>3 @ $100

  I commiserate.

> build your own

  Yes there is, there are several ones in fact. Problem: not readily
reproductible under production conditions.

  So here goes: Take a tube of inox steel about 20 mm long, 5 mm dia.
Solder a wire to it. Take a piece of dense gauze (medical supply) or a
piece of lint-free cloth, cut into strip suitable to be wound around tube
about 2 windings. Soak it in a saturated kitchen salt solution (dissove as
much salt in 1/2 glass of water as you can add until salt starts piling up
at the bottom), then wind wet around tube. Wind a piece of
corrosion-resistant wire several times around this (de-insulated
wire-wrapping wire works), and secure (do this while wet). Dry the
contraption completely. You may want to select a pair of electrodes that
withstand HCL and HCLO. Platinum and Gold are ideal, alas the price...

  This works best when sensed with AC current, with DC it tends to
polarize after a short time and cause misreadings. To obtain AC with a
PIC, drive an output pin at a few Hz, and use the 1/2 Vcc supply as
reference for the sensor ;).

  The sensor's impedance is very high in dry air and becomes low when wet
air is present. There is another variant that uses graphite electrodes.
That is much more corrosion resistant (lasts almost forever), but is very
delicate to build.

  Sensor response is slow (1/2 hour) for small variations, but very fast
for getting wet (practically instant). It should not be immersed.

  There is another type of commercial sensor for this, that is much
better. There is one of these in each VCR and camcorder. It's piece of
white ceramic with electrodes deposited on it and 2 connections. It shows
high impedance when dry and low (5 kohm or less) when humid. This is a
GO/NO GO sensor. The transition is abrupt and response time of the order
of 5-10 seconds. You can scrap one from a dead VCR for testing.  These
sensors like to get contaminated sometimes, and get stuck in the low Z
region. In this case, they are useless. To determine, measure sensor
removed from VCR in Ohms scale, should show incredibly high (much more
than 100 kohm in dry air). They are very sensitive to dew from human
breath among other things. If tripped, allow to dry for 30 minutes.
Cigarette smoke seems to be a contaminating factor but I'm not certain.
These must NEVER be immersed. They are known as 'dew sensors'.

hope this helps,

Peter


On Mon, 18 May 1998, Kolesnikoff, Paul wrote:

> Hi,
> Awhile ago there was a thread on humidity sensors. One of the
> recommended vendors was Panametrics with their RH2 sensor. This is
> supposedly a low cost humidity sensor. However, their "sampling" program
> is to sell me 3 for $100US. This hardly qualifies as low cost or
> sampling. Does anyone have any other sources for a humidity sensor or
> better yet a way to build your own?
>
> Thanks,
> Paul K.
>
> PS I will be interfacing this to a PIC. So it is not quite OT.
>
>
>