Bob Lunn
09/30/97 09:00 AM


>> Umm, well what we actually did was to use the sender that's
>> installed in the fuel tank!  :)
>
>    Yeah, I was waiting for someone to say that.
>
>    The "fuel-consumption display" that I was talking about
>    is one that shows INSTANTANEOUS consumption and forecast
>    mileage-to-empty based on the way the car's being driven...
>    The ones that I've seen seem to react more quickly than they
>    could if their only input was from the tank sender.

Ok, I was being a bit disengenuous.  We did, in fact, use the
instantaneous flow data from the engine control computer to
modify the calculated fuel-to-go from the fuel sender.  This
gives a small improvement in the perceived responsiveness of
the readout.

However, the fuel sender was very much the base-line figure.
The modified fuel-to-go was never allowed to diverge from the
fuel sender by more than a specified amount.

It's worth noting that while you typically get about 200 grads
from a fuel sender, there's also a lot of dither.  In practice
you can get pretty good figures (and even quite good rate-of-
change figures)!  Of course, it was a requirement of the contract
of purchase for the vehicle that the owner never park on a hill!

Curiously, we were required to modify the sender output to
compensate for expansion of the fuel tank with age, but not
to compensate for change of volume of fuel with temperature.

___Bob