KISS System

[Gary Derian]

The Bosch CIS system has a few subtleties that provide it more information
than you think.  Two that come to mind are WOT and acceleration enrichment.
With a 4cyl engine, the air flow through the throttle pulsates with each
cylinder.  At part throttle these pulses are not seen by the air flow
sensor.  At WOT they are and the result is an average sensor position that
indicates more air flow than there actually is which provides a built-in
enrichment.  This only works on 4cyl engines.  6 and 8 cyl air flow is too
smooth.  Systems for these engines have a separate WOT enrichment.
Acceleration enrichment is handled by keeping the plenum at 1.5 times the
cylinder displacement (see related thread in this group).  When the throttle
is opened suddenly, the partial vacuum in the plenum must be filled in
addition to the additional air ingested by the engine.  This causes
enrichment for a short period of time until the plenum reaches steady state.

Gary Derian <gderian at cybergate.net>

H.W. wrote:
>I advertized for sale a very simple mechanical fuel injection system
>the other day which has only 2 control inputs.  One is airflow, the
>other engine temp.  No RPM input, no throttle position input, no
>vacuum input, no exhaust O2 feedback loop, Just airflow and temp.
>    This system worked very well, was efficient, and very drivable
>which suggests to me that this in enough information to do the job
>well.  The objective after all is to maintain a given air fuel ratio.
>It seems to me that if one could mimic this type of system the program
>necessary to control the injection should be very simple.  The control
>of fuel flow in this system is accomplished by designing the taper of
>the servo throat so that it works correctly.  Because of this design
>the injection system is not going to be adaptable to a wide range of
>engines.  However if a airflow sensing servo were designed with a
>constant taper throat, it should not be difficult at all to devise a
>simple lookup table for each position.  Cold operation could be
>accomplished outside the computer control system in exactly the manner
>that Bosch did in this system.  By varying the resistance (mechanical)
>against the servo arm using regulated fuel pressure from a temperature
>controlled regulator.
>    I may be stupid about electronics, but it seems to me that most of
>the EFI systems out there are way too complex.  Anybody out there
>building a KISS EFI system that doesn't require operator intervention?
>H.W.
>