Another way to mix fuel and air

[HoustonZ]

[Tony D]

Swap the location of the throttle valves and the injectors, and I'm in for a set...

[josh817]

Tony, I thought you and I already have something similar. Why do you do this to me?

 

Edited January 15, 2011 by josh817

[Tony D]

Ever take a walk through a CART engine shop and see the barrel throttles built into the heads?

 

The most responsive throttle is that which is closest to the valves.

 

And as we have discussed in the past....there's a reason F1 Engines inject their fuel above the velocity stack.

 

Curiously nobody takes cues from these guys, and make beautiful things...but they could be functionally so much more....well, functional!

 

When was having one of anything...or two for that matter, enough for me? I could find a place for a set if they had close to the head throttle plates and injector bridges at the horns...

[TimZ]

Ever take a walk through a CART engine shop and see the barrel throttles built into the heads?

 

The most responsive throttle is that which is closest to the valves.

 

And as we have discussed in the past....there's a reason F1 Engines inject their fuel above the velocity stack.

 

Curiously nobody takes cues from these guys, and make beautiful things...but they could be functionally so much more....well, functional!

 

When was having one of anything...or two for that matter, enough for me? I could find a place for a set if they had close to the head throttle plates and injector bridges at the horns...

 

I guess I've never tried it, but it seems like this arrangement would work great when the throttle is open, but might have issues when sitting at a stoplight for an extended period of time with the injectors spraying fuel at closed throttle plates (something an F1 engine pretty much never has to do). Am I thinking of this wrong, or were you thinking more of a racing application where an 1800rpm idle from a slightly propped open throttle plate is no big deal?

[Tony D]

Yes you are thinking of it 'wrong'.

 

Think of the Ford or GM TBI setups and tell me how they differ from the F1 setup...other than the horsepower. How does that setup get air to run the engine? There is an IAC on there, but it's admitting air nowhere near where it will allow fuel to mix into it's bypass of the throttle plate...

 

Even using a stock SU body and closing the throttle and spraying a 1000cc injector to the top of the wall where the tau layer forms...it STILL idles fine. I don't know how, and I swore it would never work... but it did. I changed my thoughts about how closed closed throttles really were after that eye-opener.

 

And propped-open throttles is a relative thing. I don't think F1 idles that low. That would be like our engines idling at 250 to 300 rpms. You don't need much of an opening to allow an engine with really well atomized fuel to idle.

 

The head placement of the injectors is purely emissions-driven. Performance applications move it from there, further upstream when practical to do so. And the plates as close to the valves as possible for throttle response.

 

This makes the air door function very crisp, and if you have a large layer of fuel wetting the manifold it responds to transients better. But I digress...

[josh817]

Lol TimZ I thought the same and then Tony hit me in the head with a hammer. A good example of it working is also the BMW CSL with the mechanical injection. They spray at an angle from the underside of the barrel upstream of the throttle plate:

 

I was reading on a Honda forum about some Euro race teams stepped injection system and they said the best placement for the injector that sits outside of the air horn is in the center and 3/4" outside of it as that is where the best flow is so fuel will be atomized well.

 

The Hilborn system the throttle plates are about 6" up, plus another 3 or 4" for the port length? I think when I crunched the numbers (http://www.grapeaperacing.com/tech/index.cfm), I came up with a tuned length of 11 or 12".

 

Cam duration: 300º

Pressure wave velocity: 1300ft/sec

Peak HP: 7000 RPM

Runner diameter: 1.8125"

Tuned to the 3rd wave yields 12.7" length. With a 310º duration it goes to 13.2". The write up says to subtract 20-30º from the total duration so with that said, this really only get's you into the ball park. The same goes for the other variables below. The way I may do it is to get a little bit of dyno time and cut sets of PVC pipes, or metal whichever is cheaper, to certain lengths. Do a run, replace pipes, do another, etc.

 

1st and 2nd waves are apparently more prevalent but you need a substantially longer runner. With the same settings as posted above tuned to the 2nd wave the length jumps to 19.5" and the 1st goes to 40".

 

Another way to reduce tuned length is to lower the velocity of the pressure wave which is essentially the velocity of sound. The 1300ft/sec figure is at a 245ºF. If you could get the intake charge to 140º you would have a runner length of 11.7". I don't know if that is realistic though, I don't know what sort of temperatures even occur now so it may be for good or for worse.

 

So, anyway, after all that another thing I read, and it may be a false scare, is that when you set the injectors out like we're discussing there is a lag between when the injector sprays and when the valve is open thus making it unresponsive and bog perhaps. If this is true then hopefully there is a way to advance it. In my case, hopefully Megasquirt has some sort of option to take care of it if it is a real problem.

 

Also, from the last post you have there Tony, it sort of sounds like your talking about setting the idle by opening the plates. I thought you had said before its a bad idea to do that, so I was going to introduce a vacuum leak by means of air valves used to synchronize the barrels like Derek's setup.

Edited January 17, 2011 by josh817

[Tony D]

No, I never said that, I said experience with 'closed' throttle plates made me rethink how closed they really were...

 

There was a good article from someone at Ford of Europe regarding injector placement on one of their engines, and it was outside the air horn, spraying 5 degrees obliquely to the wall or some such thing. Gist of the article was outside is better, and each engine and manifold will have an optimal angle best found through trial-and-error.

[Derek]

My throttle plates are pretty much air tight and only gets idle air from my IAC block and it idles fine. It's mounted about 18" of 1/2" tube away.

 

Derek

[josh817]

I'm sorry, I was getting the idea that you were inferring that.

 

 

Edit:

Derek I'm going to go back and look at your thread. I'm starting to think my original plan isn't going to work.

Edited January 16, 2011 by josh817

[Tony D]

Derek, you are injecting at the head, the question had moved to injecting at the air horn above the throttle plate and how it would idle.

[Derek]

I just can't keep up any more!

So spraying fuel down a nearly blocked venturi and introducing air through a single point on a runner that's common with the 6 inlets is going to idle?

It seems to me injecting at the head for idle and off idle and transitioning over to the other set of injectors in the ends of the horns would be a more streetable option. Don't know if Megasquirt can do that though.

 

Getting 6 ITBs to come off idle smoothly is really tricky. At least on my car it is.

 

Derek

[josh817]

Megasquirt has a staged injection option, which is what you are talking about. Almost all the setups I have seen on Youtube has a staged system for the reason you said and because apparently larger injectors are too difficult to control at idle speeds, plus the atomization factor which as you know at higher velocities works well.

 

 

My project won't be staged injected, however this isn't my thread so whichever route this discussion goes...

Edited January 16, 2011 by josh817

[rayaapp2]

I may be missing something here related to fuel programing, but I dont see why there would be an issue with injecting up stream from the throttle valve.

 

Take a good look at GM TBI(Not an ITB example but same basic principle). TBI injection systems with injectors placed in the velocity stack of the housing. Talk about a crud but effective system. I think injector nozzle angle, position, and the length of the stacks are probably going to assist in tuning out issues and probably more so more than programing will get ya, but only trial and error will tell. It seems like a very workable setup in theory to me, there is just a lot of trial and error in setting up a good system for a particular engine build(More than I even have a grasp of I think).

 

EDIT:

damn, Somehow I missed that part where Tony already mentioned the GM TBI setups...

Edited January 16, 2011 by rayaapp2

[josh817]

I mean its the same deal in a carb with fuel in front of the throttle valve

For the sake of conversation, not necessarily. We're talking about at idle how does fuel get there because the throttle plates are closed and the injectors are outside of plates. With a carb, the idle jet is between the plate and the head, just like Derek's injectors and like the manifold posted above.

 

After reading in Derek's thread, I'm now confused on whether I should utilize the 1/8 NPT to synchronize the runners or to put barbed fittings in and make a balance pipe. If I don't care much about idle, and if after a very minute gap of a cracked throttle (Tony said .030" maybe) the air bleed sync's are moot, would one even need them? Unlike Derek's setup, all the throttle plates for my manifold are on one shaft so I don't have to worry about one plate being 90% open while the other is 100% open.

 

Like you said rayaapp2, its a lot of trial and error so I'll probably have to build this thing and then work out the rest.

Edited January 16, 2011 by josh817

[rayaapp2]

For the sake of conversation, not necessarily. We're talking about at idle how does fuel get there because the throttle plates are closed and the injectors are outside of plates. With a carb, the idle jet is between the plate and the head, just like Derek's injectors and like the manifold posted above.

 

After reading in Derek's thread, I'm now confused on whether I should utilize the 1/8 NPT to synchronize the runners or to put barbed fittings in and make a balance pipe. If I don't care much about idle, and if after a very minute gap of a cracked throttle (Tony said .030" maybe) the air bleed sync's are moot, would one even need them? Unlike Derek's setup, all the throttle plates for my manifold are on one shaft so I don't have to worry about one plate being 90% open while the other is 100% open.

 

Like you said rayaapp2, its a lot of trial and error so I'll probably have to build this thing and then work out the rest.

 

Yeah, I edited the post after thinking about it for a moment. My theory still stands though with the TBI setup from GM. At idle they are dumping fuel right on top of the throttle valve. I dont recall if the valve is cracked or not on the GM setup.

 

"Think of the Ford or GM TBI setups and tell me how they differ from the F1 setup...other than the horsepower. How does that setup get air to run the engine? There is an IAC on there, but it's admitting air nowhere near where it will allow fuel to mix into it's bypass of the throttle plate...

 

Even using a stock SU body and closing the throttle and spraying a 1000cc injector to the top of the wall where the tau layer forms...it STILL idles fine. I don't know how, and I swore it would never work... but it did. I changed my thoughts about how closed closed throttles really were after that eye-opener."

 

And propped-open throttles is a relative thing. I don't think F1 idles that low. That would be like our engines idling at 250 to 300 rpms. You don't need much of an opening to allow an engine with really well atomized fuel to idle."

 

"experience with 'closed' throttle plates made me rethink how closed they really were..."

 

In other words you will need to calibrate the leak at each valve by the largest and crack the others to match then set your idle there. Atmospheric pressure is trying to get in there despite what seal YOU think is there and atomized fuel is going to follow suit.

 

 

I actually pm'd Derek about his vacuum log and balance tube as his thread isnt clear how he has it setup. He has a balance tube cast into his "Horns" and I think he uses the vacuum log for idle control and sync between the 2 horn castings.

Edited January 16, 2011 by rayaapp2

[z-ya]

Megasquirt has a staged injection option, which is what you are talking about. Almost all the setups I have seen on Youtube has a staged system for the reason you said and because apparently larger injectors are too difficult to control at idle speeds, plus the atomization factor which as you know at higher velocities works well.

 

 

My project won't be staged injected, however this isn't my thread so whichever route this discussion goes...

 

I've never seen MS use staged injection. Most all I've seen (and tuned) are batch fire.

 

My turbo car is a different story. I use the staged injection on the Wolf 3D ECU. So at idle, it is just the 6 36lb injectors in the stock location. At 30% load I have two more 36lb ahead of the TB that gradually start opening. At WOT the two additional 36lb injectors are running at the same duty cycle as the primary 6 injectors.

 

Most all modern injectors are designed to have a non linear flow rate vs. opening time. So at low duty cycles, small amounts of fuel can be delivered from somewhat large injectors. Not a lot of fuel comes out at let say a 1.5ms (1.5 thousands of a second) opening time. So it doesn't surprise me that injectors upstream of the throttle plate seem to idle just fine.

[josh817]

Yah I'm on the same grounds for the GM TBI too. As far as MS is concerned, I'm definitely no expert on it so chances are I'm wrong but I remember seeing a picture of the screen from somewhere about it. A quick search also yielded results for like dual tables perhaps for methanol injection use. Or even like Tony said a while ago about having 2 ECU's so after a certain RPM the other one starts controlling another set. Well, you know all this, we were talking about it in your thread. Sorry. Anyway, yes, the general idea was also that larger modern injectors will work fine. Really, for my N/A application, my "large" isn't all that big to begin with.

 

This is what I have to work with:

 

As I said before I'm confused on what I should do now as far as air bleed valves to synchronize the runners. I also don't know how well brakes will work without running the booster. With the bigger master cyl., Toyota calipers in the front and 240sx calipers in the rear, the pedal may be really spongy without the booster. So, if I don't need air bleeds to sync, then I will make balance pipe and hook up the booster. If I need air bleed sync's and my brakes still suck I guess I will have to drill and tap the manifold which makes me nervous! I'm using TPS and not a MAP so a vacuum log like Derek's isn't needed unless the tiny tubing to a log is enough to "balance" the runners and utilize the booster from a vacuum log.

 

Oh and also Rayaapp, adjusting the clearance between the throttle plate and the runner wall is indeed how Hilborn says to synchronize their system, even with EFI. In fact a video by them shows the guys synchronizing a V8 (so you have 4 bodies) and when he got 7 of the 8 to match, he took the last throttle plate and tapped on it with a hammer until it came out right................ yah.

Edited January 16, 2011 by josh817

[Tony D]

I got a simple request:

STOP WITH THE "TONY SAID" because I don't EVEN want to go back and correct the crap that has been misquoted here in the past 6 posts and attributed to me.

 

"0.030"? TRY: 0.003"

 

I'm to busy now, just take anything attributed to me and delete it so I don't have to come back on a rant and try to fix the bad misquotations / inferences / etc...

 

If you quote me, LINK THE POST! Otherwise, I suggest you do not use me or your recollection of my posts as a talking point.

 

[Tony D]

Ray is on to the train of thought I got towards with the GMEFI.

 

Traditional single TB's use a sealant around the TB circumfrence. GM TBI's do not, neither do carbs or Hillborns.

 

What Ray stated was correct: the seal isn't the greatest and even with them 'closed' they will leak a considerable amount of air to the engine, and this is more than enough to let the engine idle. And more than enough to let in the fuel. Just like GM TBI.

 

That the IAC exists elsewhere (and that people are referencing it incorrectly) confuses the fact. IAC motor is NOT for ALL the air for idle. It is for a RANGE of Idle Speed TRIM, like when the the alternator or A/C comes on and puts a load on the engine at idle. Or for fast idle compensation. Having under 0.003" on the throttle plates will have the engine around 2000rpms at idle, and if you have fine enough adjusting screws you will be able to close it further for a lower 'street' idle---this will not be an issue. The throttle plates necessarily will be propped open NOT for purposes of adjusting idle speed, they will be opened to SYNCHRONISE them.

 

As mentioned, once you got the majority of them flowing similarlay and the idle speed as low as possible (then up it with your IAC TRIM if so equipped, or master bypass screw into the balance tube) then you tweak the throttles remaining to flow similarly. You are not adjusting the idle speed with the opening, you are synchronising airflow to the engine at idle---VERY DIFFERENT!

 

If they are synched properly at idle, and all linkages are identical and unbinding, all other synch checks to WOT should be identical on all barrels and you're done.

 

Remember, synch is done and then the whole thing is adjusted down to the absolute lowest possible idiling speed as opposed to the curb idle speed like on a carburettor. Sometimes you will synch directly to the lowest speed possible if you have a meter with enough resolution. This is because final idle speed is set AFTER synchronisation is set. And once they're set, the ONLY thing that should require adjustment is the bypass screw to compensate for wear on the engine, or crud buildup in the passages. Cleaning the throttle plates regularly to keep crud from building up and obscuring the synch is important as well, especially if you are not running a filter for intimidation purposes...