Ignition question

[John Dammeyer]

At 02:54 PM 06/12/1997 -0800, you wrote:
>This may sound really dumb, but I gotta ask.  In a CD ignition system 
>they "dump" a 1 MFD. cap (at 200-300v) thru the primary side of the coil.
>The reason I was told DC is to begin the arc at the plug on a specific
>electrode.  In a distributorless ignition it looks like it doesn't 
>matter.  So why not just pulse the CD charge transformer thru the ign.
>coil for say a little over a millisecond.  If the answer causes smoke
>just e-mail me, and I'll put my hand down.
>  TIA Bruce  nacelp at bright.net

My two cents worth.

Conventional ignition systems with mechanical points work by closing the
points and moving current through the coil which creates a 'magnetic field'.
This is the dwell period.  When the points close to start this current  the
inductive nature of the coil prevents current from instantly flowing;  it
takes some time, unlike a resistive load in which current flow begins
effectively instantly.  After a period of time maximum current flow is
reached.  The speed that the current increased from 0 Amps to maximum amps
can be considered an AC component and is therefore transformed in the
secondary up to 1200V given a 100:1 turns ratio.  This is not a high enough
voltage to ionize the gases around the spark plug electrode and therefore
start current flowing;  to create a spark.

When the points open at some point before top dead center the current flow
is interrupted and the magnetic field collapses.  In the same manner that a
generator creates electrical current by moving a coil of wire through a
magnetic field a collapsing magnetic field generates a voltage in the
windings.  This voltage is approximately 200 volts in the primary and is
stepped up to 20Kv in the secondary which _is_ enough to create the spark.

Trouble is,  in really cold weather and under cranking conditions the
battery voltage is rarely 12.6 volts but instead can drop down to about 8
volts.  The result is a smaller magnetic field and a lower intensity spark
or perhaps no spark at all.  Under higher compression engines the gases are
also less likely to ionize and a spark is harder to create.

The initial solution when CD ignitions were first developed was to create a
system where high voltage was immediately applied to the coil to create the
spark.  However,  transistor technology was not at the point where reliable
transistors were available inexpensively.

Instead a high frequency oscillator drove a high power pair of transistors
to switch 12 volts into a transformer as a voltage inverter and create a 300
volt supply.  The supply charged up a capacitor in a technique similar to
charging the capacitor in an electronic flash.  When the points open a
voltage is applied to the gate of an SCR which then immediately conducts the
charge and the energy stored in the capacitor through the coil primary.
This 300 volt waveform is immediately stepped up to 30,000 volts which
easily ionizes the gases and creates the spark.  The side effect from all
this was that the current through the points dropped down to a few milliamps
resulting in improved point life.  Due to the nature of the inverter
circuitry even an 8 volt battery  still produces a high enough secondary
voltage to charge the capacitor.  At the moment the capacitor is discharged
and current drops to less than the SCR holding current the SCR shuts off and
when the points close again the capacitor can recharge for the next cylinder.

That was then and this is now.  The computer revolution has produced great
advances in switching power supply technology.   We can now use this
technology to produce more efficient invertors and create a higher quality
300V supply and switch this supply with either SCRs or FET transitors.

At 12,000 RPM we have 5ms per revolution and require a spark at about 0.55
milliseconds before TDC.  This means that the ignition system must be
capable of producing a high voltage spark with a duration of at least 0.55ms
to get things going with a bang.  8-)

At 1200 RPM we have less advance but more time and we have a system in place
that can manufacture a high intensity spark 0.55ms long.  Why not do a bunch
of them from 10 degrees BTDC on until some time after TDC.  So produce a
stable high energy 300V supply that is supplied as a square wave into the
primary of the coil.  At high revs perhaps only one spark occurs but at
lower revs you get multiple sparks.

IMHO

John