Ignition coil charge time

[Matthew Lee Franklin]

Dudes,

Several of you have written useful comments in response Bill Lewis's 
questions. 

There is what I thought was an *excellent* reference that has not yet been 
mentioned.  It is "HEI-A New Ignition System Through New Technology", by 
Huntzinger and Rigsby, SAE paper # 750346 (Transactions, Vol 84, 1975).  It 
covers the charge time question as well at their solution.  If you've not 
yet read the paper, it describes some really clever innovations: 
adjustable dwell depending on how near saturation, a 5.5 amp limit in 
the driver so you could have a low resistance for fast saturation times at 
high speeds, but that wouldn't fry at low speeds.

It is a little old, but development hasn't gone much beyond that lately. I 
think that comes from a time when OEM's could afford to be much more open 
in talking about what they were doing.

I agree with Robert that 30 degrees of dwell seems to work well.  I used 
that as a constant in a 3.1 liter GM V-6 lab engine (back in '91) that ran 
between 1000 and 3500 rpm with TTL square wave fired HEI coil drivers and 
the original distributorless coils (DIS).  The TTL common was hooked up to 
the white wire (N terminal) while the TTL signal entered through the green 
wire (P terminal).  The TTL square wave signal defeated the cool, 
automatically adjustable dwell feature, but it gave my very precice control 
over the on time and off time (FIRE!) of the coil current.  The HEI driver 
limits the max current to 5.5 amps while the DIS coils originally required 
10 amps.  I left room to use two HEI drivers in parallel to give near the 
required amps, but I had no misfire problems at the conditions I was 
running, so I left it at one per coil.  This was for a 486-PC-based knock 
testing system that randomized the ignition timing for each cylinder 
individually within limits (say 20 to 40 degrees or so).  The PC would also 
read the knock sensor for a window starting around TDC for each individual 
cylinder to about 40 degrees after TDC for the respective cylinder.  The 
resulting knock intensity vs. spark timing plot for each cylinder helped in 
determing the effect of different fuels on octane requirement change.

I don't think Chevron Chemical is using it now, as the knock sensor signal 
was a little too vague, and not equally affected by all cylinders.  
Calibration based on each cylinder didn't seem to help either.

Sorry to ramble on so much, but it was an exciting system to work on, and 
it brought back good memories.

Later,
Matt