Ignition contd.

Paul E. Campbell pecampbe at mtu.edu
Sun May 14 23:53:49 GMT 1995

I did about 4 hours of research today and this is the best I found:

1. The spark duration should be as long as the flame delay; the delay between
"ignition" (the spark initiates) to the pressure rise in the cylinder. This
delay is affected by:

	The fuel-air ratio
	Dynamic and thermodynamic state of the mixture around the spark plug
	The ignition system (assuming you don't spark long enough)

More than that is pretty much a waste of energy. At that point, you will
have a good growing flame front going. Before that, the flame front may be
self-sustaining and then go into growing mode, or it may go out. Still can't
find any reasonable references for measuring the flame front.

If you do it right, then you shut the spark off at TDC since the pressure
rise now coincides with TDC, right where it should be.

This point is specified as the minimum advance for best torque (MBT) as
long as you avoid running into either the knock limit or the retard
limit (point where incomplete combustion occurs).

In practice, it appears that the retard limit can probably be ignored but
the advance limit will always be present.

2. There are LOTS of different spark ignition systems. The major different
systems are the capacitive discharge system, inductive ignition, and
gated oscillators.

For inductive discharge, connect the primary of the coil (transformer) to
your battery. It will begin charging up (storing energy). When you are
ready to initiate a spark, disconnect the battery. The primary coil will
try to discharge all that energy into the secondary coil of the transformer,
which is attached to the spark plug. This system has a 25-75 microsecond
rise time and a 1-2 millisecond duration. Higher voltages give more stored
energy which gives you a longer duration..that was the reason for the high
energy spark systems in the early 80's.

Capacitive discharge systems charge up a capacitor, then discharge the
capacitor through the coil's primary. These systems have fast rise times
(3-20microseconds) but only a 5-50 microsecond duration.

Gated oscillator systems turn a high frequency oscillator on and off which
is feeding the coil primary. They have a pretty much infinite and controllable
duration but they are serious power suckers...not very efficient.

There are restriking versions of all of these. And there are a few systems
which use charge pumps to maintain spark sustaining power in the inductive
discharge system for an arbitrary time. SAE 920575 details a hybrid
system which uses both capacitive discharge and an inductor to achieve
fast rise times and an indefinite duration.

3. The spark voltage and current are not really too important as long as you
manage to strike a spark and try to feed as much energy into the flame kernel
plasma as you can (increasing the spark voltage helps).

More information about the Diy_efi mailing list