|
|
Thread: fuel induction
|
troyjmueller
Revvin Up
Posts: 55
|
posted February 17, 2007 08:45 PM |
|
fuel induction
I am working on a project, and am trying to figure out some fuel-system possibilities.
As I understand, there are a few methods implemented by Mazda, as well as "mass-produced" aftermarket, and, of course, custom-engineered for the die-hard tuner.
My main experience with mazda fuel induction is via the 12A nikki carburetor, found on gen-1 rx-7s. This is a 4 bbl carburetor, with vacuum activated secondaries, dual floats, etc. The primary barrels feed the two central intake manifold runners(Intermediate housing intake ports), while the secondary barrels feed the outer intake manifold runners (End Housing intake ports). One question comes to mind, regarding interchange - does this carburetor bolt on to an early 4 port 13B NA Intake Manifold?
As far as I know, all carbureted engines were a 4 port setup with a 4bbl down-draft carburetor from Mazda.
The Stock Fuel Injection is a different beastie, having different variations, of which three come to mind - the 12A turbo, the 13B RESI, the 13B DEI & Turbo, and 13B REW.
From personal experience, the 13B RESI (GSLSE) has two fuel injectors, located on the intermediate housing of the engine. The 13B DEI, NA, has an additional two injectors located on the secondary intake manifold runners (outer).
I have found that the 12A turbo injected fuel into both ports simultaneously, using the compression turbulence and the rotary's unique geometry to equalize the fuel distribution. I have also found the the RX-7 GSLSE had a factory fuel injection setup refered to as a "semi-direct" injection (Mazda), and branded as Bosch "L-Jetronic" fuel injection. I have been unable to find details about timing, sensors,and other technical details about this fuel system.
Then, of course, there is the Holley & Webber carburetor aftermarket intake manifold kits.
**********************************************************
I have a 13B RESI that I a am needing to install in a vehicle, and I don't have the ECU. My preference is for carburetors, as they don't require a "brain-box" to run. I am reluctant to disasemble the engine at this time to build a 4-port 13B, (and I don't have any spare 13B 4port intake manifolds and/or carburetors). I have been seeking some small 1 bbl keihin carburetors for a custom intake, but have not had much luck (yet!)
So, does anyone know some usefull tech details on the GSLSE Fuel injection system, how to direct-wire the injectors for limp-type operation (without ecu), or what key pieces need to be integrated for a custom bare-bones ecu replacement?
I was thinking, that if someone was to build an electronic "buzzer", with a tunable frequency, and then apply the resulting signal to the injectors, that would result in a continuous "mist", similar to a caburetor venturi, and that a vehicle could be made to run ecu-less, and the idle/running mixture could be tuned by a simple radio-shack dc-oscilator circuit, and a (relatively) low pressure fuel supply. This would not be a feedback system, of course, as there is no provision for o2 signal, air mass/volume/flow/pressure, temperature, throttle possition, etc. Just a means of getting adequate fuel into the engine to be able to warm it up, and drive it a few miles, from the shop to my yard. You can't build a monster rig all at once, unless you have a wallet o' gold.
Any experience, knowledge, or advice would be welcome.
|
|
troyjmueller
Revvin Up
Posts: 55
|
|
posted February 18, 2007 06:27 AM |
|
some references and calculations
here are some interesting sites with some useful information on fuel injection:
http://www.megamanual.com/v22manual/rotary.htm
http://www.answers.com/topic/jetronic
http://en.wikipedia.org/wiki/Mazda_Wankel_engine
and, an update on the above information. I did some calculations, and arrive at the conclusion that the injectors are injecting about 3% idle (~1000, 1/3 of highway cruising rpm) 7% (city average) to 11% (highway cruising) of the time, based on 17 city mpg for 60 minutes @ 30 mph, or 25 mpg for 60 minutes @ 60mph. These calculations are from the GSLSE injector sizes.
1984/1985 RX-7 GSLSE has 2x 720 cc/min injectors,
1986-On NA RX-7 has 4x 430 cc/min injectors.
Do those numbers seem about right?
By these numbers, the "black buzzer box" mentioned above would have to generate a continuous positve/off alternatingsignal that is hot & tunable for ranges between 3% and 11% of the total time interval.
If someone wanted to get a bit more complicated, you could get an input from the tach/coil to step up the "on" percentage automatically, and maybe use the stock tps as a signal source for an "accelerator pump" style pulse. Any ideas on how tO get the system to "step down", ie, go from 3000 rpm back down to idle? Maybe a foot-feed cut circuit or the like?
Just what the world needs - an
"electronic-injection carburetor"
LOL!
|
|
klaus42
Rotorhead
Posts: 1877
|
posted February 18, 2007 06:27 PM |
|
You could...
...adapt a Nikki/early 13B manifold to one another, pretty easily... or do any of your aforementioned ideas... but, why not keep it simple? Plenty of proven options out there that likely won't cost more by the time you're done; among the less usual ones for carburetors, I've always thought that the 'Predator' carb. 'ought' to work well, being a very simple and straightforward design that can flow as much as the engine could want... (?)
Don't injectors already atomise fuel, by their very nature? (At least when they're functioning properly...)
|
|
tom93r1
Redlining
Posts: 458
|
|
posted February 19, 2007 09:51 AM |
|
you could probably adapt an 80's GM throttle body injection setup to the rotary without much expense or effort
____________
'74 lawn-green
4-port 13b
Weber 48DCO
Racing Beat Header
Magnaflow
resonator/amplifier
really loud!
|
|
troyjmueller
Revvin Up
Posts: 55
|
|
posted February 24, 2007 09:29 AM |
|
clarification
Thanks for the responses. I am not sure as to what is meant by "adapting" gm-throttle body injection, but I suspect that this would involve adapting the gm tb to the mazda intake, or fabricating a new intake to match the block, and the tb.
background: I have the complete intake tract, air-box, manifolds, etc. I have the (2) stock injectors (located on intermediate housing, in intake port from the stock 1984 13B.
Does anyone know if someone could (or how to) build a manually tunable (air/fuel mixture, idle speed, etc.) circuit box so that the system would function essentially the same as a carburetor, and be able to do the following: 1.) provide sufficient metered fuel for varying rpms, both increasing and decreasing, with regard to engine speed (controled by air speed via venturi in carb), maintain a stable minimum idle (controled by idle circuit in carburetor), control engine speed (controlled by throttle butterfly valve in carburetor, which in turn varies quantity of air traveling through venturi), and NOT require any complex IC ECM. I mean, if someone had access to some of the following electronic parts: discreet transistors or even *gasp* vacuum tubes, resistors, capacitors, wire, a soldering iron, and other miscelaneous electronical components...
In short, does anyone know, or have some thoughts, on how to make the injectors (and only the injectors) function equivilent to a carburetor (manually tunable air/fuel mixture, idle speed, automatic corolation of fuel quantity with throttle position and fuel metered quantity, with a 99% brain-box lobotomy/transplant (aka, replace einstein with a hamster), and toss out all of the "extra" feedback sensors except the throttle position sensor? This engine has an electronic distributor, so spark control is completely autonomous from the injectors. The engine will work with a carburetor bolt-on (though I am still trying to locate carbs), and I am looking for a *temporary* way to use the stock injectors and throttle body.
A carburetor can and does do all of this with mechanical linkages and cleverly laid-out fluid-dynamic principles, without any sensors, brain-box, knock-sensors, oxygen sensors, cam/crank sensors, throttle position sensors, etc. Replacing a mechanical/fluid dynamic linkage with a *very* simple electric circuit shouldn't be *that* dificult.
|
|
|
|
|
