#define TECHNO_JARGON "MAXimum" - delete now before you roast your retinas.
The fiddly bit with the Jacobs' system is that they claim to sense the
resistance of the plug gap (that's do-able), and then to adjust the coil
drive so as to insure "optimum combustion, 100% of the time". Adjusting the
coil drive is hard, because most coils basically give it all they've got,
every firing pulse. Reading between the lines, that means their coil could
be designed such that it works more as a pulse (i.e. some finite width of
input signal) transformer, as opposed to an 'impulse' transformer, like
regular coils, where you set up a DC current through the coil and then
suddenly cut it off to generate the flyback current in the secondary.
If they were to set up a high-frequency AC waveform on the primary, then you
end up generating a high-frequency, high voltage arc at the secondary, like
what the MSD does at lower RPM (I think). If you measure the current
required to cause breakdown at the plug, I guess that you can then make some
decisions about what the plug is seeing, eg. lean, rich, flood, whatever.
Then, since the system is designed such that the coil has a lot of spare
capacity, you can fry the plug clean if you detect a flooding condition.
Note that all of this only works at lower RPM, where you have enough time to
fire the coil multiple times for a single plug. At high RPM you're still
restricted to one spark per cylinder per combustion cycle, and if you've got
flooding or lean-miss conditions happening there, the ignition isn't going
to help you anyway.
I think we're basically of the same mind-set here - Jacobs claims way more
benefit from their system than what seems reasonable, if you assume you're
comparing it to a modern ignition system to begin with. The basic
pickup-amplifier-coil-distributor ignition system has become pretty darn
reliable over the years and it's only in extreme cases where you have to
look at something else. Some cases that I know of:
1. belt-driven overhead cam engines where the distributor is driven off the
cam. At high RPM the timing scatter is bad enough to cause significant power
loss compared to a crankshaft-triggered system. Solution: Crank triggered,
distributorless ignition. Jacobs can't help you here.
2. Really zoomy Chevy motors where the camshaft twists enough at high RPM to
retard the timing (Tim, did you guys run into this? I can't remember who I
heard this from). Solutions: (a) A twisted cam that straightens at high RPM,
(b) advance the timing so it's right at high RPM, (c) crank triggered
ignition
And that's it. For just about anyone else, 90% of the battle is installing a
good distributor, with a good cap, and good wires, onto good plugs, and then
making sure that all the non-ignition things are also in order. Once you've
got that you can pretty much use a nail with a bunch of bell wire wrapped
around it as the ignition coil (if anyone tries this and makes it work let
me know).
Happy Holidays!
Theo Smit
tsmit@isotel.com
tsmit@home.com
B382002705
> -----Original Message-----
> From: Doug & Rett Leithauser [SMTP:dleit@mintcity.com]
> Sent: Tuesday, December 14, 1999 7:37 PM
> To: TSMIT@isotel.com
> Cc: Tiger List
> Subject: Tailored spark
>
> I know this is 2 weeks late, but I've got to say it anyway. The coil
> output
> going to a particular cylinder is goverened by the voltage needed to jump
> the largest resistance in the circut, hopefully the spark plug gap. In
> this
> way all ignition systems "tailor" the spark to the requirements of the
> individual cylinder. It's just the way ignition works.
> Doug Leithauser
>
> I haven't read the Jacobs' techno-hype lately but on the surface they make
> some pretty spurious claims, such as being able to sense individual
> cylinder
> "requirements" and to tailor spark output on a cylinder by cylinder basis.
> To be fair, I haven't read their patent (you can get access to it through
> IBM's on-line patent search engine), but it seems to me that in order for
> an
> ignition system to sense cylinder conditions it would have to basically
> monitor the coil flyback pulse (which could be dependent on the
> characteristic of the spark generated by the spark plug, which in turn
> will
> depend on the cylinder load etc.), and then make some quick decisions on
> what to do for the current firing cycle, or the next. Maybe the problem is
> more simple than I think it would be, but it seems to me that the
> decisions
> that come out of this process would be kind of arbitrary, unless you throw
> a
> lot of hardware and software at it.
>
>
|