tigers
[Top] [All Lists]

RE: Electronic ignition

To: Theo Smit <TSmit@novatel.ca>, Steve Laifman <laifman@flash.net>,
Subject: RE: Electronic ignition
From: Bob Palmer <rpalmer@ames.ucsd.edu>
Date: Thu, 12 Nov 1998 12:53:50 -0800
Theo, et Listers,

I should have guessed my post would open up Pandoras technology box. Well,
just for you I'll waive my usual consulting fee and offer a few comments
without getting too deep into a subject about which entire books have been
dedicated.

First, regarding the Racetech guys and their experiment, I'd have to
actually see their setup to be able to comment other than generally. I do
agree that a "stock" ignition is marginal and quits working entirely at
around 5000 rpm or so, depending on throttle position (i.e., cylinder
pressure). I don't know their specific parameters for pressure, spark plug
gap, etc., but qualitatively what they say seems right. I'm not sure what
"barely fire" means; seems to me sort of like saying a match barely lights.
And you don't suppose these Racetech guys could be just a little bit less
than objective, seeing as how I presume they are trying to sell us
something. Anyway, I'm a great believer in experiments and I'm sure they
learned some important things with theirs.

Now to your questions Theo. Again, this is far too complicated a subject to
do justice to here, but let's hit just a few of the high points. Every gas
is different, but all of them exhibit a characteristic "Paschen Curve" of
breakdown voltage versus pressure, with a minimum somewhere around 1/100 of
an atmosphere and increasing above and below this point. Simplistically,
above the Paschen Minimum, breakdown occurs when the energy of free
electrons exceeds the ionization potential of the gas, so this pretty much
goes linearly with gas density as you suggest. Both pressure and
temperature affect gas density. Other factors would come in at much higher
temperatures, but without going to some references, I'm not sure what the
effect would be near combustion chamber temperatures. I think it would tend
to reduce the breakdown voltage even more than just the reduction in
density, but that's just a guess. As to the breakdown voltage of a gas/air
mixture at STP, let's assume your reasonable value of 30kV per centimeter
is correct. Your analysis is reasonable, except for one very significant
error. The calculation of field strength is not simply the applied voltage
divided by the spark gap. Only if you have two infinite planes can the
field can be calculated this way. The shape of the spark plug electrodes
has a very strong effect on the field; probably increasing the field by an
order of magnitude or more.

WRT your final comments about a "really powerful" ignition system, my
comment would be that regardless of the power, the voltage is set by the
spark plug gap. I run about 0.055" gap. Some production cars run 0.080".
With these very large gaps, breakdown becomes a serious problem, hence the
bigger distributor caps, bigger wires, etc.

>From the Land Before Digital Electronics

Bob


At 11:10 AM 11/12/98 -0700, Theo Smit wrote:
>In the development of their SDS crank-trigger ignition system the guys up
here
>at Racetech built a cylinder with an air fitting, a spark plug hole, and a
Lexan
>window. They found that at higher pressures, the stock coils would barely
fire
>the spark plug, while switching to any of a number of different 'high-energy'
>coils gave good spark even at high pressure / high RPM. Bob, how does the
>dielectric breakdown potential of air go with pressure and temp? If it is
>linear, and about 30 kV / cm at standard temperature and pressure, then at
100
>psi (call it 7 bar) you will need about 21000 V to fire across a 1 mm
(0.039")
>gap.
>
>If you have one of the really powerful ignition systems, you also have to be
>extra careful about the condition of the inside of the distributor cap,
wires,
>plugs, etc., since that bolt of blue you're trying to dump down a particular
>plug wire can skip over to the next terminal without too much trouble if it
>finds any extra impedance in the intended path.
>
>Theo Smit
>tsmit@novatel.ca
>B382002705
>

Robert L. Palmer
Dept. of AMES, Univ. of Calif., San Diego
rpalmer@ames.ucsd.edu
rpalmer@cts.com

<Prev in Thread] Current Thread [Next in Thread>