IGNITION PROJECT

[jac at wave.sheridan.wy.us]

Mark,

I have this project up and running with lots of hours on it.
The software works well but others could improve it of course.
The board is about six inched long and three inches tall.  A four 
line by 20 character LCD  sits in the front.  

For a number of reasons, the main board should be moved to a larger 
box and the present format of 6 X 3 be retained for I/O functions in 
dash mounted box.  There is lots of room for participation and 
contribution to the project.  I think that it is important to agree 
on format and direction so that the work can be parceled to the 
participants.

I will post the whole project if people want to work on it and 
improve it.  I am not the least bit interested in distributing a 
bunch of code and drawings so that folks can argue about processors 
or programming technique without contributing significantly to the 
project.

The present board needs a little work, some traces have been cut and 
jumped but about one more iteration would get finish it up.  I can 
produce printer files for the present board. I would like to figure 
out a way to post hand drawings of schematics.  Perhaps scanned 
images.  A board and code exist for a test fixture that lets one 
simulate engine conditions and run the whole thing on the bench.

I use assembled 8052 code for the core code and Basic, compiled and 
assembled with Systronix's compiler, for the '232 and LCD screen I/O 
routines. The box has buttons that can be used to modify the 
inservice action of the device.  One version of the code runs on a 
much smaller controller with no inservice I/O but has the same 
control functions.

The processor is a Phillips 87C550. When I build more of these, the 
logic will mostly move to an FPGA.  The code will move directly to 
CAN capable devices.

There is no reason to build the coil driver circuitry when GM does 
such a beautiful job of packaging the entire works for four and six 
cylinder engines.  Two units will work for eights.  GM MAP and 
variable reluctance magnetic pickups are a great bargain.  All of 
these parts are hardened for an automotive environment and are 
available at every auto parts store. The toughest procurement is for 
wire ends to match the propritory GM configurations.  They are 
available though.

The parts are a bargain, but not necessarily cheap. The same is true 
for my time and probably yours. 

The GM components add up to about $200 new, aftermarket.  I have 
several wide temperature '550 available and some cases.  More are 
available.  This may not be the proper forum for this project and I 
do not have the time to manage another one.  If I figure out how to 
use it I guess I have an FTP site available.  If you are interested 
in extending this project, post your thoughts or e-mail direct. 

-------------------------------------
john carroll
 jac at wave.sheridan.wy.us



---------------Original Message---------------
New question.. might develop into a thread, but I'm hoping to start 
a discussion, and follow the whole design process through to a 
running system.

Objective: Replace the distributor and points with a blob with some 
inteligence.

Desired sensors: Combined engine speed/crank angle, MAP, water temp, 
knock, and possibly air temp.

Desired outputs: 'points' output for as many coils as required. 1,2 
or 4. Cooling fan out put. RS232 monitor port. High speed interface 
for sensor output to possible further extension blocks.


Reasons: My engine is a 4 pot normally aspirated 1.5 liter engine. 
Its distributor is shot, and needs replacing. It eats points, and 
condensers and coils, and no there is no fault, all spitfires do 
this!

Sensing: Speed and position from a machined chain sprocket bolted to 
the front pulley (which in my case is keyed in one position to the 
end of the crank), with 36 or 24 teeth. Pickup is to be the digital 
magnetic pickup from RS components , it costs 16 UK Pounds or about 
35 bucks to anybody over there if I have to send them (if you guys 
cant buy them). MAP to be measure using MPX4250A, about 30 bucks (up 
to 36 psi, so that allows turbos up to about 22 psi boost. Knock: 
haven't even thought about this one yet, all suggestions welcome . 
Water temp, million and one ways of doing this.

Processor: 68 something (but only if I'm doing the programming, if 
you want to join, and do the software, I'll adapt to what you like)
On the fly re mapping, in flash.
All information to be made available to other systems on request, by 
serial coms.
High speed programmable pulse chains indicating TDCs, firing points 
etc for accurate timing by other systems (possible fuel inj?) and 
scope diagnostics.
Remote controll of ignition events by other systems (eg, revlimiter, 
traction control etc)

Extra sensor input from current points, to allow initial map to be 
built from currently working mechanical system.

CUT TO THE CHASE:

Order of development:


Sensors first.

Chose sensors, and design their interfaces, then test.
Design software.
Chose uP
Design system.
Test system and refine.

Up shot.. publish the sensors and interfaces , so that other scratch 
builders can just do a building block design on their own hardware. 
If at the point I go 68xx(x) they want to go PC or Z80 or anything 
else, they can, and the software design can be followed and 
implemented, even if they cant use the code.

Things needed imediatly:

Suitable temp sensor, and a knock sensor, and preferably their 
interface schematics if anybody has a suitable solution to these two 
questions, I would dearly like to hear from them.

Does anybody want to join this idea.. or am I on my own for this 
one?

A good programmer would be a bonus, I can do it, but I would prefer 
to let somebody else do it. I am prepared to add dedicated hardware 
for doing timing, and interrupt control, and stuff like that. 
Anything that makes the programming easier. ;-)

Mark

(Who has decided to stop messing around and get on with it)







----------End of Original Message----------