Re: Moisture sensor

["Karam Karam" <kzkaram@xxxxxxxxxxx>]

I should imagine its a similar idea to the gypsum based sensor but=20
just using a pair of stainless steel probes. The trouble with=20
measuring moisture levels using conductivity is that you don't really=20
know what's in the water from one location to another and this can=20
effect conductivity tremendously. So the gypsum sensor tries to=20
ensure the local ionic conditions are always the same (ie. dominated=20
by the level of solution of gypsum ions, so mostly dependent on=20
moisture level only). I've no doubt that other ions from the=20
surrounding soil must eventually get in with the gypsum ones but the=20
effect is probably negligible.

The steel probe approach will no doubt tell you whether the soil is=20
moist or not but bound to be a much cruder result. Perhaps with some=20
regular method of callibration and for a relatively self contained=20
location it might be ok to get something a little better than just a=20
logic indication. So this might suit the DIY approach.

In either case an ac signal should be used to avoid polarisation=20
(remember school chemistry ;-) ). Though I've seen comments that the=20
frequency (and so harmonic content of non sinusoids) affect the=20
measurement I'm not too sure about whether this means affecting the=20
repeatability or just the range - back to the electrochemistry books=20
I guess...=20

I would guess that for relatively crude measurements a simple=20
alternating square wave @ 1KHz should be ok and need only be switched=20
on at polled measurement times.

Karam



--- In ukha_d@xxxxxxx, "David Gumbrell"
<david.gumbrell@...>=20
wrote:
>
> Getting back to moisture sensors - I wonder what is inside those=20
cheapo
> "spike" types with the meter on ? Anyone broken one open ?
How=20
accurate
> does this application need to be (especially as it could rain=20
torrentially
> immediately after automatic watering if met forecasts are not=20
integrated) ?
>=20
> Cheers,
> David
>=20
>=20
> On 4/15/07, lexfordparc <paul@...> wrote:
> >
> >   Hi Karam,
> >
> > I have found a possible sensor which has a UK distributor and=20
costs
> > <=A325.
> >
> > http://www.irrometer.com/landcat.htm#Watermark
> >
> > I have read through the blurb and it seems to require 5VAC to
run=20
and
> > the resistance will vary according to moisture from 500 to
30,000=20
Ohms.
> > Could you suggest what would be needed to get this working with
my
> > Idratek system?
> >
> > Many thanks,
> >
> > Paul
> >
> >=20=20
> >
>=20
>=20
> [Non-text portions of this message have been removed]
>