Well here�s a little info on my EFI / Solid State Ignition System project
for all those people who e-mailed me regarding it. Sorry it took so long to
write but I�ve been up to my neck with assignments and exams.
Firstly let me say that I am not an expert in electronics. I taught myself
everything I know... so that being said here goes...
The system was inspired a few years ago after perching a Dick Smith
electronic kit. The kit was basically a parallel port interfacing device
which enabled a PC to set voltages and also read them. After playing with
this system for some months and reading more and more articles in magazines
about EFI and electronic ignition a thought downed upon me. �Why don�t I
make my own using this type of hardware (Dick Smiths)...� Well this was all
very well until I realised what I was getting myself into�
Firstly I attempted to create a device which ran directly off a PC. This
had many benefits. It was cheap, easy to program, even more easy to modify
fuel / ignition maps and I was already familiar with them. But there were
several major disadvantages. Firstly � Who the hell would drive a car
around that was powered by a PC ??? (yer I�m not gave to answer that�) and
secondly it would mean that every time the car was stoped you�d have to shut
down the PC and then upon starting the car you�d have to re-boot it. Not
even a remote possibility.
>From there I basically discovered the world of microcontrollers; and after
doing heaps of research on the different types available, I decided to use
the Atmel AT89S8252 24MHz chip.
Upon designing a system powered by this chip I had to basically decide on
how my system would work...
The system essentially works in this way:
Because at a given RPM and known throttle position the air entering the
engine will always be the same (think about that it�s VERY important) load
sites can be devoted to these positions. Ie the system can say: �The
engine is doing 4500 RPM and the throttle is in position 10 (flat out).
Therefore the pulse width will be �x� milliseconds�. This removes the need
for an air flow meter. In conjunction to these, base values other
additional fuel quantities can be added. For instance, in the previous
example the throttle was in the flat out position and, depending on how fast
the throttle moved the micro could add perhaps 1.5 times the fuel and decay
to the base value over 3 seconds. This is basically the traditional fuel
pump used in most carbies.
Another example may be if the engine was being started from cold. The micro
could perhaps double the fuel and decay in a linear fashion depending on
engine temperature.
There are many factors taken into account in finding the injector pulse
width such as engine temp, inlet air temp, throttle position, throttle
position acceleration, RPM � and a few others; but this basically sums up
how the system works.
The ignition is probably going to be run from a separate micro � just coz
it�s simpler - but is basically works on the same principle. Read all the
inputs and work out when to fire the spark as apposed to read all the inputs
and work out how long to hold an injector on.
As for the circuitry to do all this it�s not all that hard. The heart of
the entire unit is the Atmel micro, it basically runs the show. However, to
do this several other chips need to be used:
An analog to digital converter � this basically allows the micro to work out
things such as all the temperatures and the throttle position by converting
a voltage into a digital signal.
Transistors and opto-isolators � These are what�s used to drive either an
injector or coil. Basically the micro sets a pin high, which through a bit
of circuitry, (which is optically isolated from the micro) switches power to
the injectors or coils.
A few 7805�s which supply 5 volt and 10 volt regulated power to the system.
A heap of capacitors and resistors to make the thing work�
Interrupts sent to the micro: This is where I�m currently at, what is the
best way to work out the RPM and engine position ?? Welding a disk onto my
front pulley doesn�t sound to crash hot as it changes the weight
distributing of harmonic balancer which I hand on the engine when it was
being balanced (L20B by the way). Drilling holes through the pulley does
the same thing... So I�m probably going to weld up a distributor and use it
as the source for working out the engine RPM and position. Until I build
another engine and modify the pulley before I get it balanced.
Advantages in doing all this:
Well I hope to get more power basically !! Yer, I know I could just drop in
some quadruple turbo charged, supercharged, nitrous injected, water injected
454 bored out 3 feet with twenty values per cylinder BUT building a CHEAP
system which I fully understand has far more advantages than just an engine
swap� I hope
Also I can add quite a few nifty things to my system � such as a switch on
the dash �Power / economy / emissions test�. So the cops will never suspect
my little datto of anything environmentaly damaging� Another feature could
be a hard and soft rev limiter. Another could be traction control � simply
add a sensor to each wheel and retard the timing if the rear wheels are
turning faster than the front wheels. (I�d have a switch to disable this of
corse...). There are heaps of little things that can be added to the system
at a later date.
That really sums up the current system. I haven�t currently got a circuit
diagram drawn on the computer but I�ll work on that if anyone�s
interested...
Hope this helps a bit. It's prolly not technical enough but I hope it gave
you an overview of the system. If anybody would like more info just e-mail
me at:
[EMAIL PROTECTED]
Jeff.
Update � (26/06/00) - After doing some research at uni I discovered back in
1984 a student developed a �Solid Sate Ignition system� which was used on a
L20B engine donated by Nissan ! Reading through this I discovered a better
way to work out then engine position using a light weight �chopper� disk
mounted on the front of the engine. Also, after frying a couple of micro�s
whilst doing experiments with me car I�ve redesigned the hardware and
started drawing them up on my computer. These coming holidays might be
interesting...
Jeff
________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com
--membersozdat-------------------------------------------------------
OZDAT Mailing List Please Note:-
Send (un)subscribe requests to [EMAIL PROTECTED]
Send submissions to [EMAIL PROTECTED]
No unauthorised redistribution of this email
http://www.datascribe.com.au/ozdatonline/index.htm
http://www.datascribe.com.au/ozdatonline/listindex.html
http://www.mail-archive.com/[email protected]/
---------------------------------------------------------------------
