Update on Roller Barrel ITB project for STI

[grimm]

good to see it's nearly done, now i know you've taken your time with things and have been trying to design and fabricate the system from scratch, or at lesast just from ideas. So my question is, what have you learned? what would you do differently?

[wgknestrick]

Originally Posted by grimm

good to see it's nearly done, now i know you've taken your time with things and have been trying to design and fabricate the system from scratch, or at lesast just from ideas. So my question is, what have you learned? what would you do differently?

I wish I had found this site first: www.extrudabody.com. That would've saved me a little money. The injector locations really hurt the overall design because of space constraints. I have some close ups of the fuel rails that I can post when I get home, but I really cornered myself (space wise) by placing them so close the heads. The STI heads have a big solenoid that is integral to the head that made a space sandwitch out of the fuel rails and the ITBs.

Once I start getting a little further along, I will surely see other things that I wish I'd done differently.

Right now, have 0 vacuum at idle is proving to be a problem. I need to start removing assessories from the vacuum manifold/accumulator to see where I'm loosing it all - PCV is the first place I'm looking. I expected to loose "some" vacuum and resolution off boost, but not "all". If the PCV doesn't help at all, I am trying restrictors in the vac lines to each port.

I also think the car will still idle with the barrels all the way close (they aren't air tight then due to clearance). This should further help idle (TB opening is always biggest vac leak). The vacuum manifold would then supply the air instead of each TB and average across all the TB openings. This should function more like a single TB system than a multi at idle.

We will see, but getting it running was a major milestone. Refinement will be an on-going process.

[shvrdavid]

Originally Posted by wgknestrick

Right now, have 0 vacuum at idle is proving to be a problem. I need to start removing assessories from the vacuum manifold/accumulator to see where I'm loosing it all - PCV is the first place I'm looking. I expected to loose "some" vacuum and resolution off boost, but not "all". If the PCV doesn't help at all, I am trying restrictors in the vac lines to each port.

I don't think that the assessories are much of the problem...
I think the biggest problem is the lack of volume between the throttle bodies and the valves...
(Keep in mind that all a vacuum is, is a lack of air pressure within a space or volume) Your design doesn't have much volume to create a vacuum in...
With the stock manifold, you have all 4 runners and the plenium combined to create the vacuum in... With your setup, you only have a few inches to create the same effect...
The problem that crops up with a setup like this is time...
I know, sounds strange, but that is the problem... With the idle at the factory setting, there is too much time to refill the space between the throttle and the valves...

There are a few ways around this, that may not be in your design, but could be added to it...

The first thing would be to get a vacuum pump to run all vacuum assessories, freeing the engine from the task completely...

The second thing is to move the injectors to above the throttle body to form a liquid seal in the throttlebodies... If you look at many F1 designs, the injectors are in the plenium, or ontop of open stacks (to form a liquid seal at the throttlebodies, and to raise the mach index of flow, within the runner)...

The third thing is to raise the idle... Probably not what you want to hear, but you may have to do that to get any vacuum that will last untill the next intake valve event...

The last thing is to program the car completely differently at idle, sort of like antilag, to promote even more cylinder scavanging at lower rpms...(timing after tdc/rich/etc) You can probably do it with the Hydra, it is a very user friendly system in speed density mode... Ditch the MAF base if you are using that now, your going to need speed density and some creativity to program this setup...

Bill your setup will work, it just needs to be looked at a little differently from the tuning department...
Pm me if you have more questions...

Scott

[wgknestrick]

Originally Posted by shvrdavid

I don't think that the assessories are much of the problem...
I think the biggest problem is the lack of volume between the throttle bodies and the valves...
(Keep in mind that all a vacuum is, is a lack of air pressure within a space or volume) Your design doesn't have much volume to create a vacuum in...
With the stock manifold, you have all 4 runners and the plenium combined to create the vacuum in... With your setup, you only have a few inches to create the same effect...
The problem that crops up with a setup like this is time...
I know, sounds strange, but that is the problem... With the idle at the factory setting, there is too much time to refill the space between the throttle and the valves...

There are a few ways around this, that may not be in your design, but could be added to it...

The first thing would be to get a vacuum pump to run all vacuum assessories, freeing the engine from the task completely...

The second thing is to move the injectors to above the throttle body to form a liquid seal in the throttlebodies... If you look at many F1 designs, the injectors are in the plenium, or ontop of open stacks (to form a liquid seal at the throttlebodies, and to raise the mach index of flow, within the runner)...

The third thing is to raise the idle... Probably not what you want to hear, but you may have to do that to get any vacuum that will last untill the next intake valve event...

The last thing is to program the car completely differently at idle, sort of like antilag, to promote even more cylinder scavanging at lower rpms...(timing after tdc/rich/etc) You can probably do it with the Hydra, it is a very user friendly system in speed density mode... Ditch the MAF base if you are using that now, your going to need speed density and some creativity to program this setup...

Bill your setup will work, it just needs to be looked at a little differently from the tuning department...
Pm me if you have more questions...

Scott

Thanks for trying to help me with this project, but I don't really agree with most of your suggestions or theories with the exception that the car needs to be tuned to maximize vacuum (untuned now). I certainly don't want this to go into an arguing match, but rather present my thoughts to you in an effort for you to learn something out of this. I don't claim to be the authority.

Vacuum is dependent on air leaks, period, not volume in intakes. To maximize your vacuum, you must minimize your leaks, and you WILL always leak for the engine to idle (it does need air to run). The TB is usually the biggest leak in all systems. ITBs have low vac not because of the small volume, but because they have 4,6,8, or 12 times the TBs leaking. Each ITB is open the same amount as one single was open to flow the same amount of idle air to the engine. No matter what volume you are dealing with, you will not draw a vac if you leak more in than what the piston can draw. You will always loose vac faster than you can pull it with an ITB system unless you are "tricky". I am trying being "tricky" next week.

Vac is important to me and many others not because it is helpful to run petty accessories, but because it is critical to having resolution within an ECU to tune different loads. If I have 0 vac at idle, that means there is 1 load column of resolution over what used to have 15. I can only tune 1 amount of fuel vs RPM that has to be sufficient to fuel many different load conditions until the turbo can build boost. This is a nightmare because most daily driving is done in these vac parts of the map.

Moving the injectors above the TB is just plane silly if you aren't reving to around or above 10K rpm. Even then, it is a safety nightmare and requires a really high idle speed to keep the fuel suspended in the air. One backfire and it's pyro time in your engine bay with these. These are ONLY found (cannot stress ONLY any more) on race cars. This along with pneumatic valves are the reason F1 cars have to idle at 5K RPM. I am thinking about shifting at this point in a Subaru.

Yes I will have to raise the idle as it will provide more vac, you are right.

"Ditching the MAF" is certainly not applicable to me because of the Hydra. Hydra uses MAP not MAF for loading (MAF is already "ditched"). Actually being able to use a MAF is very helpful because they measure the air flow (not pressure) into the ITBs. Idle vac is much less of a worry then. Nissan uses MAF on both the Skyline GTR and Pulsar engines stock and this is why. Both of these engines have ITBs and turbos from the factory.

Once again, I really appreciate the effort in trying to help me and I hope I was able to teach you something in a positive way. The internet should be used for learning, not arguing.

[shvrdavid]

I was under the impression that the Hydra could be set up as MAP or MAF...
I could be wrong on that.... I'll ask Phil next time I talk to him...

I dont agree with you statement about injectors on top of the throttle plates...
TBI systems had this for years on tons of cars, and did not have the pyro problems, or the fuel distribution/idle problems you mentioned... Actually, it made it better...

Most of what I mentioned in my prior post was based on previous experience with "different setups"... Your setup falls right into that description...

I worked with a design that was very similar to yours years ago and ran into the same problem with the first couple of setups... And we did find a way to increase the vacuum by making changes to the rollers and the barrels to make it seal up better... We also moved the barrels to the top of the runners just under the bell, and the engine held more vacuum... The only difference was the volume of the runner between the barrels and the valves... More volume takes longer to refill with the same bleed rate, that is plain physics...

I can understand your point about resolution in the maps as far as setting the mixture and the timing... That can be a major problem, but can be delt with if the ecu can fake a map signal based on tps... And I think that the Hydra can do that as well, I am not sure if it is a 2d or a 3d map fake or not......

I am not here to argue, I just thought I would give you some food for thought...

Everybody does things differently when a problem presents itself, and I am curious as to your fix for this problem... Your idea may well be a better fix than I ever thought of... And I will be the first to tell you that if it is...
You have obviously put a lot of thought and time into this project, and it shows...

Keep working on it, and you will get it all sorted out...

[wgknestrick]

I am trying check valves on the port feeds to the vacuum manifold once I get them. My theory is to install 3 (that number is critical and I will explain why it isn't 4) check valves so that they only allow their cylinders to contribute vacuum, but prevent any vac from escaping back through them. 1 port is left open to allow positive pressure into the manifold so it and therefor the Hydra still see the full spectrum of pressures + and -. This should "in theory" function very similar to a single TB setup with respect to vacuum since there is now only 1 TB that can have vacuum escape.

I will see early next week how well this works.

[wgknestrick]

Project is moving along well. I had a small problem with the coolant overflow tank and it's lines that was hopefully solved today. Evidently the whole cooling system functionality is dependent on the small line going from this to the top of the radiator inlet. The car will hold temperature very well now and I can't even feel the radiator getting warm (hopefully that is a good thing).

Air velocity meter arrived and works pretty well to give air velocity over each ITB at idle for balancing. I think I have it within +-2% now and haven't even spent much time on it. I just don't know when or what level is "good enough".

I just have to fix a small leak (rrg leaks!) on the PS pump fittings, and install an ISC valve so that I don't have to manually choke the car on starts. Now that I have most things running smoothly, it is obvious that I need to tune the throttle pump enrichment as the car wants to climb RPM so fast, the old enrichment settings can't keep up and cause the car to have hesitation from lean out on throttle blips.

My "trick" with the check valves got idle vac up to 15-18inHg once I disconnected the PCV system from the vac accumulator!!!!! There shouldn't be any more huge stumbling blocks ahead of me except getting the boost leaks taken care of. This vac issue was the only thing worrying me because there really wasn't much else to do if that didn't work. This check valve trick should make running ITBs on ANY speed density only ECU cake. Idle vac is almost identical to what it was with the stock manifold and single TB. Hopefully next week, KevinWRX can stop by with his digital video camera so we can see it in "action" (by action, I mean running in my garage).

The Subaru engine doesn't seem to sound much different than a Honda with ITBs (intake noise wise). The exhaust still sounds like a Suby, but with the manifolds off, it has the typical ITB sound on revs. Almost reminds me of an old VW bug with multi carbs. This is almost gone when the manifolds are installed too. I was hoping for something magical noise wise, but I guess this is it.

It does sound bad-ass on starts when the idle jumps up to 3K RPM instantly then falls right back down to 1.5RPM (high idle speed for now).

Bill

[Bobert540]

Glad to see you are making good progress. I need to stop over and see this thing. Last time I saw it the manifolds were still in pieces. Keep up the good work!

[shvrdavid]

I am glad to hear that you have vacuum now, that cures that part of the tuning problem...
The tip in will be a little tricky to get right...
You will have to change multiple maps to dial it in...
Tip in, iat correction, and coolant temp correction come to mind...

I would think that 2% flow difference is a very good spread, cudos to the person that did the machine work, and to your design...

I am wondering why you would need an isc valve with a drive by wire setup...
I was under the impression that the hydra can do that all by itself...

Don't be to upset about a different sound...
Your setup will probably sound different because of the seperate plenums, that will be pulsing independently...

The design in itself, should make for a very distinct sound under boost...
Think of it as sounding sort of like it did before, but is stereo, and a deeper tone...
Your getting there, cant wait to hear it when it is running... I might have to come up and hear and see it myself...

[Ender81]

Originally Posted by wgknestrick

Almost reminds me of an old VW bug with multi carbs. This is almost gone when the manifolds are installed too.
Bill

That would be awesome if it stayed like that because I have a VW bug that I want to throw dual carbs on hahaha.

[wgknestrick]

Originally Posted by shvrdavid

I am glad to hear that you have vacuum now, that cures that part of the tuning problem...
The tip in will be a little tricky to get right...
You will have to change multiple maps to dial it in...
Tip in, iat correction, and coolant temp correction come to mind...

I would think that 2% flow difference is a very good spread, cudos to the person that did the machine work, and to your design...

I am wondering why you would need an isc valve with a drive by wire setup...
I was under the impression that the hydra can do that all by itself...

Don't be to upset about a different sound...
Your setup will probably sound different because of the seperate plenums, that will be pulsing independently...

The design in itself, should make for a very distinct sound under boost...
Think of it as sounding sort of like it did before, but is stereo, and a deeper tone...
Your getting there, cant wait to hear it when it is running... I might have to come up and hear and see it myself...

DBW was abondoned because the STI TB motor is super weak. It was taking about 10sec just to reach 4K RPM when I stepped on the gas. It idled the car superbly, but I would run into safety measures if I kept the motor in place and used the cable to "help" it.

It has a nice throttle cable now that I attached to the STI pedal assembly and reaches 4K RPM almost instantly.

Machine work actually wasn't very accurate to start, but (given my background and profession in molding & tooling) I was able to hand work the barrels into this state. I will probably keep refining them, but it is a tedious process to hand file, then measure, rinse and repeat (we are talking thousands of an inch here). The ports were a little off axis when machined in the barrels, and is the problem. Basically, I insert a .010" feeler gage in the largest throttle opening, then file the other to match, .001" at a time.

The problem is that .010" open on the top doesn't equal .010" open on the bottom (underside) of the ITB. On one side it is closed underneath, and the other is open .025". I am forced to use the meter to balance with this situation.

[wgknestrick]

Car is going together pretty well, but boost leaks once again are fighting me. If I can get these solved, the car only has the meth injection system for me to install, then off to tuning. I am pretty happy with how everything came out so far.

Bill

[SpoolinDSM]

Dude, that looks ****ing amazing

[gary]

So any ETA for its arrival at a meet? You still owe me a ride.

[AREA1320]

shouldn't you be at a wedding slacker??

[wgknestrick]

http://videos.streetfire.net/categor...c2006678b3.htm

Someone just posted some new footage of the original Jun Hyper Lemon for which this engine bay is based off of.

[Ender81]

Damn that car is incredible.

[Honest_Bob]

I cant wait to see this thing in person.

[wgknestrick]

Got it all back together yesterday and arrggggg more leaks. They are getting less and less, but this is certainly frustrating. Nothing like ripping everything apart, reassembling and finding out you still have a leak.

[shvrdavid]

Do you know anyone that has an evap smoke machine???
If you can find one, fill the system with smoke, then pressurize it...
You can find the smallest leaks with the smoke, and a very bright light... It isn't as messy as soapy water either....

Sounds like you have most of the bugs worked out of your setup already...
I am curious to see if all of the time and planning you put into this setup, shows up on the dyno... Something tells me it will...
I can tell you from past experience that the only way to tell is to strap it to a dyno, and tune it...
And that that is the part that I always looked forward too the most in my builds...

You are almost to that point now...

Before you know it, you will be driving your car again...
And I can imagine the grin on your face the first time you drive it after it is tuned...

I want to see this in person when it is done... You have put a lot of time and thought into this setup, and it shows...