Ignition questions

[Darrell Norquay]

Doug Rorem wrote:
 
> Well this art isn't much better, but here's where you want to place
> your fast recovery diode (switches on from a reverse bias condition
> very quickly).. This provides a path for the current to keep flowing
> in the coil after the transistor turns off [the coil wants to keep current
> constant]. Without something like this, the coil can create very large
> voltages and cause the transistor to break down.
 
 
First, to reply to Doug's comments.  I tried this years ago with an old 
points style ignition.  I figured if I hung a fast recovery diode across
the coil, as you've indicated, it would allow the coil's magnetic field
to collapse faster, and induce a higher secondary voltage.  As a side 
benefit, it should have absorbed most of the voltage spike from the primary 
and made the points last a lot longer as they wouldn't arc over as much.

WRONG!  It didn't work.  As soon as I connected the diode across the coil, 
the engine stopped dead.  As near as I could figure, what was happening 
was that all the energy stored in the coil was being dumped into the diode
instead of being allowed to excite the secondary.  With the diode, you're 
limiting the negative voltage developed across the primary to under a volt 
(the diode forward voltage), so even with a turns ratio of 100:1, that 
still only allows a secondary voltage of about 100V - definitely not enough 
to tickle a plug.

So, the only way to get the coil to work properly is to allow the primary
back EMF to increase at will.  Clamping it to a small value will simply
kill the output of the coil.  The easiest way to protect the coil driver 
transistor is to simply turn it back on if the coil back EMF increases 
beyond a certain value.  This can be accomplished by putting a couple of
150V or 200V zeners in series from the transistor collector back to the base, 
with a suitable current limiting resistor in series with the base to avoid
pumping too much current into it.  (this will destroy the transistor as
surely as exceeding it's VCEO rating)  This will allow the primary back
EMF to go up to 300 or 400 V, then the zeners will begin to conduct and
turn the transistor back on, thus limiting the back EMF to the value of 
the zeners.  If you use a 500 or 600V 15A transistor, it should live quite
nicely for a long time.  

But, you say, won't turning the transistor back on start charging the coil
again and kill the spark?  Yes, but the primary will take some time to
build up to this voltage, and hopefully your spark has already triggered
by the time the transistor turns on again.  Also, the energy stored in the 
primary is relatively low, and it will only supply the base current of
the transistor for a limited amount of time, as soon as the voltage on the
primary drops below the 300 or 400 V threshold, it will turn off again.
By this time, most of the stored energy in the coil will have dissipated
and it will be ready for the next ignition charging sequence.

Secondly, on the subject of ballast resistors, someone mentioned using
a transistor driver with the existing coil and ballast resistor.  This
amounts to simply replacing mechanical points with a transistor, and in 
fact a lot of the earlier "electronic" ignition systems did this. (most
noteably the Dodge little silver box with a wierd 5 pin plug)  While this
is certainly an improvement over points, it has little effect on the energy
supplied by the ignition system.  The ballast resistor is used as a current
limit for the coil, and it was usually a big honking ceramic body resistor
with a rating of a couple of hundred watts.  The disadvantage of the ballast
resistor approach is that, while it limits the maximum coil current, which
protects the coil from self destructing, it also limits the charging current
to the coil and increases the necessary dwell time significantly.  This 
also limits the maximum energy stored in the coil, and ignition energy drops
dramatically at higher RPM's because there is simply not enough time for the
coil to charge to full capacity.

With the later electronic ignitions, such as GM's HEI, there is no ballast
resistor, and the coil has full unlimited current as soon as the module turns 
on.  The coil quickly charges up to full capacity, and THEN an electronic 
current limit in the module kicks in to limit current to a safe value for the 
coil.  This has the effect of reducing necessary dwell time, and giving much 
better ignition energy and performance especially at higher RPM's.  Most of
these use Motorola's MC3334 chip (or equivalent), and it's still the cheapest 
and best solution for a coil driver.  It has all the necessary smarts built 
in to control a coil, including automatic dwell adjustment, current limit, 
mag pickup preamp, etc, an it's cheep!  The coil driver transistor itself 
will be the most expensive part in the system.  The newer modules use a 
different approach tailored to computer control, the Allegro Microsystems 
ULQ2460 is one such device with similar circuitry to the 3334, but with 
additional pins for computer control of dwell and timing.  

Check out the DIY_EFI web page for schematics of one system using the MC3334.
This circuit is pretty much right out of the Motorola App Note.  Note that the 
schematic has an error, however, the current sense resistor is missing.  
(my apologies to the author if it has been corrected).

regards,
dn

--

-----------------------------------------------------------------
 Darrell A. Norquay              Internet: dn at dlogtech.cuc.ab.ca     
 Datalog Technology Inc.         Voice: (403) 243-2220
 Calgary, Alberta, Canada        Fax:   (403) 243-2872            
 
       "Absolutum Obsoletum" - If it works, it's obsolete    
-----------------------------------------------------------------