Laptop Management

Phil Lamovie injec at ains.net.au
Sat Aug 14 19:57:27 GMT 1999


Try this for size.

RPM = 5500 or 91.6 RPS or 32,999 degrees/sec.

People live in days and hours engines live in degrees.

In order to digitally plot 33,000 degrees to an accuracy of 1 degree
you will need to have some very flexible timers.

 You are going to have to calculate RPM a minimum of 4 times per
revolution otherwise your ignition timing will fall in a heap as the
engine accelerates. If your rate of change is 2500 rpm/sec that is
from 2500 to 5000 in one second ! Then the rate of change per rev
is on average 7 degrees.

If you wait until after the end of the revolution and then start your
calculation you would be lucky to be within 12 -14 degree of
MBT timing.

The very first thing is to establish the task. What am I controlling and
what are it's requirements. Is it a requirement that the system be of a
certain standard ? Do you need closed loop for catalytic converter
feeding or can it be dispensed with.

Do you need one pressure sensor or two ? I wish someone would slap
me with a solid answer on this as I feel it's still not firmly established
what is going on.

As much as every control engineer loves a closed loop. It is the solution
that works the slowest. The sensor takes  for ever to register a change.
most brand new broadband sensors take 200 - 300 ms that's enough
time to do 25 to 30 revolutions or 120 firing strokes without a clue.

Most carbys can do as well if not better.

Your system, if it is to inspire others, should be of a Linux vs M$ battle
do better than the big companies not worse.

> If I need to be able to control at 1 msec intervals then I would need to
be
> able to output at 1 kHz. I think I will require clever electronics in
order
> to address 8 injectors individually or in pairs (any ideas for this)?
Would
> I need a system with 8 digital outputs at 1 kHz each?

The resolution for injector control is 0.01 ms or 10 micro seconds which
ever comes sooner.

The grouping of injectors is very much a peripheral issue compared to
accuracy of the injection duration. If your engine performs 120 firing
strokes
without  you updating the  output buffer then 119 of them will have the
wrong A/F ratio regardless of your target.

The A/D on your load sensing should be updated a a minimum of once per
combustion so as to give minimum accuracy.


>  If I was to control the ignition timing electronically, it would
probably
>  be using a control to advance and retard the timing using a servo

I can't think why an engineer would suggest such a thing. Maybe your
pentium has a valve with a faulty heater element. Or perhaps the
clock work spring in the computers dynamo power supply has run down ????

Nissan often uses 180 pulses per revolution to measure engine speed.
Toyota 12/rev, Ford 8 /rev, GM 18/rev.

Why would they bother if it could be done using the postal system.

I know that this might sound a tad sarcastic but I'm desperately trying
to bring you a couple of orders of magnitude closer to the problem.

It is some what akin to me thinking that the "K" in my processor is dead
because when I hit it on the keyboard it doesn't come out onto the screen.

We both know that there are many many layers of software and hardware
between the two events but I'm standing in the shop asking the repair man
to
put a new "K" in my chip.

Once again please forgive my tone I'm just an aging fool with a
bee in his bonnet


 Regards

 Phil Lamovie

 injec at ains.net.au

     cogito ergo zoom



More information about the Diy_efi mailing list