, could you confirm?


The only way to properly do this is with a clutched SC pulley, with bypass valves (it's been tried and it's even more $$ and a pain in the @$$).

Okay, is there such a thing as an auto-clutched pulley for superchargers that starts to transfer torque through to it's spindle when the input speed goes above a certain threshold?

I'm thinking in basic terms here, mopeds and most go-karts have a system where the moment the motor starts to spin up, arms inside a centrifugal pulley start to engage and move the vehicle forward, could such a system be employed?

Regarding the pipework, yes, that's a complexity, but not sure how expensive this would be, control system aside (read ECU developments below) I can personally picture using solenoid controlled exhaust cut-out valves to direct airflow. It's a mess and the sudden changes in airflow could cause pressure / flow issues, but the theory is there :) thoughts?


Aside from that there's anti-lag, and now we're working with rolling anti-lag. I can build full boost while sitting still at the stop light before I ever take off, so no need for a SC anymore. Just retards the piss outa the timing and generates stupid unuseable amounts of boost before we even launch. Or if you want to do a roll-on "highway" race. No, problem, works there too....Rolling anti-lag(tm), just get set going the speed you want to start from (ie 40 roll), press the button and smash the gas. The boost builds, but the car doesn't accelerate. When you're ready to run, just let go of the button, and Ape $#it!!! It goes crazy!

There's quite a lot of ECU development going on at the moment for our cars, I've been helping people such as Kenneth, BCX, Shtiv and many others with their development work. One thing Kenneth did recently was release switchable maps, 8 maps each able to pull from a pool of 8 unique tables for fuel, timing advance, target boost, wastegate duty cycle etc.

One thing I did with those switchable maps was implement both stationary and rolling anti-lag. I have to say, rolling anti-lag is too dangerous for your average user as it can easily build extreme levels of boost, I've seen 1.8 bar (26 PSI) of boost whilst doing 30 MPH in 3rd gear from our paltry TD03s :) It does take off like lightning though afterwards!

Other developments in the works are user-programmable outputs which could be used to control solenoids / valves. This opens many doors for us such as changing airflow routing depending on engine RPM or load, we could look at sequential turbocharging or using valve control to skate around the issue of a superchargers power drain :)


So in conclusion, it does work, but the efficiency of it, and how practical it is, is debatable. But my hats off to anyone who is attempting this in the pursuit of knowledge and finding out "what if". I would be more than happy to help and add in my .02 for what it's worth. If nothing else, it makes funny noises, and gets ALOT of attention!!

That's pretty much what we're doing right now, the pursuit of "is this viable?" :)

Regardless, thank you for your time CK, hope you hang around a while longer :)

[Jesus-Ninja]

To step in in anticipation of the reply to the first question about bottle necks: That would be the whole point of tripple charging a 6A13TT (on the assumption that the original twins are left in the system as the second) - they wouldn't be a bottle next because you're not asking them to flow any more volume. They'd flow exactly the same volume as they do normally, but that volume is denser. Hence the comment that the bigger tub needs to be fist, as that will be flowing the same mass of air as the twins (which are second in parallel), but at a lower pressure, hence more volume.

[SEAN-NZ]

just to revive this, as i was thinking about it tonight, could we have our big turbos with to start with a .5 bar wastegate that could get its boost pressure straight after the turbo, then assuming the first pic is how our current wastgates work
http://www.turbov6camaro.com/gate.htm
could we plumb the open air to piping before little twins? to produce a further .5 above turbo intake pressure, so that would be 14.7*1.5=22 22*1.5=33, so itl be boosting 18psi, or leave our current wastegates as they are and have an ebc set to 18psi? i guess ebc would be easier to work with, hope what ive tried to say is easy enough to understand, if it could work this way, it could make it a lot easier to set up, only plumbing being an issue, i could try do up a pic if asked for, but im far from an artist haha

[Davezj]

you should be able to boost to 18psi an EBC fitted without any mods to standard turbos, you just have to take care of the knock before running tthe boost hat high. when i say just take care of the knock it is far to simply put but in essence that is all you have to do.
our standard turbos will run about 20psi if you ask them to, but that is not advised.

[Nick Mann]

18 psi is one thing, but 18psi at low temperature is another. The general feeling is that our tubs are fast leaving their efficient range at 14psi The difference in 18-20psi on the stock set up is negligible, due to the extra heat build up by the tubs in making that extra pressure. The actual mass of air doesn't increase much and the higher temps means you are more likely to get knock.

The whole point of attempting compound charging is to achieve efficient boost isn't it? It's the main reason for going to bigger tubs - a good sized pair of TD04s at 5k revs and 18psi should be producing significantly more power than stock tubs at the same revs and boost.

I'm not convinced by compound charging, TBH. I suspect the investment and headache would be more than just changing turbos to something more appropriate to your aims.

[Davezj]

i see what you mean about usable boost as an example,
18psi at inlet temp 40deg C is much better than 18psi at 80 deg C.
as the you are physically getting more air into the engine due to the lower temp.

[scott.mohekey]

This is one of the main benefits of compound charging; high boost levels at workable temperatures. The other benefit, is that you get the lower spool times of the smaller turbo(s). I'll agree, if you're not going for some large boost levels, then you're probably better off with just a pair of TD04s or something. I'm aiming for around 30psi, and obviously don't want the slow spool time normally associated with that, so compound should work out well.

[Davezj]

this is why i did not understand what sean was trying to get at with compound charging if you are only going to aim for 18psi.
as nick says you could achieve this with a tdo4 upgrade so there would be no point in compond charging.

[scott.mohekey]

Exactly. While you MIGHT be able to get to 18psi with lower temperatures, you're overly complicating the piping for essentially no reason. It's only of any use when you want to run far more boost than is usual for a given engine.

[SEAN-NZ]

i only used 18 for ease or working with the number, as 14.7 x 1.5 x1.5 gets you 33psi, didnt want to make it too complicated for everyone

[scott.mohekey]

Oh, were you meaning 18psi from each 'stage', that makes more sense. Instead of this, I'm going to run around 13psi on the stock turbos, and closer to 20psi on the larger turbo.

It's not really accurate to talk about running a given psi on each stage, as its more about the pressure ratio.

[SEAN-NZ]

i was just meaning the big turbo will boost to 1.5 intake pressure, then little turbs will boost to 1.5 the intake pressure, giving overall 18psi of boost, working with say 1.7 of each would get around 28psi, just trying to make it sound too complicated, just had that figure as a base for ease of maths, but with the first example it would only be 18psi of boost

[Kenneth]

Don't be so sure about the temperature. You have to take into account the efficiency of both stages, which will also compound.


This is one of the main benefits of compound charging; high boost levels at workable temperatures. The other benefit, is that you get the lower spool times of the smaller turbo(s). I'll agree, if you're not going for some large boost levels, then you're probably better off with just a pair of TD04s or something. I'm aiming for around 30psi, and obviously don't want the slow spool time normally associated with that, so compound should work out well.

[SEAN-NZ]

if space permitted, although i doubt it would, could there be a possibility of a small air to water inter cooler between stages?

[Nick Mann]

if space permitted, although i doubt it would, could there be a possibility of a small air to water inter cooler between stages?
You don't need a huge amount of space for these though, and they can be placed anywhere with the cooling for the water being completely remote. It's not a bad plan.

[Nick Mann]

http://www.ebay.co.uk/itm/Barrel-style-water-to-air-chargecooler-intercooler-/290752728587?pt=UK_CarsParts_Vehicles_CarParts_SM&hash=item43b2368e0b#ht_720wt_932

You can get all sorts of shapes and sizes, but something like this (It would need to be able to flow more!) should be able to be squeezed in somewhere.
Part of me wonders if one of these would help any stock setup.

[SEAN-NZ]

could run 2? just before each little turbo, that would add up to 580, althouh that could probably still be too low for some people

[Nick Mann]

There's loads of options. I have just bumped an old chargecooler thread as I thought we were going a little off topic here.

http://www.clubvr4.com/forum/showthread.php?28971-Water-air-intercoolers

[SEAN-NZ]

ahh, i just finished reading that thread, some interesting stuff on it

[Davezj]

just thought i would let you know, i did some measurements on inlet and exit air temps on a standard and 600x300x76 intercoolers fitted to my car. can't remember what boost presure this was i have it in another thread somewhere i think.
at the lower end of the temp range obviously the intercoolers do not cool down below ambient temp so below about 30 deg C inlet temp the outlet is about 20 deg C

the stock intercooler very roughly cooled the exit air to half the inlet up to 70deg C inlet then the exit air started to rise by more than half.
inlet exit
40 20
50 25
60 30
70 35
80 45
i presume it was coming out of it's effecint range and from then on it would just get less and less efficient at disapating the heat.

the bigger intercooler, cooler the exit air down to about 1/3 of the inlet temp. up to at least 80deg C.

my monitor for this has stopped working which was a UTCOM trip computer, so i can't do any more tests on it at present. but the limit of temp measurement is 80 Deg C so i could not go any high than this with the inlet temp.

when it is up and running again i might do a divide by 2 mod to the inlet temp so it would give me the ability to measure up to 160 deg C, althought it would diplay up to 80. not that the inlet temp should get anywhere near that figure, just out of interest though.
but first i have to get it working again and at the moment i have no idea why the screen is blank.