"rmp" > wrote in message >.. .
> Usually strain gaged loadcells have very small deflections. So movement
> sticking is unlikely to be an issue.
>
> Software issue may relate to low cost analog to digital conversion, reading
> frequency, etc.
>
The deflection doesn't have to be large for friction to be a factor
and anyway the deflection on *my* digital scale is large enough to be
visible. If you go back and look at the data from the original
poster's experiment, the only real "mystery" here is why the results
are different between adding weight quickly and very slowly. The
easiest (and therefore most likely, IMO) way to explain that result
is, as Dave has said a number of different ways, by the effects of
friction: there is something "sticky", if you like, in the mechanism
between bowl and strain gauge; that stickiness is overcome if you add
(or withdraw) weight quickly but not if you sneak it on slowly. Any
other explanation is a stretch. The other possibilities that have
been mentioned - software, strain gauge/gage, a/d conversion, and so
on, are unlikely to vary with time. In other words, I'm with Dave.

- Mark W.