CD Ignition

[tom cloud]

>
>Oh contrare.  Back when I lived in Edmonton (-40C temps.) I installed a CD
>ignition into my Datsun 510.  

 [ snip ]

>So if I know when to fire the plugs I can program the micro to appy current
>through the coil for n milliseconds to keep dwell constant and coil heating
>down.  So lets say 5ms dwell at 12 volts and a coil resistance of 12 Ohms.
>Then we have 1 Amp for 5 ms and when current is broken, fild collapses and
>ignition occurs.  At cold temperatures and a battery voltage of 6 volts
>during cranking we get only 500ma and much less energy.  So.... do I step up
>the voltage to 24 volts with a switching power supply and decrease dwell by
>50% and then still trigger the coil with the FET transistor.  
>
>In other words, before uComputers was the CD Ignition with the Capacitor and
>the SCR just an elegant solution to the problem?
>
>John.
>
>

No.  A coil is used (could use Cockroff-Walton, or other voltage multiplier) to
amplify voltage to the 10 - 20 kV required to jump plug gap at high pressures.
The coil is actually an autotransformer -- and only works on AC, not DC.  Well,
a 'change' in voltage / current looks like AC to it, so applying and removing
current causes a pulse on the secondary.

Now, there's been lots of clever ways to do this.  Points is just a switch.
Transistor and FET (still a transistor) systems are still just switches.  With
this system, the coil field is expanded by passing current through it.  Once the
field has reached its fullest, additional "dwell" time is of no further value.
It's the collapsing field that generates the spark (i.e. opening the switch /
removing the current source).  Problem is, at high RPM's there's not enough time
to build the field up -- hence less spark energy at high RPM's.  To get extra
energy for starting, the coil is made to operate at less than normal battery
voltage, hence the need for the series 'running' resistor.

CD systems store the spark energy in a capacitor.  With a short time constant
circuit, this can be done mucho quickly.  Then, the energy is discharged
into the coil,
producing the spark.  This removes the RPM limitation, for all practical
purposes.

Either system can cause excessive heating in the coil if not properly designed.

Any pulsed coil will ring.  Some CD systems clamp the ringing.  MSD has made
their
trademark (multi-spark) out of a naturally occuring phenomena.  The rate of
ringing (pps) can be changed by the coil design or external components and
it can be clamped after a pre-determined period of time if so desired.

Methinks my Ford EEC-II is a transistor switch system.  Apparently most of the
aftermarket systems are some form of CD.  Don't know of any better way --
just better implementations of it.

I'm interested in hearing of any newer technology.

Tom Cloud