My compound turbo set-up

[Kapok6]

It sounds like your trying to avoid modifying your oil pan. You can run a -20AN line but the hole in the oil pan is still to small to drain two turbos. Oil pan modification is going to have to happen. I wouldn't take any short cuts when it comes to a compound setup. Do it once and do it right with each turbo having its own drain. Those are my thoughts on it atleast.

I agree with the "do it once, do it right" idea, however just because one person does something one way doesn't mean that a different way a person accomplishes the same task "isn't right." I have done a few things to my DSM for which I was laughed at and told it wasn't the "right" way to do it and here I am a few thousand miles later running 30+psi on a 65lb/min turbo with no issues as a result of me doing it the way someone else said was "the wrong way."

I agree with you though that 10an may be a bit small, but instead of messing with welding fittings on my current oil pan, I did see a pan with one -12an fitting already welded on for sale which would make things much easier for me than taking my pan off, buying 2 fittings, then taking them 10mi away to the welder and having to pay him to weld them on, then pick it up, bring it back, and puit it back on the car. If I bought a pan with one -12an fitting already welded on, I simply pull one pan off, put the next pan on, and away I go. I want to do it in a manner that is functional and requires the least amount of labor/money possible.

I run my 50 trim and S475 into a 5/8th plastic Tee, which then goes to the pan. The Tee is at the level of the pan fitting. In fact, ,the 50 trim is level with the pan fitting, so that whole damn thing is all horizontal, only the leg going to the 475 is vertical. None of this should work. But it does.

With a BB turbos that uses very little oil and an MHI, Garrett, BW, or any other OEM or OEM based turbo, I would try it. PTE turbos are the only ones I've used that try to burn the car down if the drain line isn't completely vertical all the way to the pan. The seal system they use is very sesnsitive to oil backing up in the drain even slightly. OEM turbos are pretty tolerant.

That being said, I've been meaning to cut off the existing 5/8ths stub on the oil pan and weld on 2 10an fittings since I built this setup in 2009. There hasn't been a pressing need, other than cleaning it up a little, so I keep putting it off.

Yeah, MHI and Garrett are the turbos in question. I probably wouldn't try this with a Holset based on what I have read. The other thing I am wondering if it would help in running one line, I have a crank evac system like Paul's, so would a system like that creating vacuum in the crank case help pull the oil into the pan as opposed to relying on just gravity to get it from the turbo to the pan?

 

[Marc91GSX]

Paul, I'm just spit balling some ideas here but I was wondering if going to an evo 8/9 manifold and turbo for the primary would be good setup. Since you've already removed the PS pump and relocated the alternator you would have plenty of clearance for the evo turbo and iirc that turbo is alot more efficient than what your previous setup utilized. I also am wondering if having it setup like that would make the exhaust flow more efficient since you could theoretically have the hot sides of the turbos practically welded together since you wouldn't have the compressor housings interfering hitting.

Either way this is a really sick setup and thanks for proving it's functionality with on a DSM platform. I want to do this to my 4G Starquest project that I will finish one day.

 

[M4s73r]

Well I'm inspired/impressed right now. Read the thread. Watched the videos. Yep I'm Sold.
I have a S16g sitting on the shelf. I have a FPRed in the car now. I like the feel of it after about 4.2k rpm. But I'm really missing my part throttle/low RPM power. I do however have a couple of questions.
Any of you guys used Link to control your boost on this yet? Ive been using it. And really enjoy having different psi based on what gear i'm in.
What about heat? As tight as everything is in there what have your intake/underhood/coolant temps been looking like? did you do anything different with fans? Ducting?

 

[boostitall]

wow! no one posts here any more??? i love this idea for my autocross/dailydriver. i also love fabricating too. i was thinking much lower on power though.

i dont see alot of posts here that shoot as low as i want though...i hope i dont offend?

my idea is using a td04l-13t from a WRX as the HP primary, and a gt3076r for the secondary LP.

my goal is to get high torque/boost at 2500-3000rpms and a max power of like 400whp around 8500rpms.

any input?

 

[Kevin Jewer]

The problem with low power is low PR. Then divide that over two turbos and it's even lower. If you can pick turbos that can work at very low PRs, you can probably make it work.

 

[boostitall]

The problem with low power is low PR. Then divide that over two turbos and it's even lower. If you can pick turbos that can work at very low PRs, you can probably make it work.

from my research the gt3076r at 11psi (.75bar or 1.74 PR) will be 71% efficiant at 40lbs/min, or 550cfm, and 65% at 45bls/min, or 625cfm...so, 400-500chp ish?

the tdo4l-13t at 14.5psi (1bar or 2 PR) is efficiant to 27lbs/min or 405cfm.

myboost goal is around 20-23psi.

i just dont know how tee=ll if this is workable or not? i found this formula for compound turbocharging on a vw tdi site: http://www.vwdiesel.net/forum/index.php?topic=28781.msg243586#msg243586

m2=m1*sqrt((T/Tmap)/(P/Pmap))

but it doesnt make any sense for this type of setup. even the OP 99gst_racer's setup doesnt work with this equasion either?? but it did work well!!

i think my idea for this set up should work based on scientific guess work, but it would be nice for some concrete fact of yes/no it works.

 

[pneumo]

You'll have to do your homework to get any compound turbo setup to work at it's best. The trick to your proposed setup is that you're trying to make two small turbos work at low boost, when the two main benefits of compound turbos is the ability to run very high boost at high efficiency levels with reasonably quick spoolup. You're really not gaining much efficiency by running two turbos at such low PR. IMHO you could reach your goals more easily with a stroker motor and a modern small/medium sized turbo. A stroker motor will make nice power around 2-3k rpm without needing to make a ton of boost. If you pair it with a twin scroll EFR6758 and you'd have a quick spooling setup that can easily make over 400awhp.

 

[Kevin Jewer]

from my research the gt3076r at 11psi (.75bar or 1.74 PR) will be 71% efficiant at 40lbs/min, or 550cfm, and 65% at 45bls/min, or 625cfm...so, 400-500chp ish?

the tdo4l-13t at 14.5psi (1bar or 2 PR) is efficiant to 27lbs/min or 405cfm.

myboost goal is around 20-23psi.

The problem is that with those two PRs, boost is 36 psi, far more than your goal. Airflow is going to be far more than what you mention there. At 20 psi, you're looking at a PR of 1.5 on both turbos. That's pretty low. Efficiency is bound to go to shit that low for most turbos. If you go for that 36 psi it might work. If you really want boost/airflow that low I'd consider running the big turbo all the way to full boost and seeing what happens. The more time goes on, the more I think about that strategy for these smaller setups. I don't believe it's been tried.

Throw that formula out the window too IMO. It just isn't that helpful. Pick the boost/airflow you're going to run (airflow can be estimated pretty closely), then try splitting the PRs different ways to get decent efficiency on each turbo. Divide total airflow by big turbo PR to ballpark where you are for airflow on the small turbo's map. That's not exact, but it's close enough to see if you're even in range or not.

 

[boostitall]

The problem is that with those two PRs, boost is 36 psi, far more than your goal. Airflow is going to be far more than what you mention there. At 20 psi, you're looking at a PR of 1.5 on both turbos. That's pretty low. Efficiency is bound to go to sh** that low for most turbos. If you go for that 36 psi it might work. If you really want boost/airflow that low I'd consider running the big turbo all the way to full boost and seeing what happens. The more time goes on, the more I think about that strategy for these smaller setups. I don't believe it's been tried.

Throw that formula out the window too IMO. It just isn't that helpful. Pick the boost/airflow you're going to run (airflow can be estimated pretty closely), then try splitting the PRs different ways to get decent efficiency on each turbo. Divide total airflow by big turbo PR to ballpark where you are for airflow on the small turbo's map. That's not exact, but it's close enough to see if you're even in range or not.

FIRST: thanks alot for the input!!! there is just not alot on this topic i have seen except in the diesel world.

here is the link to the calculator figures i am getting:
http://www.not2fast.com/turbo/gloss...=21&SFC=0.55&AFR=10.5&maxInjectorDutyCycle=85

so, i need roughly (cuz calculators are just a good start..) 48lbs/min for a little wiggle room on my goals. the gt30r is about 48lbs @ 1bar still at 70% efficiantcy. so then i try to find the right smaller turbo to compliment the big next. the part im alittle fuzzy is how a PR of 2 (14.5psi) on the big or LP turbo, and the PR of the small is 1.5 (7.5ish psi) come together? i know my #'s are changing a little since the last post...sorry. so is it simple addition of the 2 PR's? 14.5psi+7.5psi=21psi total manifold pressure? it seems im not quite sure on the math. if i get an efficiant 14.5psi on the gt30r all the way to my total airflow needed, how much boost or what PR do i need on the small turbo to get 23 psi or so manifold pressure? just 1.5PR on the td04l-13t is kinda efficiant at 60% to 320cfm and at 1.6 PR is much more efficiant at the same cfm of 320, and goes to 345cfm at 60% efficiantcy. so 1.9 on the gt30r and 1.6 on the td04 gets my 1.5bar boost or 23ish psi?
am i getting close to understanding this? i am an excellent fabricator, but the science part of this i am lacking...!

the reason i am only wanting 400whp is my drive train. i will upgrade later to a drivetrain that will handle 500-600whp and the stroker someday too! i was just hoping to get something like this working now, and with minor changes later go for the big #'s!!!

 

[Kevin Jewer]

The PRs multiply together. The boost may add together depending on how you set up the wastegates, but PRs always multiply.

Be careful when trying to hijack the not2fast calculator for compound turbo use, it's got a few issues. Info from the diesel guys can be tricky too, they do it completely differently than the way I think we should do it and for different reasons. Some of their trusty guidelines just don't carry over to us. Some of it does.

Trying to model this setup for you at ~20 psi and 40 lbs/min, I can't make it work. Meaning, one turbo or the other is always off the map. That doesn't mean you can't run it, it just will be less than ideal. But for the record, my compound setup that ran 8.80 at 157 wasn't ideal either.

If you did try to run the turbos so that the big takes over completely, as I mentioned above, you'd be right in the meat of that compressor map for the 3076, about as good as it gets. Checking the numbers though at higher psi when you decide to turn it up, I can't get the small turbo on the map. It's just too small. So I would still try to phase out the small turbo by max boost. This will require a fair bit of wastegate capacity around the small turbine. You could go to a bigger small turbo, like an EVO3 16g, but then the 3076 becomes too small for a big turbo (only another ~100 hp capacity), so your whole setup gets much bigger, maybe much bigger than you even need. This is the problem with compounds, they really want to make power. IMO, they start to make more sense around 4:1, or 45 psi. But I don't want to discourage you from trying it! There are other reasons to do it than just max power.

Build it so you can upgrade each turbo within it's given foot print so you have the ability to easily upgrade and you can always change things later. For example, 14b to ev0316g, 3076 to 3586hta, etc. You can get quite a range of options without having to start over with the fab work. If you wanted to try to optimize it a bit, it looks like replacing the 13t with a small 16g just gets you in range. It doesn't do too badly. Even at the max of about ~35 psi, it can be on both maps. If the small 16g meets your spool up needs and the 3076 meets your max power goals, it can be made to work. If you were interested in going this route I could model it more closely and estimate back pressure as well.

Keep in mind too that a stroker makes things worse with the smaller turbos (airflow goes up at the same PR, moving even further to the right). Luckily a stroker can spool bigger turbos so there's hope, but you might want to go bigger from the start and deal with the spool to avoid having to buy turbos twice.

Sorry for the long post, but I hope it helps.

 

[viperlp01]

That WILL not work ESPECIALLY to 8500 rpm. 400whp can be achieved with an evo FP white and have almost insta spool. This is designed to work with a LARGE secondary and a primary to compliment it to reduce lag and have a wider area under the curve. What you are trying to do doesn't make much sense. Just the compressor flow capabilities to make power to 8500 without spinning the secondary to the moon is more than the 3076 can do.

If your trying to be innovative and have awesome torque design a nice t4 twin scroll setup based around a borg warner s256 and integrate a quick spool valve. Simple, less weight, ect.

 

[boostitall]

both of you make complete sense.

i am now running a single td05h-small16g. i love the midrange, but the bottom is at like 3700rpm's and my boost falls from 23psi at 6000rpms, to 17psi at 7500rpm's. idealy i want a perfect turbo that gives better low response at 2500rpm and better top end at 7500-8500 rpms. there is just not such thing as a perfect single turbo as all options put me one way or the other but not both.

my proposed setup for now will only see 7500rpm's, but my next engine will be a stroker and better heads to get the 8500rpm and 2150cc displacement. i am sorry my numbers are confusing cuz of my multiple possible setups, and i am confusing myself along the way...HAHA!

there are some other td04 options that fall between the td04-13t and the td05-b16g. there are 15t, 18t, and 19t. the 19t is very close to the big 16g on the map. i was just picking the one with best spool response, as each upgrade does sacrifice a little spool along the way, and to add i already have the td04l-13t in great shape.

i am totaly using a fabrication design that compliments many different turbo options, that has been in thought from the start.

so if the big turbo takes over in the top end, i was thinking that i could use a big wastegate to bypass the charge pipe past or around the td04 at a certain boost level or rpm/boost level, but that would be going to a compound/sequencial design...is that doable, or do i need some kind of reed valve, or popoff system? sorta like some supra's and and diesel systems?

 

[Kevin Jewer]

Intake switching stuff starts to get complicated IMO. I would do it all on the hot side. The small turbo will free wheel and stay out of the way. You just need enough wastegate to keep the small turbo PR very low. If the PR is low enough, it doesn't even matter if the efficiency is 40%, it doesn't add much heat to the air. The simplest form of control that comes to my mind is to just set both wastegates to your target boost level, with the big turbo WG several psi higher. That should force the small turbo WG all the way open once the big turbo comes up. One potential problem is that with the small turbo at very low PR, back pressure evens up across the turbos and reduces the effectiveness of the wastegate, so it needs to be pretty large. Relatively inexpensive, no fancy gizmos, personally I like this option. But I've never done it, unfortunately. It could be the answer for these really small compound setups, so hopefully it gets tried eventually. My setup might be good for testing since it has two small turbo wastegates. I could vent one of them if needed and if it helps.

At any rate, keep the dream alive. There's probably way, that just hasn't been done yet. There are a number of people currently building/running really small compound setups but no good data yet.

 

[boostitall]

WOW! you just read my mind! that is exactly what i was thinking before i looked at anything in this forum or the internet. as i looked i didnt see much like you said actualy done and documented. most of you and others on this forum are running true compound setups where the turbo exhaust supply, and charge supply, and even down to wastegate control is all set up compound. this idea that you are sharing that was my original conception would still be compound in the turbo's and piping, but the wastegates will be referanced by the manifold pressure (well pre intercooler/post HP turbo) and referanced in common with both the LP turbo & HP turbo. just using different spring/controller settings.

i really didn't like the over complicated sequencial valves and such being in the charge piping or the exhaust sides. same reason i don't do t4 swinscroll w/ exhaust valve controllers. it's not that they don't work, i just like the compound design. it is for some, and not for most. i like what some one said earlier in this thread that compound setups are not going to be a marketable item really, more of a DIYer, or custom ordered special fabricated to a customers asking.

yes...i would like to set the td04l-13t HP turbo at 1bar and using a dutycycle map based controller to raise the boost to 20psi or so at 3000rpm's, and when the gt3076r comes in at 4500rpm's i would raise the dutycycle to 100% to bring the HP turbo to idle cuz the bosst will be higher than 1bar, then the wastegate on the LP turbo would have a spring set at 1.5bar(23ish psi) and be flat and consistant to redline. all while the HP turbo wastegate is soooo open at 23psi it would really be idle or bypassed to the point of not being an issue.

so the next question is wastegate sizes? the OP was using 44mm for both. i think that would be ok for me, but it wouldn't hurt to go to a 50mm one would it? is wastegate over sizeing even an issue other than external size and placment issues? i say 50mm for the HP turbo cuz the td04l is 41mm exducer and 6cm housing....choke choke/hot hot! 50mm should be enough cuz my up pipe is 50mm, so it would be close to full bypass as u can get w/ 90 degree bent wastegate. then a 44mm should be fine for the gt3076r cus the OP had one on his 60-1, or should i just get 2 50mm's? like i said, IDK if wastegate over kill is an issue or not? i would rather be safe, then sorry!

 

[Kevin Jewer]

Keep in mind that as soon as the small turbo spools, the big one will right after it. Whatever notion you had of what RPM the big turbo will spool goes out the window. Another way to look at it is to multiply the displacement of your motor by the PR of the small turbo an consider at what rpm a motor of that size would spool the big turbo. For example, once your small turbo gets to 15 psi, you effectively have a 4 liter engine, and so on.

Aside from size and cost, there's no reason to avoid bigger wastegates. My setup used two V44s on the small turbo, and 1 v44 on the big turbo. A single 60mm is almost the same size as two 44s.

For the boost control, I would start simple. Set the small turbo to a relatively low boost, but one where the big turbo is starting to spool. Once that WG opens the big turbo should spool even more quickly since unmolested exhaust gasses will hit the big turbo ("early wastegating"). You should be able to get pretty good performance just by balancing two simple MBCs, with no complexity to speak of. If for any reason it doesn't work, you can move to something more complicated knowing that you have to. In your ~20 psi example, I would start with the small turbo at 15 psi and the big turbo at 20 psi. If it works, I think you'll be surprised how quickly the big turbo spools. The wider the spread in boost pressures, the better it should bypass the small turbo. The narrower the spread, the quicker it should spool, but there will probably be a point of diminishing return since you want those hot unmolested exhaust gasses to reach the big turbo. But again, I haven't tested this. If you haven't started building yet, you've got plenty of time to think it over.

 

[boostitall]

yes, what you describe is what i was thinking also. 50mm wastegate is what i think, like i said before, my up-pipe is 50mm. does it make sense to go bigger than my up-pipe?

as far boost control goes we are on the same page too. if i can, i will just use the springs w/o controllers to see how it works and prove the concept. like 7psi small, and 14psi big to see it work, and feel the responses. 1bar on the small and 1.5 on the big seems about where i would like, but i'll see how it goes...nice and easy like! i would only use complicated controllers if in the end i need to.

the best thing to me about the compound set-up is other than having more piping/turbos, it stays rather simple if you want it to.

 

[Kevin Jewer]

I'm not sure what you're referring to by up-pipe, but in general a WG of a given size isn't going to flow as well as a straight pipe of the same size, so there may be some benefit to going larger. On my setup I couldn't keep boost down with a single 44 on the small turbo, it would hit the 50 psi boost cut immediately. That was a bigger setup though, with a 50 trim and an S475, so I can't say for sure how the 55mm would do.

 

[boostitall]

first i want to say that i am a car/truck/anything driven on the street enthusiast. i love all cars makes and models new and old. one of the cars i have always loved is the eclipse/talon/laser awd turbo-ed cars from the 90's. they are impressive in many ways for sure...much respect to you all who do amazing things with them! i have been hesitant to say the car i'm working on is a Subaru gc8. mostly cuz of the whole EVO/STI animosity. i wasn't sure if i would be welcome here or not. i hope this doesn't change the game of this topic, because like i said before, there isn't much talk about compound turbo elseware that gives true results and encouragement for it.

so, the Subaru manifold is connected to an up-pipe that leads to the turbo in some foreign land far away. subi's are rather notorious for lag because of this amount of piping between the turbo and engine. most people go with the 2.5l sti block and don't think twice about building a 2.0l as a low rpm torque machine. others will do a low mount setup that i'm not to keen on. i know some may say there is a reason to there logic, but i actually get a thrill out of doing and building some thing that makes some say "why?", and others say "i can't believe you actually did it?", or "you did not just beat me with a 2.0l on the autocross track?"

i hope we can continue our conversation with mutual respect, or maybe i'm over worried? i hope!

 

[Kevin Jewer]

Doesn't bother me in the least. I've been involved in or at least privy to dozens of compound setups on all kinds of cars. They all present unique challenges, and are therefore all interesting to me. The vast majority of them have stopped in to the compound turbo thread on Yellowbullet, which is the most comprehensive collection of gas compound turbo projects I've ever seen. When you start doing some work, be sure to stop in there. Or check it out now for more ideas.

 

[Ohhchhittt]

Wow very impressive

 

[TomLV]

Hi there,
I started to read this forum last year beginning and now I have tested my first compound on Honda. Too bad it wasn't right and I have missed something. These last comments are really useful, but still I don't get it how did You calculate, that "boostitall" small turbo isn't on map? My goal for now is only 300hp and smaller turbo is even smaller than "boostitall's", but all points I get on map following Your advice. Big turbo is the same GT3076R, small one KKK K14 or similar GT1749v.

I added a graph from dyno. The big turbo which was T3 60-1 didn't even show up (could get only 7psi), so now I plan to change it to GT3076R. Small one was producing 22psi and still no clue from big one. I guess there is some restrictions, maybe in the piping from LP compressor to HP compressor because it's only 44mm diameter.

If you did try to run the turbos so that the big takes over completely, as I mentioned above, you'd be right in the meat of that compressor map for the 3076, about as good as it gets. Checking the numbers though at higher psi when you decide to turn it up, I can't get the small turbo on the map. It's just too small. So I would still try to phase out the small turbo by max boost.

 

[TomLV]

Last night read also yellowbillet site, understood, that I will start with small one by itself and then chose big one. Maybe You have already seen this, but small turbo really works, here is proof (it's diesel, but still):

http://www.darksidedevelopments.co.uk/golf-mk4-1-9-8v-tdi-arl-4-motion/

 

[Kevin Jewer]

If you want, post or send me some more info. Pressures on both turbos, intercooling, displacement, rpm, etc. I'll plot it out for you and let you know where I think it falls.

 

[Archer Fabrications]

Hi there,
I started to read this forum last year beginning and now I have tested my first compound on Honda. Too bad it wasn't right and I have missed something. These last comments are really useful, but still I don't get it how did You calculate, that "boostitall" small turbo isn't on map? My goal for now is only 300hp and smaller turbo is even smaller than "boostitall's", but all points I get on map following Your advice. Big turbo is the same GT3076R, small one KKK K14 or similar GT1749v.

I added a graph from dyno. The big turbo which was T3 60-1 didn't even show up (could get only 7psi), so now I plan to change it to GT3076R. Small one was producing 22psi and still no clue from big one. I guess there is some restrictions, maybe in the piping from LP compressor to HP compressor because it's only 44mm diameter.

I'm curious why you would even attempt a compound setup with such low power goals.... 300whp? easily obtained with a 16g by itself.... but i understand the want to be different. it's just a lot of unnecessary work.

on another note, are there any block fittings that anyone has found to be adequate for an oil drain if you didn't want to use a pan fitting, or if in a compound setup, an additional spot to drain. balance shaft inspection port ported out on the front of the block maybe?

 

[Kevin Jewer]

Here's another compound setup I've been working on recently. It's works very well, this converter is extremely tight and it still stalls it up eventually. The turbos are both Forced Performance, 73HTA and 4205HTA. I've been saying for years that the most critical part of compound setups is wastegate control, and that's the only thing I'm still working on with this car. Once I can get boost to stay up over 35 psi this thing will be pretty quick. 6.0 at 115 in the 8th so far.

You must be logged in to view this image or video.