Date:         Fri, 1 Mar 96 9:48:40 PST
Sender:       Vanagon Mailing List <vanagon@vanagon.com>
From:         Brian M. Miller <miller@spk.hp.com>
Subject:      Re: CDIs

Mike Radtke writes.....


It turns out that time was the overriding issue.  The time constant of a 
inductive discharge system (Kettering) is L/R and a CDI R*C.  Thus, in a 
CDI, a lengthy spark requires a lot of R.  And, since the energy of the 
system (other than the not insignificant radiated impulse) is dissipated in 
R, the R better be in the spark, and not the wiring or other components. 
 The solution is a large gap.

****************************************************************************

I have never studied ignition discharge waveforms and I do not dispute Mike's
observations.  However I do question his technical explanation. The discharge
time constants of both a CDI and IDI ignition are strongly influenced by the 
presence of the coil inductance.  I contend that the CDI time constant is not
R*C.  The capacitor of a CDI ignition discharges into the primary of the 
ignition coil.  The secondary of the coil ignites the spark.  The effective
resistance of the spark plug is scaled by the coil turns ratio (squared) and
reflected back to the primary of the coil.  We can then simplisticly model the
CDI circuit as a charged capacitor discharging into a series RLC circuit.  For
such a circuit, the discharge voltage (across the resistance, ie. spark plug)
is a exponentially decaying sinusoid.  The frequency of oscillation is:

1/2*PI*sqrt(1/(L*C) - R^2/(4*L^2))

The exponential decay time constant of the voltage is R/(2*L).
Note that the decay time constant is dependant only on L, not on C. 

Note:  My explanation is technically correct, but I believe that a simple RLC 
model is not adequate to fully explore ignition behavior.

Also, for what it is worth...
I have used CDIs for the last 18 years and have never noticed a performance
degradation.  I have used CDIs on:  ,74 Fiat 128, '67 Datsun 2000 roadster,
International Scout (yr ?), '72 VW type II.  The CDI has always worked well.
I gap the plugs per manufacturer specs.  For the last several cars I used the
Tiger CDI from Universal Corp (Grand Junction, CO).  I have had no problems.  I
should mention that I have always lived in relatively dry climates.  I am about
to put one in my newly acquired '80 Vanagon.  I really like the CDIs because
the points don't wear.  One caution though;  The points still rub on the 
distributor cam and the rubbing block will wear out if the cam is not
appropriately lubricated.  This happened to me once.  I wore the rubbing block
down so far that the points no longer opened!  But the CDI kept the car 
running normally until the gap was immeasurably small.

Brian Miller
'72 Type II, '80 Vanagon  (both campers)