Project ITBs: Part 1-parts arrive
 

Some of you already read the thread about building some individual throttle bodies. Here is the link to the first idea thread

http://forums.pelicanparts.com/porsche-911-technical-forum/179053-custom-individual-throttle-bodies-fab-your-own.html

Today the parts were waiting when I got home. Here are a few sizes and pictures of the tear down.

Here is the entire assembly as received. This is a 2001 Suzuki GSXR 600 throttle body setup. I bought it from a motorcycle parts dealer on ebay for $20. You can see that it comes with four individual throttle bodies bolted together in a complete setup.
http://forums.pelicanparts.com/uploa...1094085548.jpg

the first thing to do is strip the vacuum lines, electrical connectors and fuel rail. There are two #2 phillips screws that secure the fuel rail. These were tight but the injectors just pulled out after removal.

Next we look at the parts that are usable. This system as two internal butterflys. One set is hooked up to an electronic actuator (silver plates). I assume this is part of a choke system for low speed response. To remove lock the 17mm nut and spin off the 10mm nut shown to the right of the photo. next remove all of the small screws that secure the throttle valves. These slide out of the trottle shaft.

http://forums.pelicanparts.com/uploa...1094085826.jpg

Next its time to split the TBs. on the end plate there are two really long allen cap screws that tie all of the throttle bodies into one unit. These simply unscrew and the trottle bodies will fall apart. The throttle shaft you disassembled above will then slip out. Careful to catch the bushings and springs that set the clearance and adjustment.

http://forums.pelicanparts.com/uploa...1094085644.jpg

You are now left with 4 throttle bodies like this. There is one end unit that contains a TPS and the actuator for the secondary butterflies. I doubt I will use this and focus on the other three.
http://forums.pelicanparts.com/uploa...1094085999.jpg
http://forums.pelicanparts.com/uploa...1094086651.jpg

How do they measure up? This setup has 41mm upper openings(45mm OD) which will make them comparable to the 40IDAs that I currently have on my 3.0. I will be building a 3.2SS so this size is comparable. The TBs neck down to 38mm throttle plates and this size continues to the bottom side(44mm OD) where I will adapt to my heads. I really think this will be a fantastic match for this engine. Andy, I think you will agree here. The overall height is 83mm.

http://forums.pelicanparts.com/uploa...1094086093.jpg

Next, I will start to think of how to adapt these to fit on a 911 motor and also design a linkage. I think I am going to fab up simple plates that will bolt to the heads. I will use a PMO CIS to weber insulating plate as my template. To this will be welded a section of aluminum pipe (near 40mmID). Once I decide on the total length of the system the TBs will be added to the top.

I would like to use the stock injector ports on these TBs. I don't know if there are injectors that fit. There is a 21mm diameter with 14mm hole on the step. Can anyone help ID some injectors?

There will be more progress soon but don't look for daily updates. I am trying to finish my ST for its DE debut in 3 weeks.

Jamie

GSXR600 is basically a debored GSXR750. The injectors are Mikuni.. widely available. I'd guess you'll need more fuel though. 4 into 600 vs 6 into 3200.
Well at least your RPMs will be lower. Good luck.

Very cool! Yes Jamie, with stock ports I think those are a good match, especially for a street car. $20 seems like a cheaper than usual price for these, nice work.

what's the center-to-center measurement on these before you take them apart?

OOPs I had that measurement written on my diagram but forgot to post.

the stock ITBs come with 80mm on center spacing.

I will have to widen them a bit to get them to fit the 911 head spacing. Even less if I use a weber manifold or make my own. One main issue is how to connect the butterflies between adjacent TBs. They use their own internal spring setup. There is a setscrew on one side and an L bracket on the other. They are arranged so that the set screw seats on the L bracket. The setscrew is spring loaded and serves the purpose of syncoronzing the TBs, sorta like the majic 140mm number for MFI but much simpler.

Hi Jamie,

Thanks for that on center measurement. In the alfa or honda thread I think someone said that some of the suzuki bodies have 90mm on center spacing, as well. I don't know which ones. With the construction they have it shouldn't be too big of an issue either way.

I remember that one of the guys just lengthened the L brackets when he spaced them out by welding some stock on. I think he showed pictures, they came out nice.

I would encourage you to investigate using a carrera 3.2 fuel rail unless you already have that figured out. It should certainly have the correct spacing and if you can fit some injectors to the 'zuki TBs and to the Carrera 3.2 fuel rail it would show you exactly how to space them. This is, as you probably saw, what some of the honda guys did with things like a prelude fuel rail. It's possible, though, that there aren't injectors that fit the carrera fuel rail on one end and the 'zuki ITBs on the other, but I bet there are.

I also want to clarify my statement above regarding a good match with "stock ports". I was referring to stock early SC 39mm intake ports. We discussed this more in Jamie's initial suzuki ITB thread.

and I want to know if you're going to that chile festival!

$20 !! These are exciting threads -- and I already paid $3,000 for PMO's...

I am no expert, but my mechanic is. When I was recently looking at some third hand ITBs, we talked about sizing. His comment to me was that when you have fuel and ignition controlled by an engine management system (ie., Motec, DTA etc) you don't have to worry about air velocity like you do with carbs. So while a 40mm Weber/PMO may be correct for you engine, 40mm ITBs while workable are too small. Perhaps less than ideal is a better way of describing it. On the engines he is building, he is using 46mm ITBs on 3.2s and 52mm ITBs on 3.6/3.8s.

He has built about a dozen ITB engines in the last 4/5 years all for race applications. I guess for street use, YMMV.

Just trying to be helpful...

Richard

The guy on the Alfa board used toyota injectors with the throttle bodies. He didn't mention any additional machine work. I may contact him to see if they were a direct fit.

Custom fuel rails are relatively cheap at $10 - $12 per linear ft. of extruded aluminum. Some of the place I contacted will also machine the injector ports at any spacing requested. Those will run about $85 each.

I believe the honda guys are using honda injectors with no machine work as well.

Richard, it would be interesting to know what intake port sizes his 3.2s had. Race and street engines are definitely different, as you noted. If anything, a throttle body should be effectively bigger than an equvalent total diameter carb, because you don't have the venturi. Unless I'm missing something. I do think that you still need to worry about port velocity. Please don't think I'm slamming you but the difference between race and street may be the major factor here.

Jamie, we need to get your some clean calipers......:D

Andy,
My first impression is that the throttle body should be about the same as the carb body, since the carbs neck down in the venturi area. AKA my 2.4T motor needed ~32mm venturi, but the carb size would be a 40mm and throttle plates on the MFI are 38mm. My first impression is that the throttle plate size should be the same as the carb body sized, and have nothing to do with the venturi size, but I might just be repeating what others have said.

I think when it comes to port velocity, port size is more important than throttle plate size. I'm basing this on the MFI data, where the 2.4T used 29mm ports, but had a 38mm plate size,and the 2.7RS, which used 36mm ports still had 38mm plates.

I am fairly sure they are standard port sizes.

Certainly no slam intended or taken :) I only contributed because he specifically said that velocity is not really important compared to carbs - which seems to differ from the consensus here.

I have alot of respect for my mechanic - his training and experience are impeccable and the results his clients get are amazing. On the extreme side, he is building 800hp TT engines that last a race season not just a weekend. Among other things, he has built engines for podium finishing Le Mans 24 hour cars. I have every reason to trust him.

Anyway, I was just throwing the comment into the mix for you all to digest and consider. I have no axe to grind.

ciao,
Richard

Richard, thank you for sharing all that. Your mechanic has more experience than me, that's for sure! That surprises me that the engines had standard 3.2 ports!

I agree.

true BUT you want to make a smooth transition from the throttle plates to the ports. For this reason 45mm throttles on 39mm ports don't make much sense to me, but the T example you used is even worse. How is this transition made (smoothly?) on a 2.4T?

Andy,
Right after the throttle plates, the throttle bodies themselves neck down to 29mm from 38 just before the port on the head. The strange thing is interface from the stack to the throttle body itself isn't smooth. If I remember right the bottom of the stack on a T motor is 34mm or so and the opening of the t body is 38mm. Kinda strange but they do this consistantly between all of the stacks and t-bodies across the range.

Here's my two cents -- there are two important things to keep in mind...

TB size: If your TB chokes down to 38 mm's at the butterfly, your effective diameter there will be much less then 38 mm's since you will have the shaft running right through the widest part of the TB (ie -- across the diameter). Just guessing, the effective diameter might be down in the 35 or 36 mm area, so these TB's may not be any better then running a 40 mm weber with 36 mm chokes. Weber wasn't stupid - by putting the butterflies downstream from the venturi where the body has a 40 mm ID, they avoided this issue. But as far as peak HP is concerned, your intake is only as good as it's weakest link -- or most restrictive element. On the webers it is the venturi. According to what I understand of your TB's, it will be the butterfly area. From a peak HP perspective, whatever happens upstream and downstream of the restrictions is secondary.

The shape of the intake. Ideally, you want to have something of a cone aspect to the intake. So that the smallest ID is close to the valve. I've posted previously information about the acoustics of closed tubes versus closed cones, and for engines cones appear to have a number of benefits from a tuning perspective. As a general rule of thumb, I've seen something like a 5 degree angle to the cone is enough to get the affects. If you've got your restriction at the beginning of the intake track near the butterflies -- you don't have a cone.

I really like the idea for this project, but I think that you're starting out with TB's which are too small. In order to maintain the relative performance of your engine, you are going to want to use TB's from an engine which shares roughly the same cylinder capacity as your 3.2 -- so about 533 cc's. Figure that your 3.2 at peak RPM (say 6500 RPM) will be drawing about 1.733 million cc's or 1733 liters of air per minute through the TB. The GSXR600 has about a 150 cc cylinder. If it generates peak HP at about 10,000 RPM, you're talking about 750 liters of air per minute through the TB. Using these "back of the envelope" comparisons, it looks like the TB's are going to be pretty tight for your application.

Well, the largest four I can think of is the 1300cc Hayabusa. Don't know the dimensions of the tb's though. Either way, this is one of the more interesting threads, right up there w/ Noah's EFI conversion.

John, one question for you since your numbers add up and yet at the same time don't.

Does suzuki and porsche assume the same mm of diameter/ liter of air required in fitting their throttle bodies? It seems that the suzuki's are really large compared to the porsche versions. 38mm plates are the same size as a 2.7 with a 7300 rpm redline. that's about 440cc's per cylinder @ 6800rpm, that seems alot closer to 1733liters of air than 750. (I'm pulling all these example numbers from the top of my head)

I don't think that it's a question of either company's assumptions. One difference is that I suspect that the Suzuki spec'd considerable overhead into their TB's given a number of reasons.
1) The boundry conditions within the TB make up a larger percentage of the TB's area in a smaller TB then a larger one.
2) As engine speeds go up, pumping resistance apparently becomes a larger issue.
3) Having a relatively smaller TB on an engine will increase velocity and turbulance at lower engine speeds. Given the relatively light weight of a motorcycle, I suspect that pulling off of idle is a smaller issue in a bike then in a car. So the bike engine is most likely designed to have a relatively "peaky" motor since it is less of a handicap then in a car.

I'm not an engine designer, but those are my suspicions of what is going on.

The ITB's on the GSXR 1000 are 41 mm.

I wonder if we can come to some kind of consenus on which would be better to bolt up to the weber manifolds that I have on my 2.7, the 600 or 1000's?

Maybe I'll just try the 1000's and post my results.

Here are the sizes I collected, might not be accurate but I did my best. GSXR 1000 should be 42mm:

http://forums.pelicanparts.com/showthread.php?s=&postid=1481866&highlight=hayabus a#post1481866

just a data point...

the people at TWM Induction recomend 44's for 2.7L and 46's for 3.0L

I have no technical data to back up the validity of this information, but they guys at TWM seem to be the biggest players in the ITB game.

That's interesting data, Alan. However, this type of recommendation should really be based on the nature of the engine (cams, port sizes, exhaust, etc, etc.), and in particular whether it is a race or street engine.

maybe one of the engineers here could post the formula for velocity. I will dig up some books tonight to get some guidelines. I picked up acouple of sets when Jamie did. Since the 3.3 has 32mm ports, porting them to match 38s seems like the right thing to do. For a turbo, you want the velocity to reach a certain # much earlier in the rpm range. Then let the turbo pack it in. I will post the formula tomorrow if nobody else does.

Jamie, keep up the good work!

Seems to me that most indicators point to using a bigger ITB than 40mm on a 3.2. Even if you use the weber analogy, opinion is divided on whether 40s or 46s are best for a 3.2 - certainly on a more highly tuned 3.2 you would opt for 46s.

Fascinated to see how this project turns out :)

Richard

Wow, what great dialogue. It seems I missed a bunch yesterday. Amazing that you miss when you stay home sick and sleep off a cold all day.

Everyone here has valid points. I am not an engine builder or engineer so all this is very helpful. I agree that these may end up being too small for my 3.2. Right now, I plan to get them working on my internal stock 3.0 to set the proof of concept.

Richard, I think that a truly track engine will require larger intakes for the fuel system. Its no surprise to me that your mechanic specs bigger TBs on his race engines. I assume these engines have lots of port work too.



John, You bring up some valid points. Sounds like the Cone shape will be best and that having the TB close to the head as the limiting point on the diameter may end up in my design. It might make fabrication easier too. btw, I tried to find your old posts with the spread sheets for calculating HP from intake speed and volume. I have had no luck. Can you email me those formulas?

Aaron, The intake velocity will be:

v = volumetric flow(cm^3/min)/cross sectional area of flow(cm^2)

Yesterday, I thought about fabrication. There will be less work to space the TBs and connect them in sequence. The linkage will be easier to setup if I mount these to the weber manifolds. The downside is that I will have to make mounts to hook up with the ball/socket arrangement. A cable operated setup may be better and easier to attach with the current config on the TBs.

Mounting them directly to the heads will be more difficult, especially since I no longer have access to a mill or lathe after changing jobs a few months ago.

This is great fun. Keep up the suggestions. Everything is being considered.

Jamie, I think you should definitely use the 38mm TBs for your proof cf concept on your internally stock 3.0. You'll learn a lot about how you want to make these work, and whether you want characteristics found in suzuki throttle bodies other than the ones you have.

I'm still not convinced that the 38mm bodies are a terrible choice for this street engine. Anyway figuring out how to use these is the hard part, not selecting the diameter. In particular, you did well by spending just a little for your proof of concept. it seems that the suzuki throttle body "architecture" is pretty consistent, so you should be able to apply your experience to another set if necessary..

I can't visualize how the linkage will hook up. maybe you can draw a diagram or take some photos as you get closer to setting that up. It would be cool to hook directly to the heads, but getting this set up in some way is the most important part. Did you get the rubber boots with the TBs? They appeared to hook right up to the alfa ports, which don't look too different than porsche ports.

Get well soon, Jamie!

The required size is pretty much about the amount of HP needed, so using the HP per cylinder of the bike as a guide should let you know if it will be suitable for your monster 911 project!

I will try and get some pics of potential linkage mounts this weekend. I need to measure the throttle shafts and see if they are the same as the webers. That would make finding linkage parts easy - especially if they are the same size as the VW versions.

From "Maximum Boost" by Corky Bell

"For race applications where one throttle plate per runner is employed, it is quite adequate to sum the throttle plate areas and calculate accordingly, or simply to use one cylinder and one throttle. The 300ft/sec figure should still offer a suitable guideline."

"Simply making the throttle bigger will alleviate the problem, but low speed driveability can become an on off, jumpy proposition. A big throttle plate open a small amountcan let in a lot of air, and smooth low-speed throttle respnse will suffer. A maximum air velocity of approximately 300 ft/sec will keep flow losses acceptable."


I don't know if 300 ft/sec works for n/a engine, but I thought I would post.

40mm ports at 6Krpm is about 65 ft/s. Should be within the limit of a boosted engine. What does the target of 300ft/s give you? throttle response? cylinder packing/filling efficiency? air column momentum? Just looking for more context.

Funny you bring up the large TB opening and letting a small amount of air in. I chased a high idle problem when I first started up the 3.0. Turns out that two plates were out of adjustment and not closing. they only let a sliver of light through on one bank and it was enough to drive the engine at 3K rpm. This example just shows that it doesn't take much air to get things going.

Jamie

My guess is for throttle response. He doesn't get specific as to his findings. I'll look around for books on intake design. That should hold more info.

What is the dia of you original throttle? That should be a good starting point.

I am currently running weber 40IDAs. So 40mm is the trottle body. This size should also be good for a 3.2 unless its a full race application. the 38mm throttle plates should not give up too much for this first shot at building them.

Jamie;
For street use, I agree that 38 mm TB's should fine for a 3.0. As far a HP/port size formula, I'm not aware of one -- although some would disagree. I've rarely found a non-CIS's 911 (carb'd, MFI'd) that generated peak HP much above 100 m/s (about 300 f/s). I've come to the conclusion that peak torque gas speed is more important. In the case of non-CIS'd 911 it tends to occur at around 80 m/s. If the speed is much below 60 m/s you start to lose low end performance. Above 80 m/s it seems that torque tends to trend downward (off set by the increase in rev's) until the peak HP engine speed.

Where you put the butteflies and the injectors also makes a difference. In general, the further upstream that you place the injectors, the better the performance at peak rev's. If you put the injectors down close to the head, the engine runs better at low revs.

I don't have any definitive conclusions on where you put the butterflies except that I believe that the total volume of air downstream of the butterflies will have some sort of affect on idle quality. But if they are no closer or further then the standard MFI butterflies, it shouldn't be an issue.

Now if someone like a Steve Weiner weighs in here, I'd take his $0.02 over mine. I'm not of any observations that agree with Corky's conclusions. I don't think that it's a safe assumption that a common intake of 6x is the same as 6 separate intakes of x diameter or area. Look at the performance the smaller Le Mans Prototypes who have a fairly small single intake restricter of about 32-35 mm's.

Check out these two threads for some explanation of what I'm trying to describe. Here are the rest of the videos. They may start to give you something of an idea of the way that intake designs work.

I've run across some info on the GSXR 750 TB's for anyone that is interested. I am looking for similar info on the 600's

There was a change in configuration from '99 to '00. Something to check on when buying two sets of TB's. It may be a good idea to buy from the same model year or confirm that the design is the same between years.

http://forums.pelicanparts.com/uploa...1094239725.jpg

http://forums.pelicanparts.com/uploa...1094239831.jpg

Thanks Mike, I know there are many changes year to year on these systems. What I measured was different than the info I found on the web. I guess a factory manual would be the only way to find correct info, unless you happen to know a good Suzuki tech/engineer.

John, That information about intake speed vs Hp/torque was what I referred to previously. I guess I was searching for the wrong topics. Now I can go back and reread the previous threads.

Jamie, the main throttles are in the wider part of the cone, right?

John, That 911R inspired thread was very interesting, thank you for your research. I will probably ressurect that thread once I get some time to think about it.

Best Regards,

No, the primary plate is in the narrow section, at the bottom. The secondary plates, which I have not figured out yet, are in the wide section. Interestingly, the secondary plates have a diameter that is about 4mm smaller than the body at that point, 2mm/side.

That's odd. it seems like the injector would spray against the throttle plate at WOT, from the picture I posted in the other thread. it might be important to use injectors with a similar spray pattern if you replace the suzuki ones. You ought to have more HP/cyl than the suzuki even with a stock 3.0, but sometimes there's lots of margin.

here's the picture I was thinking of. It's a 1000 throttle body (of some year, not sure which year), maybe Jamie could note if there's any significant difference versus his.'

http://forums.pelicanparts.com/uploa...1093539275.jpg

I believe the secondaries are computer contolled . There is a vacuum sensor also.Now here is the leap I made, the throttle bodies are oversized to take advantage of high rpm performance. The computer monitors vac and other sensors and modulates the secondaries to keep velocity at the lower rpm range. Sorta like varioram....are you buying it? 38mm throttle bodies are pretty big for a 600cc engine.