On Mon, 20 Jul 2009 07:34:29 -0500, Jules wrote:
On Fri, 17 Jul 2009 19:40:11 +0100, Andy Burns wrote:

Jules wrote:

what's the best (DIY or "professional") way of
measuring the frost line during Winter and/or monitoring ground temps at
different depths?

Make a string of 1WIRE temperature sensors spaced a few inches apart,
and bury it, you'd have to give some though to waterproofing it, pot the
individual sensors in epoxy? seal the lot in some heatshrink tube?

True... I was thinking along those lines, but not liking the need to have
sensors all over the place. As you say though, I can probably just make up
a 'rope' with them at different depths and that should be quite neat. My
only worry there is whether whatever I seal them in will somehow distort
the readings.

I did something similar with my fishpond (not measuring the frostline, but
the temperature at various depths, to see how the little blighters fared
during the winter).
I used LM335 sensors - cheaper and less hassle than 1-wire jobs, as these
output a voltage of 10mV/degree[1] so you can just attach them to a DVM.
I sealed these with bathroom caulk to keep out the water and it worked very
well. The temperatures won't vary at all fast, so there should never be a
gradient across the caulk. Also these sensors take very little current
so the self-heating is negilible.

However, for the frost line, what you'll really measure is how deep the
frost penetrated *that* *year*. If you need this info to base the depth
of foundations on (as opposed to simply wishing to know for your own reasons)
How much of a margin you add for bad winters is a matter of judgement.

[1] when powered, the voltage "output" is really the voltage drop across
the device, which is temperature sensitive and remarkably linear.