Simple EFI ?

I'm considering trying to make a simple EFI
set up for a "racing only" application on a 350cc
2 stroke . The application doesn't require idle
capability or a whole lot of "drivability " . Roughly
90% of the operational time will be WOT , the
other 10% not less than 1/4 throttle .
I "think" I can accomplish this with a very
simple computer linked with a TPS and I "don't think"
that timing of the injector will be an issue due to the
rpm involved ( 9500-14,000 )
I normally run carbs on this stuff but have been
forced to use some creative manifolding brought on
design parameters . This has brought to light a
problem with pooling ( fuel falling out of suspension )
and blow back .
I figure I'll need a pressure pump , pressure
regulator , injectors , butterfly body with a TPS
( that's not on/off ) and a little help from my geeky
friends for the computer end .
Anyone have some insightful input to add or opinions
about this idea ?
Will

wrote:
>
> I'm considering trying to make a simple EFI
> set up for a "racing only" application on a 350cc
> 2 stroke . The application doesn't require idle
> capability or a whole lot of "drivability " . Roughly
> 90% of the operational time will be WOT , the
> other 10% not less than 1/4 throttle .
> I "think" I can accomplish this with a very
> simple computer linked with a TPS and I "don't think"
> that timing of the injector will be an issue due to the
> rpm involved ( 9500-14,000 )
> I normally run carbs on this stuff but have been
> forced to use some creative manifolding brought on
> design parameters . This has brought to light a
> problem with pooling ( fuel falling out of suspension )
> and blow back .
> I figure I'll need a pressure pump , pressure
> regulator , injectors , butterfly body with a TPS
> ( that's not on/off ) and a little help from my geeky
> friends for the computer end .
> Anyone have some insightful input to add or opinions
> about this idea ?
> Will


Check out http://www.diy-efi.com/contents/complete05.php
as a start, then look into the Megasquirt injection computer
kit at http://www.megasquirt.info/ms2/

If it was as simple as just turning on the injectors you
could do with a square wave generator and an amplifier
to run the injector coils, but it's much more complicated
than that. For one, you need to vary the injector pulsewidth
based not only on TPS, but manifold pressure (or vacuum) since
both sensors indicate in part how much actual air mass is
entering the engine. At WOT you can have a wide range of
air mass flow, so running a single injector pulsewidth will
result in wildly varying air/fuel (A/F) ratios, a bad thing
for most engines. Another variable is that as manifold vacuum
varies the pressure drop across the injector varies as well,
and that directly affects the fuel flow through the injector,
so you need some way to compensate. Older cars do it with a
fuel pressure regulator connected to manifold vacuum such that
the fuel pressure rises and drops in synch with the manifold
pressure. Newer cars accomplish that function in the ECM by
modifying the injector pulse width.

JazzMan
--
************************************************** ********
Please reply to jsavage"at"airmail.net.
Curse those darned bulk e-mailers!
************************************************** ********
"Rats and roaches live by competition under the laws of
supply and demand. It is the privilege of human beings to
live under the laws of justice and mercy." - Wendell Berry
************************************************** ********

wrote:
> I'm considering trying to make a simple EFI
> set up for a "racing only" application on a 350cc
> 2 stroke . The application doesn't require idle
> capability or a whole lot of "drivability " . Roughly
> 90% of the operational time will be WOT , the
> other 10% not less than 1/4 throttle .
> I "think" I can accomplish this with a very
> simple computer linked with a TPS and I "don't think"
> that timing of the injector will be an issue due to the
> rpm involved ( 9500-14,000 )
> I normally run carbs on this stuff but have been
> forced to use some creative manifolding brought on
> design parameters . This has brought to light a
> problem with pooling ( fuel falling out of suspension )
> and blow back .
> I figure I'll need a pressure pump , pressure
> regulator , injectors , butterfly body with a TPS
> ( that's not on/off ) and a little help from my geeky
> friends for the computer end .
> Anyone have some insightful input to add or opinions
> about this idea ?
> Will
>

Why do you want ELECTRONIC fuel injection? Guys around here use a
manually adjustable system, works fine. No need for fancy electronics.
Look at the old Hilborne system. Most of the local guys that build
their own FI use a copycat system much like the Hilborne.

I agree with Don about machanical FI being the way to go but I think
you all forgot that he said the engine is two stroke! I think this
would be HUGE engineering problem, (intake port position, exhuast port
location, etc.) The injectors would have to mounted to the head and
crank position factored in somehow. How is this done on a two stroke
diesel? OK! now your brains are racing!! Man I Love this stuff!!

Rick

thetoolman wrote:
> I agree with Don about machanical FI being the way to go but I think
> you all forgot that he said the engine is two stroke! I think this
> would be HUGE engineering problem, (intake port position, exhuast port
> location, etc.) The injectors would have to mounted to the head and
> crank position factored in somehow. How is this done on a two stroke
> diesel? OK! now your brains are racing!! Man I Love this stuff!!


Two cycle diesels have camshafts and exhaust valves and injectors
operated by a cam lobe. At least the Detroit Diesel does. There are
valveless two cycle diesels such as the Stuart Turner, Cerlist and the
rare 51 Series Detroits but they are museum pieces now. There are
probably a couple of Fodens still in England and maybe in the odd
English doubledecker in America (most had Gardners and have ben
refitted with Cummins or Detroits here) and I have no idea what they
use.

Anyway, FI is a waste of time in this application, crude metering
devices like the Posa-Fuel or Ellison spray carbs would work as well.
But the Hilborn is a crude system indeed-the RSA Bendix is far better.

Perhaps if I shed a little more light on what I'm up to .
I build and race PRO 350cch outboard boats . I'm allowed to do
ANYTHING to these EXCEPT run fuel that is not liquid at room temp and
supercharge . My current set ups use carbs on long manifolds . This
means the motor is mounted high on the tower ( high CG = "interesting "
handling issues )
The idea behind this lowering the CG with out having to go to longer
runners where the charge will really fall out of suspension. The charge
is initiated right at the port and
should get well mixed by the time it's blown through the transfer
ports. Means I can place the butterfly TPS pretty much where ever I
like ( dry would be good ) with out having any pooling issues . At
12,000 rpm I think the injector is going to be spewing pretty much all
the time and having it "timed " to the port opening would not really be
a necessity .
Mechanichal FI would require driving the pump with all the
attendant bits and pieces . This a "racing only " application on a 2
cylinder 350cc motor that will spend most of it's run time at WOT .
"Drivability" is not too much of an issue .
I want to keep this as simple as possible .
I appreciate any input being as this is not a car . As far as
mounting the injectors , fuel rail and so forth , that should be "a
piece of cake " considering some of the stuff I've pulled off in the
past .
Thanks , Will

Well... That could give me a head ache .
During my oridginal R&D I found some really interesting things happened
concerning manifold length vs the choice of tuned pipe .
There were certain rpm ranges wre you would get a "coupling " effect .
Hmmm...
Will

thetoolman wrote:
> I agree with Don about machanical FI being the way to go but I think
> you all forgot that he said the engine is two stroke! I think this
> would be HUGE engineering problem, (intake port position, exhuast port
> location, etc.) The injectors would have to mounted to the head and
> crank position factored in somehow. How is this done on a two stroke
> diesel? OK! now your brains are racing!! Man I Love this stuff!!
>
> Rick
>
Why not use throttle body injection. Most of the homemade FI I have
seen at our tracks were tbi, though one is homemade port injection.

One big thing, though, that I forgot to mention. All these homemade FIs
were on cars running methanol. The old fashioned mechanical fuel
injection systems ran VERY rich at part throttle. This is acceptable
with alcohol- engines will burn a VERY rich mixture of alcohol. When
these systems are tried on gasoline they don't work well at all. Lots
of folks tried Hilborne injectors on small offies for road racing. Was
not that successful- in fact, it was very hard to get Offies to run on
gasoline, carb OR injection. So I would only try homemade FI on alcohol.

wrote:
> Perhaps if I shed a little more light on what I'm up to .
> I build and race PRO 350cch outboard boats . I'm allowed to do
> ANYTHING to these EXCEPT run fuel that is not liquid at room temp and
> supercharge . My current set ups use carbs on long manifolds . This
> means the motor is mounted high on the tower ( high CG = "interesting "
> handling issues )
> The idea behind this lowering the CG with out having to go to longer
> runners where the charge will really fall out of suspension. The charge
> is initiated right at the port and
> should get well mixed by the time it's blown through the transfer
> ports. Means I can place the butterfly TPS pretty much where ever I
> like ( dry would be good ) with out having any pooling issues . At
> 12,000 rpm I think the injector is going to be spewing pretty much all
> the time and having it "timed " to the port opening would not really be
> a necessity .
> Mechanichal FI would require driving the pump with all the
> attendant bits and pieces . This a "racing only " application on a 2
> cylinder 350cc motor that will spend most of it's run time at WOT .
> "Drivability" is not too much of an issue .
> I want to keep this as simple as possible .
> I appreciate any input being as this is not a car . As far as
> mounting the injectors , fuel rail and so forth , that should be "a
> piece of cake " considering some of the stuff I've pulled off in the
> past .
> Thanks , Will
>
Well, alcohol is liquid at room temperature- more so than gasoline, so I
would say go with mechanical injection and methanol.

Thanks to all for your thoughts . Lot's of "food for thought".
I'll have to digest for a bit . Would like to stay away from alcohol
due to the , roughly , twice as much weight issue .
It "seemed" like a pretty simple at the time but I can see there
are some issues that need to be "re-thunk" ( so to speak) . Thanks
again , Will