At 14:10 01/11/99 +1100, James Cameron wrote:
>Then we remembered that the ants tend to crawl up the watering pipes,
>and so the volume of water delivered will vary over time.  Sure, the
>moisture level feedback would help to fix this, but not rapidly.  I'd
>prefer to design the system so that a "volume" of water is delivered
>rather than a duty cycle of valve opening.

maybe you can hinder the ants crawling up? something like a "siphon"
(german, didn't find it in the dictionary -- goes down and up again, so
that there's always water in the bend) maybe, in some central places in the
main pipes.

difficult to measure volume flow if the active cross section of the pipe is
not known. and if the ants can considerably change your cross section, they
sure can block any moving parts, too (like a paddle wheel to measure the
speed -- but you still wouldn't know the cross section, =if= the ants
change it). are you sure that the ants change the volume flow so much you
have to care about?

ge