Poor man's ITB fuel injection for a Sierra engine....

Fireblade

Fireblade

Active member
Fellow Frogger
Joined
Mar 22, 2012
Messages
564
Location
Melbourne
Welcome all to a bit of a thought experiment. While diving down some rabbit holes during COVID lockdown I had a probably crazy idea that I have only shared with one dear friend. It involves a Lynx DCOE manifold, a Megajolt ECU and a Jenvey Heritage DCOE throttle body. You'll have to hear me out on this.

First of all, on reading up on Fuel Injection, I was surprised that even with ITB's there is really no timing to the squirt - it seems that in all but rare circumstances - all injectors are fired as bank and all that varies is the metering (duration) of the squirt. I believe this is because - that at a few thousand RPM - the intake cycle is so fast it would be damn near impossible to hit, and would require a really high volume injector anyway, just to keep up. Anyway, let's not get too sidetracked on this!

Enter the Lynx manifold. With its pairing of ports 1&4 and 2&3, it is supposed to give a less lumpy signal on the Venturi's - which gets a bit moot at high revs anyway. Now let's pair this with a Jenvey Heritage and now we have a solution with fuel rail, injectors and throttle position sensor all packaged in to the one unit. All OK so far.

But how about running this with an ECU that has the option to fire 2 banks of injectors - independently. Could we set such a system up to fire each of the 2 injectors in the Jenvey in alternating cycles? Does the Lynx twinning of alternate cylinders give us enough space to effectively "time" the squirt for each cylinder?

I wouldn't have brought this idea to light - except that there is a single Jenvey Heritage on FB Marketplace for what seems a reasonable price. Now if only I had a Lynx manifold, and a project that might benefit from such a set up....
 
Fireblade said:
First of all, on reading up on Fuel Injection, I was surprised that even with ITB's there is really no timing to the squirt - it seems that in all but rare circumstances - all injectors are fired as bank and all that varies is the metering (duration) of the squirt. I believe this is because - that at a few thousand RPM - the intake cycle is so fast it would be damn near impossible to hit, and would require a really high volume injector anyway, just to keep up. Anyway, let's not get too sidetracked on this!
Click to expand...

Hi Fireblade
I hope the following helps refine your lockdown thought experiment

Injection time, 1 revolution at 6000 rpm takes 10 milliseconds so on a 4 stroke you have 20 msec to deliver the fuel required for max horsepower. Pretty much any injector from a mid 90's car can do this for your regie sierra engine. If what you have on hand is borderline you can increase the fuel rail pressure to squirt more in a shorter time.
Bank injection (and waste spark coils) is used when you don't have a 720 degree dizzy or cam sensor to identify the compression TDC that is required for individual cylinder injection and ignition control.

Many modern aftermarket ECU have tables for start of injection, the smarter way is to specify the end of a sequential injection squirt to before inlet valve closes. Do a search on msextra forum for 'megasquirt injection timing'


Fireblade said:
But how about running this with an ECU that has the option to fire 2 banks of injectors - independently. Could we set such a system up to fire each of the 2 injectors in the Jenvey in alternating cycles? Does the Lynx twinning of alternate cylinders give us enough space to effectively "time" the squirt for each cylinder?
I wouldn't have brought this idea to light - except that there is a single Jenvey Heritage on FB Marketplace for what seems a reasonable price. Now if only I had a Lynx manifold, and a project that might benefit from such a set up....
Click to expand...

Megajolt ECU does ignition control only, a Megasquirt should easily do your 'thought experiment' so get the Jenvey and start your project :jig:
 
Yes, Megasquirt!

AlexB said:
Injection time, 1 revolution at 6000 rpm takes 10 milliseconds so on a 4 stroke you have 20 msec to deliver the fuel required for max horsepower.
Click to expand...
But if each injector is feeding two cylinders, we're back to 10 msec to deliver the fuel required, with each intake valve open for around 5 msec.
 
Fireblade said:
Yes, Megasquirt!


But if each injector is feeding two cylinders, we're back to 10 msec to deliver the fuel required, with each intake valve open for around 5 msec.
Click to expand...
If an EA falcon can feed six cylinders from one (maybe two….?) injectors in the “centre point” injector throttle body, I reckon you’ll be fine ;)
 
Came here to say this too.

Early Mitsubishis also had a big, or maybe 2 injectors that just fired into a plenum. You are in the right track, it will work :D
 
Fireblade said:
But if each injector is feeding two cylinders, we're back to 10 msec to deliver the fuel required, with each intake valve open for around 5 msec.
Click to expand...

It all works like a carbie, there is no requirement for the intake valve to be open, instead of the carby venturi atomising the fuel and plenty of it being left on the manifold walls, spraying the fuel at the back of the valve when it's closed, along with all the turbulence of pulsating air in the manifold keeps the droplets 'in suspension' and gets sucked in when valve opens.
There's 30 years of OEM vehicles doing this, some R19 had a single injector, Cordia and Starion turbo had a single injector.....
The benefit of sequential injectors and timing the squirt to end when valve closes reduces emissions when engine is running at lambda at low rpm where there is potential for heat soak to clump the droplets together increasing unburnt hydrocarbons, once above say 3000 rpm there's very little benefit for timing the squirt as the fuel mixture is richer and the fuel droplets cool the valve and top of the piston (reducing knock)
Two 330cc per minute injectors is likely to be sufficient to power a 100hp engine.
 
Fireblade said:
But how about running this with an ECU that has the option to fire 2 banks of injectors - independently. Could we set such a system up to fire each of the 2 injectors in the Jenvey in alternating cycles? Does the Lynx twinning of alternate cylinders give us enough space to effectively "time" the squirt for each cylinder?
Click to expand...

Fireblade said:
But if each injector is feeding two cylinders, we're back to 10 msec to deliver the fuel required, with each intake valve open for around 5 msec.
Click to expand...

Fit a normal 4 cylinder crank trigger and configure the ECU as a 2 cylinder two stroke and you should then get the alternating injection events with upto 10msec of injector duration for each cylinder at 6000rpm. A tiny gotcha is to source more modern 16 ohm (high impedance) 330cc/min injectors with quicker opening/closing times than early 90's low impedance (approx 2 ohm) injectors.
 
Actually a lower impedance injector (4ohm) will open faster as it draws more current making a stronger magnetic field to open the injector against a stronger spring allowing it to also close faster than a more common 16ohm injector.
Many programmable ECUs have a sequential injection available, and as said it could be configured as a two cylinder two stroke with the set up proposed. Older Motec and Haltech and other ECUs can be bought for $500.

I would suggest any impedance of injector would work fine, sizing being more important for your use. I admit to using low impedance injectors on my road/race car mostly to use 5 bar of fuel pressure to achieve better atomisation.
 
Fireblade said:
Welcome all to a bit of a thought experiment. While diving down some rabbit holes during COVID lockdown I had a probably crazy idea that I have only shared with one dear friend. It involves a Lynx DCOE manifold, a Megajolt ECU and a Jenvey Heritage DCOE throttle body. You'll have to hear me out on this.

First of all, on reading up on Fuel Injection, I was surprised that even with ITB's there is really no timing to the squirt - it seems that in all but rare circumstances - all injectors are fired as bank and all that varies is the metering (duration) of the squirt. I believe this is because - that at a few thousand RPM - the intake cycle is so fast it would be damn near impossible to hit, and would require a really high volume injector anyway, just to keep up. Anyway, let's not get too sidetracked on this!

Enter the Lynx manifold. With its pairing of ports 1&4 and 2&3, it is supposed to give a less lumpy signal on the Venturi's - which gets a bit moot at high revs anyway. Now let's pair this with a Jenvey Heritage and now we have a solution with fuel rail, injectors and throttle position sensor all packaged in to the one unit. All OK so far.

But how about running this with an ECU that has the option to fire 2 banks of injectors - independently. Could we set such a system up to fire each of the 2 injectors in the Jenvey in alternating cycles? Does the Lynx twinning of alternate cylinders give us enough space to effectively "time" the squirt for each cylinder?

I wouldn't have brought this idea to light - except that there is a single Jenvey Heritage on FB Marketplace for what seems a reasonable price. Now if only I had a Lynx manifold, and a project that might benefit from such a set up....
Click to expand...
Very funny, thank you!
 
Thanks everyone for the knowledgable commentary. My take away from the discussion is that using the Jenvey/Lynx manifold does make some sense in that we can set up virtual sequential injection, where each squirt is the benefit of an individual cylinder - but requiring half the injection hardware!
 
Aln said:
Actually a lower impedance injector (4ohm) will open faster as it draws more current making a stronger magnetic field to open the injector against a stronger spring allowing it to also close faster than a more common 16ohm injector.
Many programmable ECUs have a sequential injection available, and as said it could be configured as a two cylinder two stroke with the set up proposed. Older Motec and Haltech and other ECUs can be bought for $500.

I would suggest any impedance of injector would work fine, sizing being more important for your use. I admit to using low impedance injectors on my road/race car mostly to use 5 bar of fuel pressure to achieve better atomisation.
Click to expand...
Yes, low impedence open/close quicker (why nearly all single injector systems are low impedence) but the ones from the 90's will likely spray like a fire hose or you peeing up a wall, hence the trick of higher fuel pressure so the pee stream splashes back when it hits manifold walls. What @Fireblade needs is an injector to mimic the DCOE venturi with a wide conical spray to get the most atomisation of the mixture as it begins it's journey down the Lynx manifold that is able to open/close twice in under 20msec with the least dead time. Injector solenoid design has evolved that you can now get 16 ohm 1000cc/min injectors with wide spray cones and minimal dead time even at idle.
 
  • Like
Reactions: COL
Slightly related - does anyone have a photo of how a fuel line is connected to twin heritage Jenveys?
 
Found out all you need are the T piece, elbow and retaining bolts as fitted to 45mm Twin Webers that will do the job perfectly.
 
You must log in or register to reply here.
Top