I have an application in which I will be measuring temperature and relative humidity at four locations in a building. I will use the same setup in 5 different buildings.
Excellent interchangeability and precision of the temperature sensors is critical, particularly 20 and 32 degrees C. The "accuracy" of the sensors, i.e., the closeness to a known standard, is much less important than interchangeability in this application.
I am planning to use CS215 sensors for the relative humidity measurements and I am trying to decide whether I should a) rely on the C215s for the temperature measurements; b) add a separate thermistor for the "definitive" temperature measurement; or c) adjust the CS215 temperature readings (probably by adjusting the offset in the program) based on the difference between the sensor readings that I observe in a side-by-side test of the four sensor or in a comparison of each sensor to a known standard.
According to the various spec sheets and docs that I have been looking at, it appears that the CS215 has a temp interchangeability of 0.3-0.4 degC, while something like the CSI 107 or the Apogee ST-100 has an interchangeability of 0.1-0.2 degC. Therefore, I'm leaning option B above. However, this might put me over budget and will add complexity to the system. Therefore, I'd prefer to go with option A or C for those reasons, but not at the expense of measurement quality.
I'm wondering if anyone has any recommendations.
How many meters of cable is needed between the sensor and the logger?
It will be one or more than one logger, for example one for every building?
What kind of logger you use?
How frequent are the measures?
The cs215 has a signal practically immune from cable length while 107 could degrade the measure with a lot of cable.
You can choose later to install the 107 instead of cs215 temperature?
Appraise the major savings in the cost of the cables due to the type of wiring is possible due to the characteristics of the SDI-12 protocol present in cs215.
You may check in a test environment (stable and insulated) if the 20 sensors measure equal, even as annual or semestrial control with offset report.
Smile
Here are my replies:
- There will be up to 25 meters of cable between each sensor and its associated datalogger.
- There will be one logger in each building.
- I am using CR1000s
- I would like the scan rate to be 5 seconds, but I can slow it down to 15 seconds, if necessary
The CS215 uses a "band gap" semiconductor temperature element that are usually considered quite stable and reasonably linear, although there are other components in the measurement path that will add extra drift.
Generally most sensors when tested are better than the 0.3C figure, at least at room temperature. You could improve this with an offset correction. The 0.3 figure is probably quoted more due to limitations in the calibration rig used by the manufacturers of the sensor element where they calibrate multiple elements en-masse in a chamber. I believe a relatively simple two point calibration is done in the ambient temperature range.
Over a wider temperature range the sensors are not so good as basic sensors I suspect due to temperature drift of the other components mentioned above. You do though need to consider those errors too when looking at other probes, i.e. the error of measurement by the datalogger if using a non-digital probe.
Of course the advantage of a separate temperature probe is that it is more easy to calibrate if required, i.e. it can be put in a water or oil bath.
I agree with Andrew