Multispark

[Clinton L. Corbin : Backgrind/Gold : Pager 0544]

>(assuming by this you meant waste spark). In addition, half the spark
>plugs on wastespark systems erode the ground electrode faster than the
>center electrode, so if you want 100,000 mile tune up intervals, you
>need double platinum plugs OR customers/dealers/mechanics who are
>willing to take the care to install the two kinds of single platinum
>plugs in the correct cylinders. Guess how often we count on them to
>do it right! Thus, we have to install the fancy plugs in each cylinder
>to idiot proof the tuneup.

>and burning change. Surely ion sensing is easier and more robust across
>a single gap compared to two in series.

Ed,

Ok, I can see the advantages of one coil per plug.  Of course, most of this
started out of a discusion about DIY EFI, so the 100,000 mile tune up
is kind of irrelevant.  But then, you are not looking at this from a diyers
point of view.

What problems would there be in having one TPU channel fire two coils?  It
would still be a type of wasted spark (so you save 4 TPU channels), but you
would have the same polarity on all of the plugs.  And, you would only have
to do the ion sensing across one gap.

If you really can't stand to have the plug fire on the wasted spark, could
you drive the ignition driver with a latch?  The cylinders would be paired
in and "odds or evens" type system (just the way they are in the wasted
spark system) and a normal output port from the micro would select which 
cylinder out of the pairs would fire.  The TPU channel for the cylinder pair
would feed both cylinders.  The cylinder select line would go straight to the
even cylinders and be inverted to feed the odd cylinders.  The TPU and the 
select line would be ANDed together to control the latch.  Possible?  What
problems do you see?

But anyway, thanks for the input Ed.

Clint Corbin
ccorbin at intel7.intel.com