Long Lead Thermistor Temperature Measurement

EMPTY_C

Hey all,

I would like to try an utilize the following thermistor in one of my temperature probes (PS103J2) located at: http://www.ussensor.com/node/6. This probe will be mounted remotely (Between 10 to 25 feet from the BS2). I was planning on using the RCTime command but it seems that there could be some additional error on the length of my leads and the tolerances of the capacitor. I am trying to keep to the +/- 0.1 deg. C accuracy of the thermistor but a litter lower accuracy can be tolerated.

- Should I use the RCtime command and calibrate it with a known resistor value using the required line length?
- Is there a better way to get the information from the thermistor to the BS2?

Thanks for your help,

MT

Tracy Allen

I'll recommend against using RCTIME for a measurement that requires ~0.1% resolution, particularly for a thermistor out of the end of a cable. The RCTIME command depends on the switching threshold of the Stamp itself, which has temperature drift over time and is also subject to internal processor noise. A sensor out of the end of a cable may be subject to additional noise pickup from external sources, although you can probably improve that with proper shielding and grounding. It's not that it won't work. It will just be hard to obtain that level of accuracy.

Better if you can use an ADC located close to the thermistor, in a ratiometric scheme. Or a different sensor such as a DS1620 or TMP04 that returns data in a digital format, with calibration.

Leon

I'd put a suitable MCU with an ADC (such as a small PIC) close to the thermistor and transmit the data to the BS2 using RS-232. or RS-485.