EAS327. LABORATORY 4: Calibration of Temperature Sensors

• A pair of diode thermometers set up in series in a constant-current circuit (the supply voltage is nominally 25 volts, and need not be set exactly as the current regulator will compensate). The voltage drop across each diode is of the order of 200 millivolts, and may be expected to vary linearly with temperature.

• Two thermistors, each operated in half-bridge configuration. One of these has room temperature resistance of order 50K, the other order 11K; accordingly the control resistors are chosen as (nominally) 51K, 10K (you should measure these with the multimeter).

Method: Sensors will be immersed in a controlled-temperature water bath. The bath is well-stirred, thus uniform in its temperature T_b, ensuring all sensors are at the same (controlled) temperature. Your "standard" will be the temperature estimate provided by a thermocouple, one junction of which will be immersed in an equilibrium mixture of melting ice and water ("ice bath") that we will assume to be 0^o C.

Procedure:

1. Familiarize yourself thoroughly with the equipment and circuitry assigned to your group. Before commencing any measurements, you should be able to create a circuit diagram that describes your setup, paying particular attention grounds and noting that all power supplies are floating (as an example, here is a circuit diagram for the set-up used in an earlier year). Be sure to note which signal corresponds with which input to your logger.

2. Create a suitable program for your Campbell Scientific 21-X Data-logger. Single-ended voltage measurements will be suitable, and Full Scale Range should be 5000 mV for all signals except the thermocouple, which requires +/- 5 mV FSR. Enter a multiplier for the thermocouple such that the logger reads out the bath temperature directly (the sensitivity of your thermocouple will be 40 or 60 [microvolts per degree Celcius]).

3. Set the half-bridge supply voltage for the thermistors to 2.00 ± 0.01 Volts.

4. When you are satisfied you understand the various signals, perform a calibration over a range in bath temperatures T_b from -10 ^oC to + 30^oC, in steps of about 5^oC. At each bath temperature, take at least one 60 second average of your signals.

Analysis:

Tabulate and graph each temperature signal versus the bath temperature T_b deduced from the thermocouple voltage. Perform a linear regression (if appropriate) to determine for each sensor a calibration equation. Comment on any signs of non-linearity, if observed.

Back to the Earth & Atmospheric Sciences home page.

Last Modified: 13 Mar. 2005