[aklug] Re: Making a temperature sensor

Thanks for the feedback, y'all. The DS18S20 1-wire digital device is
most interesting -- especially seeing what you can do with owfs,
owhttpd and rrdtool. Very cool!

As far as finding an appropriate potting compound, a bit of research
turned up http://www.epoxies.com/therm.htm. After discussing my
application with sales, I ordered a sample of their 50-3151 FR
product to test. Will report...

Another question. In addition to temperature, I also want to measure
solar radiation, and came across a $20 pyranometer kit available at
http://mysite.verizon.net/brooksdr1/pyranometer/pyranometer.htm. It
returns a 0-2.5V signal. Dr Brooks says it needs a minimum of 12-bit
resolution. Until reading this feedback on other temperature
measurement ideas, I had been looking at a labjack USB A/D converter.
http://labjack.com/u3. But maybe this signal can also be measured
with these 1-wire devices and owfs?

Thanks,

Bob

>Date: Sat, 21 Nov 2009 16:14:39 -0900
>From: Jim Courtney <courtney@ieee.org>
>Subject: [aklug] Re: Making a temperature sensor
>
>You should look into the maxim-ic DS18S20 "1-wire" digital devices. They
>are about the same size as the analog parts, and with linux you can get
>up and running pretty fast. They have three leads, 5V, return, and data.
>You can get a 1-wire bus master for USB or RS232, and daisy-chain dozens
>of devices quite a distance away. Each has it's own unique address and
>the software goes out and finds them all. I'm running owfs and owhttpd
>to talk to the devices with a USB bus master. I use rrdtool to grab the
>temperatures from the owfs (one-wire file system) periodically and make
>graphs I can check out online. You can see them at 99645 dot com port
>3002 . For my only outside sensor I heat shrinked the whole thing and
>left the top of the device exposed so it responds quickly. If you want
>to pot the device, 2-part epoxy and empty shell casings should work. I
>think they'll fit in a .22. The only downside I have found with these is
>that they seem to bottom out at about -32F.

---------
To unsubscribe, send email to <aklug-request@aklug.org>
with 'unsubscribe' in the message body.
Received on Tue Nov 24 11:06:41 2009