Re: MMIR X-10 Macro IR Module

[Marc_F_Hult <MFHult@xxxxxxxxxxxxxxxxxxxxx>]

On Fri, 21 Sep 2007 12:28:01 -0400, "Robert Green"
<ROBERT_GREEN1963@xxxxxxxxx> wrote in message
<YsSdnXodl4fobG7bnZ2dnUVZ_s-pnZ2d@xxxxxxx>:

>Although thermistors are non-linear, aren't their outputs fairly
repeatable?

As stable as the voltage reference in most HA controllers/gizmos as long as
the self-heating is negligible or held constant.

>If I calibrate the output with a good digital thermometer, can't I always
>expect the output voltage to be the same for each calibration?

See above. Adding thermal mass and surface area reduces self-heating at the
expense of response time.

>I would imagine I'd want to know where things like 32, 65, 68, 75 degrees
>F, and maybe a few others are at precisely, but I'm not sure I'd need
>anything else.

Right. And you don't need to use  a 'look-up table' per se. You can create
a small number of distinct binary 'events' with IF:THEN or CASE statements.
(A "look-up table" implies a set of paired values, and are often used as
the starting point for interpolating intermediate values.)

>Depending on how I was using them, I might want the ability to set a point
>slightly above and below each of those marks.

Here's where a low-rent DIY solution can bite you. If you don't add some
sort of hysteresis or dead-band to your system, you can burn things up
and(or) cause other undesired  or erratic results. Think (eg) of a relay
that chatters because of noise/instability in the voltage reference that
your analog-to-digital converter depends on that causes it to toggle back
and forth between values.

>However, I just bought a number of alarm-capable thermos and hygros from
>Sper scientific that should be able to let me set alarm points right at
>the monitored location.  Instead of sounding their internal buzzers, I
intend to have them close a pair of
>contacts instead with that voltage instead.  At about $10 a unit, there's
>*nothing* I can build that would be cheaper or better looking.

In other words, you create the digital event external to the controller.

[snip]

>He's got some new stuff that
>looks very interesting,  though:
>
>http://www.phanderson.com/arduino/arduino_index.html

which depends on the user's programming of firmware in Yet Another Language
that itself is based on a two-year-old experimental program language that
is not yet out of Alpha development.. That should make for "very
interesting" results ;-)

... Marc

Visit my porch Sale of Home Automation and electronic gear at
www.ECOntrol.org/porch_sale.htm

Marc_F_Hult
www.ECOntrol.org