----- Original Message -----
From: "Pic Dude" <picdude@AVN-TECH.COM>

> Almost every fuel-tank sender I know of is resistive and
> runs electricity thru the tank.  Except for that floating
> stick protruding out of the tank on a Piper J3 cub.
>
> For the most part, they work fairly well on an automobile,
> though not very accurate -- both because of the meter
> innacuracy and the fuel being uneven in the tank at most
> times.

So, how could we implement a fuel gauge that was accurate?  I have a few
ideas:

The obvious, but expensive solution springs to mind.  Why not create an
optical grid and detect fuel level by maping the surface in the fuel tank?

Well, it's hard to do and it's not cheap.

The next thought I have is to measure flow rates in and out of the fuel
tank.  Some cars have a hot-fuel return, which prevents vapour lock, and
such like things, so it would take 2 flow rate sensors.  And there would
need to be an input system of some kind, which would allowthe user to input
how much gas was added when they last filled up.  You could even add a
sensor which operated off of the existing system to auto calibrate itself,
but which used several mercury switches (or something similar) to determine
what its orientation was.  That way, you could avoid the issue of the user
inputing the wrong number etc...  What would be a good system in that
respect is that you could use the user's input as long as it was within a
certain tolerance of the output of the potentiometer (on level ground as
detected by the mercury switches, or what have you), but then go with the
potentiometer once the stored value came too far out of tolerance.

Eh.. just some random ideas, maybe someone can have some fun with them.

--Brendan

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu