closed loop dwel

[steve ravet]

DIY_EFI Digest wrote:
> Date: Mon, 13 Dec 1999 03:47:48 PST
> From: "Robert Sandor" <sandorr at hotmail.com>
> Subject: virtual closed loop dwell regulation
> 
> I am working out an idea to fire sparks of a 4cyl. carburetted engine in
> right time without use of lookup tables or a knock detection nor ionization
> current measurement.The only input to the regulator will be the RPM. The
> RPM-dwell lookup table itself is not enough, because the  temperature of the
> air and fuel varies. The knock detection and the ionisation current
> measurement are expensive methods. I hope a cheap method "virtual closed
> loop dwell regulation" will work. The main idea is to advance-retard dwell
> until maximal RPM is measured at fixed throttle position(Is this a fuzzy
> logic). I want to use separate coil for each cylinder. At very low RPM (like
> starting the engine) even multiple sparks could be applied. I also want to
> drive switching transistor without a MC3334 or alike IC. If the battery
> voltage and the inductivity of the coils are nearly constant, why to use an
> intelligent ignition driver IC? What do You think is this a good idea?
> 
> Robert Sandor

Robert, are you talking about dwell or timing?  Timing is when you want
the spark to occur, dwell is how long you let the coil charge before the
spark.  I can't speak as to how well your feedback scheme would work as
far as timing goes, but it seems like a lookup table vs RPM would be
simpler than a control loop based on RPM that continuously changes the
timing.

As far as why a driver IC is used, it allows the coil to charge faster
because higher voltage is applied at first, then reduced as current
builds.  More efficient than the "point" style ignition with an inline
resister and probably a higher impedance coil, although that worked fine
for years.

--steve


-- 
Steve Ravet
steve.ravet at arm.com
Advanced Risc Machines, Inc.
www.arm.com