I need your Small Displacement Turbo Help! (Long)
 

Hi All,

Big lurker, first time post to the Pelican Turbo forum. I am embarking on a new project that is pretty atypical to most of the projects I read bout here, so am hoping I am not so far in left field that no one can help. I know there is no shortage of informed opinion and knowledge here, so fire away. Thanks in advance.

I recently retired my little MFI ‘S’ motor and since I’ve always wanted to put a turbo motor for my early 911, figured I’d build one in the spirit of my old motor. Because of this, the typical turbo doesn’t appeal to me, meaning 3 liter and up motors won’t meet my desires. Frankly, they just make too much power! :) Yeah, that shouldn’t be possible, but I said it anyway. Please don’t reply with the standard “Just by a 3.3 and be done with it” comments – they aren’t going to fly; my course is set.

The motor I am building is mostly specced and I already have most the parts to build the shortblock. It will be a small bore turbo, along the lines of the Baby RSR. Basic spec is for an 85mm bore x 66mm stroke, 2.3L intercooled engine. HP target is between 320 and 340HP @ 7500RPM, which calculates out favorably against my goal for boost at a conservative .8 BAR. The motor is being built to rev safely to 8000 RPM and my cam choice makes power to suit (see below).

My final decisions are down to figuring out the turbo and IC configuration. There seems to be a lot of debate between twin turbo configurations vs. using a single unit, especially regarding the throttle response. Seeing as this motor can’t rely on displacement to overcome off boost deficiencies, I want to make sure that my selection maximizes spool up.

During my research, it seems that the preference is for ball bearing turbos over journal bearing turbos, due to the efficiency gains they provide. Unfortunately, when matching my application for twin turbo applications, it seems the only ideally sized turbos that meet my needs are journal bearing turbos.

So the decision looks like it is between:
1. Using two journal bearing small turbos like the GT15
2. Going with two larger, but less ideally sized bearing turbos like the GT25
3. Going with a single ball bearing turbo like the GT28

I was convinced that twin turbos are the way to go, but modern, water cooled turbos w/ variable valve timing like the WRX, Mitsu Evo, Audi and VW 2.0T cars all make excellent throttle response and power from one turbo. From my reading, I notice a consistent theme that most turbos today are most efficient over 1 BAR, typically in the 1.2 BAR range; mine is much lower, so efficiency should be down. Dilemmas, dilemmas…

There is obviously a lot of tweaking that can be done on the A/R and compressor/turbine pitch to optimize any of these, but my question is:

Given the above options, what would you do to maximize spool up for a motor like this and get similar throttle response to what these modern cars make?

As far as full specs, here’s where what I already have:
85mm bore – Nickies w/ 7.5:1 JEs, thermo-coated
66mm stroke – SC rod journal grind, cross drilled
Custom Dougherty turbo cams: GT2 intake with the DC43 exh on a 108 lobe center (@ .050” Duration=248I/242E, lift .485”I/.484”E, 108 lobe centers)
Coated STD main bearings, grooved for cross drilled crank
R&R chromoly rods w/ ARP bolts from LN Engineering, GT3 spec (same size as Pankl Ti-rods)
2.4T 7R case, fully prepped by Ollies: shuffle pins, oil bypass, drilled for cross-drilled crank
Supertec head studs
2 liter heads w/ ‘69S valves, 39mm ports I/E, twin plugged
930 oil pump, race prepped
Properly sized George Narbel headers
ITBs

Don’t have it yet, but will soon:
EFI - Will use MegaSquirt II Sequencer system (waiting for final availability)
Coil on plug
Crank Fired
Wideband 02, w/ MAP, TPS and intake temp sensor
Apparently this ECU supports knock control and electronic wastegate/blowoff control – would love to integrate this

I am not totally in the dark here. A similar motor to my spec (but with a less aggressive cam) produced this dyno run:

http://forums.pelicanparts.com/uploa...1243356292.jpg

Sorry to be long winded, but thanks again to everyone who battled through.
Thanks!

So, 320-340 hp out of 2.3 L at .8 bar?
I don't think you can get there that way.
Better up the boost to 1.1-1.2 bar. You might need to run a lower static compression ratio than 7.5:1.
As a data point, Porsche's first turbocharged racing engine had a 2.1 L displacement and ran 1.4 bar of boost, producing 490 hp @ 7600 rpm (source: Porsche 911 Story, 7th edition, pg. 269).
Sounds like an interesting project. Good luck.

You are totally right - I did some calcs and .8 Bar is closer to 305 HP w/ a reasonable efficient IC. Better to keep the half point, since the small displacement will need all the static CR it can get for off boost acceleration.

1 Bar gets ~350HP, based on said calcs. @ ~12 HP per psi (basic rule of thumb, all other variables remaining static), this basically jives out. Water/alcohol injection would support this on pump gas. Maybe a future project? ;)

BTW, the 2.1 RSR still pencils out at 380 HP a .8 Bar, per my calcs...something is fishy.

I was going to build a small displacement turbo engine for my own car a few years ago....You are following the right path building the bottom end as robustly as possible.

Why are you limiting to 0.8 bar boost?

Put a 0.7 spring in the wastegate, and use a boost controller to select your boost..

Why 85mm? 84 is standard 2.2/2.4 bore... You can get the nickies or Mahle cylinders... Then get a JE 2.4S piston.. It will be forged, and low c/r..probably 8.5:1 which is fine for a EFI controlled turbo engine.. This may be a more economical path

I'd use other than 2.0 heads..go for some heads with as large port/ valves as you can get..More aggressive cams i.e. more lift and duration.. Also you make no mention of lightweight valve components i.e. Ti retainers. Also hi perf springs will be necessary if you want to spin the engine that high..

Narbel headers are mild steel...M&K exhaust has some interesting mods of stainless SSI headers, the secondaries are shortened and a v band clamp is welded on.. This is ripe for easy twin turbo apps.

You could save a bunch of money by using a 3.2 manifold... in lieu of ITBs.. The Nth degree of performance will be realized with ITBs

Are you going to twin plug the engine?

I'm a bit confused on your description of the EFI set up... Will you be setting up the MSII sequencer for sequential injection and direct spark? If so in addition to the crank sensor you will need a CAS (cam angle sensor) you can take this obviously from one of the cams, or use an old distributor to provide the signal.. You can control a COP system with just a crank sensor though ...

That MSII sequencer looks very interesting...I may have to pick one up..

No unless you can get variable valve timing and variable turbo vanes...

Sorry I strayed from which turbo to select...I'll ask Bill which turbo he told me I should use when I was going to build a 2.0 turbo...

Not an expert but here are some thoughts.

7R case starts reaching its limit at about 270hp. At the heat of a turbo with severly increased combustion pressures and you might be taking a risk. Might be better to find an early sand cast aluminum case if durability is important.

Small displacement, short stroke, low compression, cams that will support a usable power band up to 8000rpm -- and I suspect low end power and response is not going to be in the cards.

For example, a K27-7200 turbo supports 320 hp. Boost starts to come on at 2500 with full boost at about 3200rpm on a 3.3. Put in on a 2.2 and boost will start at about 3700 rpm with full boost at 4500rpm.

Further, you will not likely get low overlap cams to work well at the rpm you want to reach. Thus, boost will come on later with the same turbo.

Also, you may need a larger lower restriction hot side for less back pressure so there is less hot gasses being pushed back in through the exhaust valve and past intake valve up into the intake manifold.

Think a step beyond a 911S 2.2 (906?) in terms of power band and throttle response. You will probably want the turbo and cams to come in at about 6k to work right. No reason to go twin turbo that adds complexity and weight and you will be looking for turbos that make about 160hp each.

What would be fun would be to run an MFI set up with your 2.3 turbo set up.

The best.

Interesting project.
How far outside the box are you willing to go? With engine management combined with tiny displacement you can try something really different - a single VNT from the 996tt.
Max reduction in lag and just about the perfect size for 0.8bar from 2.2L engine.

TimT, I was really hoping you'd chime in. OK, one thing at a time...

Damn, jackpot! 911ST, I was hoping you'd chime in, too.

Holy Trinity; thanks RarlyL8!

So obvious, I can't believe I didn't think of it. I am sold unless someone can come up with a reason not to.

From what I can read, the stock 996 turbo is the K16; is this correct? For some reason I thought the VTG turbo were Borg Warner. I thought these were 997 only, though. Any clarification would help.

Tim,

Can you explain why 8.5:1 is OK when it sounds like getting even .8 Bar at that CR sounds like it would be impossible? What are the variables that would allow this with EFI? Ignition mapping w/ load sensing, etc? Fuel enrichment for charge cooling? Love to hear some toughts on what you gain/lose at this static CR...

I thought about the VAT also. Historically managing them was out to the abilities of most. I do not know if that has changed. This is the only way to reach the goal of low rpm full boost and carrying 8k I suspect.

As to the hot side I was referring the the turbine, not the tube size. If it spools at low rpm it will make a lot of back pressure up top. With the cams you will need to carry 8k this should be an issue. Talk to your cam builder.

Trying to reach boost by 3k and pull to 8k is a sizable task.

Did SW know you are going w a turbo on you R case? We are talking a lot more heat and potently expansion. Having said that there has been one builder I read about the built a 2.7 twin turbo he used to drag race that made big HP w/o a problem. Road Racing might be different.

Best of luck and look forward to the project.

This motor is a change of direction, so wasn't factored when I built the case. In talking to him later, he didn't feel it would be an issue, so have decided to move forward. I am still tryin to tease out all of the particulars in this regard, but all looks to be OK as long as I'm not greedy.

I assume managing the variable vanes is through electromechanical control, not by pure mechanics (like a bypass). I will have to research the ECU capabilities more to confirm that this isn't beyond my capabilities - I am not so skilled as to write custom code for vane actuation.

OK, looks like I cleared up a misconception on my part. Only the 997 uses VTG (variable turbine geometry), not the 996.

As a recap, these adjust flow by altering the air incidence angle upon the turbine. Basically, the 997 Borg-Warner VTG turbo is just a more modern version of the variable vane technology that was piloted in production back in 1989 in the old Chrysler Shelby CSX turbo POS. Older variable turbos were indeed mechanically actuated, but were unreliable in gasoline engines due to high exhaust temps (compared to diesels). The new Borg-Warner VTGs use aerospace materials and thus are very reliable and make great pressure.

The 997 Borg-Warner VTG appears to be computer controlled, but have yet to verify...anyone know for sure?

One other note: apparently "variable vane", "variable turbine" and "variable nozzle" turbos all refer to the same thing.

Sorry about the 996/997 confusion. Yes 997 is the variable speed turbo I was refering to.
I'm sure it can be made to work with less complicated engine managment electronics.

Cool - was just looking at some Garrett VNT-25 installs on other marques. Very interesting, although they are finicky (can get sticky). You can use them without ECU management, but then your boost is 100% RPM dependent, so forget regulating boost to load. They are a neat turbo, though. The CSX Shelby did use ECU control. They are limited as manufactured to ~250 HP, but supposedly you can install a GT28 turbine in them, which apparently supports 400HP with minimal impact on spoolup...interesting option.

I'd bet that the 997 vane actuator (and the VNT-25, for that matter) is analogous to a stepper motor, which MegaSquirt can control. BTW, those 997 turbos are like hen's teeth - you pretty much have to buy them new, which is an EXPENSIVE proposition.

Since you already own a bunch of the pieces you need roll with it..

Yes better control of all aspects of the engines operations..I'm going to delve further into the Microsquirt sequencer and what it can do.. It may have available channels where you can use knock sensing.

With 8.5:1 you will get crisper off boost throttle response.. though maybe not so much with a small displacement engine. I just put that out as anecdotal..We have built a number of 8.5:1 turbos... some with CIS.... and haven't experienced problems that were purely c/r related...

The new Porsche 997 turbo runs 9.0:1 cr Granted they are water cooled... .. On a related note our 935 clone is 8.7:1 and we sometimes run it at 1.6 bar... usually it runs at 1.2 bar...The only failure this engine sees is if it is over revved... other than that bullet proof..

Back to the engine management side.. you can run fat in areas of the map to keep EGT in check..

Your engine will be a real fun engine... especially in a light car...

My car weighs 1846 wet ;)

MUST KNOW MORE. How are you getting these pressures at this CR w/ such large pistons? Is it ECU regulated for short bursts? FAT fuel maps to cool the charge on load?

Damn, I can't trust whatIi read. High CR, big boost and reliability...WTF??? ;) And to think I thought I needed water injection to get over 1 Bar...

SICK. :eek:

I will probably come in at 2100 lbs - this car will have some creature comforts, albeit not many. Mostly carpet, engine pad, some tunes, etc. All steel body hurts weight...

Stole this pic from another thread of the MFI pump with a boost enrichment can mounted on it incase you want to rethink you EFI idea. I would keep the ITB's w the pump mounted in its original location and put the IC on top of every thing.

Or you could mount it off to the side like the factory did. ;)

Well, my ITBs are MFI ITBs and I do have a really nice PFI rebuilt RS spec MFI pump...I was talking to Burn Bros again and he thinks that piping up this kind of enrichment circuit isn't too hard. He thinks I should go this route.

You guys are killing me. :D

I have looked into this too. I have spec'd a similar build for the next year or so.

btw, the 997 Borg Warner VTG turbos are electronically operated. I have not been able to figure out if they are direct voltage or PWM actuated. yes, Megasquirt can run this system on either a boost value, rpm or possibly both depending on your ability to code. You would use the configurable output to run the system.

Now just need to figure out how to fit a ducktail under the decklid for 350 HP of hidden, sleeper fun!

Off to continue looking for a donor engine.

FYI, I don't think you need to wait for the sequencer to arrive. The standard MS unit should function just fine - especially considering the rpm range where you will be building boost.

I'd sell off the goodies and pick up a 84-84 3.2 long block for about $5k with the Nickisel (sp) cylinders. Already has some of the best turbo cams, intake manfold, fuel rails. Just fit J&E pistons and almost there.

Is it correct to say that if it is PWM, we need to puzzle out the frequency?

Yes, This should be easy to do on a bench with an adjustable frequency generator. Find the turbos and I can do the test. I can even start to code the .ini file for MS. This would be a custom mod to the code and I can almost guarantee there is not a specific example out there. However, the IAC and other controls work on PWM. It would be a matter of porting the code to the correct output and path.

I was looking into a mechanical system for variable boost geometry that I would drive with a servo using the MS. I am thankful for the identification of the VTG systems. Sure make life easier.

I assume since we won't use an IAC on a 911, that output could be reprovisioned for this kind of usage. Glad you can code - I work at a big software company, but my codin skills go as far as parsing code...can't write a lick, except for some basic SQL.

Yea, I know MS-II will do the trick, especially if going MFI. The only real advantage of the Sequencer is sequential injection. There are plenty of uys out there already running single spark COP set ups using run-of-the-mill v3 boards.

Now to find the turbos - this is going to be a task as I have only been able to locate one new PAIR and you don't want to know the price (as much as a Supertec rebuilt 2.7RS, soup to nuts :eek:). I am a fan of used parts and the discount they bring, as we all are. Someone out there is going to bin their 997TT eventually.

BTW, are you going with an ECU controlled BOV and pump enrichment? If so, there will oviously be even more custom coding to be done. Are there enough ouputs to control VTG, BOV and Enrichment from a v3 MS-II board?

Yes, I understand the logic, but I am a contrarian. Soapbox time... :D

Nothing grates at the fabric of my motoring being more than continual rote recital of "there's no replacement for displacement". I say there is "no replacement for a unique creation." Convention has its place for those who prefer to see a beautiful picture vs. relishing the challenge of solving the puzzle in pursuit of the picture. Both are great and totally subjective...the former just isn't for me.

My recipe: Add serious revs, mix it with positive manifold pressure, throw in some exotic alloys along with a healthy helping of currency and voila, nirvana.

The road less travelled may have more bumps and tollbooths, but the scenery is FANTASTIC. That's why it takes me 3 days to drive to California vs. 12 hours.

Hell, I may never even get to the end of a given journey, but damn wasn't it fun taking that last corner just a little too fast...

- Edit - Did I ever mention that I have a real disdain for common plenum intakes? Even Varioram. ICK.

You must use the MFI :)

You are such an enabler.

BTW, since we are completely off track and ignoring the bounds of sanity or putting limits on complexity, why not add water/methanol fogging to the mix?

http://www.snowperformance.net/product.php?pk=8

http://www.snowperformance.net/produ...ge_image-8.jpg

Could MS-II control that, too? Yep...

A decision has been made on turbo selection! I just sourced this bad boy:

Mitsubishi TD05HR-16G6-10.5T From The Lancer Evo VIII.
http://i25.photobucket.com/albums/c7...n/IMG_1160.jpg
http://i25.photobucket.com/albums/c7...n/IMG_1161.jpg
http://i25.photobucket.com/albums/c7...n/IMG_1162.jpg
http://i25.photobucket.com/albums/c7...n/IMG_1163.jpg

Here's the compressor map - it's good for over 400HP:

http://forums.pelicanparts.com/uploa...1244068173.gif

All of that power and lightning fast spool up for $300. Not bad. This is a dual scroll turbo, so I am going to have some stainless fun fabbing up the equal length pipes from my headers into the split mouth flange sometime soon:

http://forums.pelicanparts.com/uploa...1244068302.jpg

You'll notice that you need to keep the two exhaust banks completely separate until after gases exit the turbo to reap the benefits of twin scroll:

http://forums.pelicanparts.com/uploa...1244068877.jpg

Here's the O2 housing I procured used, as well. Amazing the things you can pick up used from the Sube/Evo sites for next to nothing:
http://forums.pelicanparts.com/uploa...1244068527.jpg
http://forums.pelicanparts.com/uploa...1244068539.jpg
http://forums.pelicanparts.com/uploa...1244068552.jpg

We are on our way...

Yeah, that's a nice turbo. I'm running TE05-turbine turbo on my 2L 4-cylinder clunker and it reached full boost @ 2500 RPM before I fiddled with ignition and cam timing which made it a bit laggier. But full boost around 2500-2700 RPM will be fully possible on 2.2L.

How are you going to solve water mantle cooling? A short cooler loop with electrical pump? I wouldn't run it without water.

P.S. Be prepared for slight boost taper on high revs if used with mechanical wastegate control. You can straighten up the boost on high revs with ECU-controlled PWD-operated valve but make sure you don't overspeed the shaft by asking it to boost 1 bar at 7000 RPM or such.

++1

2.1 liters, MFI, 540hp @8000 RPM's... it doesn't get much sweeter than this
http://www.youtube.com/watch?v=zWeLuglJr9I

Officially sub'd for the long haul on this one.

Good luck!

One better - one of these driving off of the passenger side cam housing:
http://i13.ebayimg.com/04/i/001/49/84/729d_1.JPG
Will use a rear fender well mounted radiator or heat exchanger.

Full ECU boost control is planned. Potential for boost taper is fine - worth the risk given the low price of the turbo. I can always upgrade the housing, compressor wheel or the entire turbo later with an Evo IX or Evo X turbo and still have it bolt up.

This is my inspiration.

Alright, the cat is out of the bag and it is time to revisit an old thread as things are are moving again. I was BARknucklehead trying to be stealth while in research mode...

I bagged the turbo above for an Evo IX version, which stock is both more efficient at spoolup and has longer legs on top. Couple that with the fact I just sent it out to Chad Block for a full performance build to BBK Lite spec and you have a turbo that makes 380WHP at 25 psi (1.7BAR) topping @7500 RPM on 93 octane in an Evo IX:

http://forums.pelicanparts.com/uploa...1270098732.jpg

Latest progress is Jamie has all of my engine guts and is busy TFL coating the bearing surfaces, as well as heat coating the combustion bits. Once those come back, the bottom end goes together.

I have also settled on Brian's 1.5" headers. Ben and Brian have been saint's in dealing with my questions, BTW!

Once the long block is done, I will need to finally commit on the induction and engine management. Electronic MFI or pure EFI? Water/meth injection to support 25psi or even...29 psi??? Oh we'll see!

Anyway, the build is not dead and very much in motion. I'll start a fresh build thread once I start wrenching!

I vote for the MFI. There have been a few people in the MFI thread talking about using a linear actuator directly on the rack to eliminate the chunk of metal ballast also called the regulator section. :)

My $0.02
I think there's nothing more beautiful than an MFI setup. But........
If you're toying with water/meth injection, varying boost levels, different turbo combinations, go with the ITB EFI. You just can't beat the programmability of a computer controlled injection system. An RSR-style ITB EFI is a thing of beauty. Besides, the plenum and IC is going to block the view to the MFI pump anyway. ;)

To be clear, my final fuelling and ignition solution is still up in the air. Once the longblock is built, I will make the final decision on where I go with it. That said, there are a few absolutes:

• I will be using my custom bored MFI stacks and TBs for this motor
• I will be using a custom water/air IC, designed and fabbed by me
• I wil be using knock control
• Water/meth injection will be used
• Visual stealth (early MFI motor) is a primary design goal

I have already conceptually laid out the IC and it won't obstruct an MFI pump. The design also allows the use of the stock airbox, with proper modifications, to remain in place for the ultimate in stealth looks and would hide the water/meth injectors, placed above the intake trumpets on the stacks.

There are still design challenges to overcome and although I do like the benefits of sequential EFI control, I am somewhat hesitant to have a big, honking fuel rail and EFI injector adapters below the TBs. Stealth is the main reason why I am considering MEFI (from this day forward, this will be my acronym for MFI w/ electronic fuel trim adjustment).

One thing at a time, though. The longblock design is sound and the rest can come after that is done.

So, stay tuned - once this litte project is out of the garage (a few more weeks), the motor build will begin. I'll also update my overall project thread on the car, but the motor will be a new thread in the Engine forum.

I'll keep you posted.

Well, well the secret is out. :)

I should have the coatings done shortly. I know you have busy with the paint project. I can't wait to see both of these projects going forward.

I think we should hide the fuel rails in the airbox for a "stealth Madonna" type setup. Then you can have the EFI and vintage looks. Crap now I gave away a secret to my 2.5 build. Then you just need to figure out how to get the stock airbox to seal against boost. I think a feax filter cover on the front with turbo inlets on the back would be cool. This shouldn't be too hard with your liquid charge cooling.

Had to happen sooner or later. :)

We should talk - I believe that I have that solution in hand.

I'm looking forward to seeing your maximum effort small bore turbo motor materialize.