>No steps would work so long as the transistion from one step to another =
is
>good and repeatable. Seems to me like your solution would work fine in =
this
>case. Any info on the mechanical construction of this (like width of =
rod,
>how deep to make each step ec.)

I do not have any direct experience with the method, but heard of it =
being used for fluid level, I think it was in petrol tanks, where it was =
important to have no electric's inside the tank.

Time for some ASCII art

------\_________
                \---------\_________
                                    \
                            ________/
       _________/----------/
------/

If you get the idea. The LED and sensor (e.g. a reflective sensor unit) =
would be looking down the rod from the left hand end. each step causes =
some light to be reflected back, and the liquid covering each step fouls =
up the reflection coefficient of that step, meaning light does not get =
reflected back.

If you have access to a suitable lathe you may wish to try turning a =
piece of Perspex rod to try out the idea. The other trick which may work =
better for you, as I envisage the rod being on a slope inside the tank, =
is to use a rectangular rod, and steadily mill away steps on one side =
only. It may require some experimentation to determine if you get better =
results with the steps on the top or bottom side in this case.

happy experimenting

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.