At 12:09 PM 1/15/00 , Tom L. Neal wrote:
>The dynamic range of the fuel injection feedback control loop is probably
>a factor of 100 or so (pressure and injection time).

It's strictly pulse time -- the fuel pressure is fixed (i.e. it is
regulated to a constant difference from the manifold pressure).  At 6000
rpm the available pulse time per revolution would be 10 milliseconds.

>  It's my
>understanding that the two primary feedback signals are the AFM, which gives
>the approximate point of operation within the dynamic range,

Up to 3500-4000 rpm at which point it runs wide open and the basic timing
(i.e. before mods for air temp, engine temp, engine throttle, O2 sensor) is
set by a pulse-time vs rpm map in system ROM.

>and the O2
>sensor, which provides a fine adjust within a very small percentage (1% or
>so?) at the optimum point within the dynamic range.

Close -- the O2 sensor gives a very sensitive indication of rich or lean,
but outside of an extremely narrow range it doesn't say *how* rich or
lean.  The adjustment range for the injection time is controlled by the ECU
electronics -- I dunno how much it is.  But it's much wider than the
measuring range of the sensor.

>   It's difficult to
>imagine a "manual closed loop"  configuration that would work better in a
>Vanagon, and easy to imagine that any would work far worse.

What, can't you turn the knob ten times a second?  Tsk, tsk  <g>.

>  Note that in
>the straight and level airplane example the throttle is fixed and the load
>is fixed, both of which are impractical when driving.

Not to mention that in the airplane you have two concerns: getting there
and not burning up your engine.  If you had to deal with emissions as well
you'd be taken out of the loop right smartly.

david
David Beierl - Providence, RI
  http://pws.prserv.net/synergy/Vanagon/
'84 Westy "Dutiful Passage"
'85 GL "Poor Relation"