Calibration of MET3A Temperature Measurement System

The temperature measurement system of three MET3A prototype units were calibrated over the range of –50 to +60° C.  Subsequent testing demonstrated system temperature conformance that was well within +/- 0.1° C, referenced to a calibrated and NIST-traceable reference thermistor probe.

The MET3A features a precision thin-film RTD to sense temperature.  This technology requires a calibration technique that is somewhat different than the technique commonly used to calibrate temperature probes.  For MET3A temperature calibration, a thermal chamber is used to control environmental temperature, and an independently calibrated and NIST traceable thermistor-based thermometer is used as a temperature reference. 

The MET3A humidity/temperature sensor is surrounded by a small thermal mass that features a bore designed to accept a standard 0.125” diameter temperature reference probe.  The bore terminates near the humidity/temperature sensor, assuring isothermal contact between the sensor and the temperature reference probe.  The MET3A RTD sensor and electronics can therefore be simultaneously calibrated by taking reference temperature and MET3A temperature measurements at various stable temperatures over the range of –50 to +60 °C, and adjusting the MET3A temperature coefficients to correct for combined temperature sensor and electronic nonlinearities.   This process yields system temperature measurement accuracy of +/- 0.1 degrees C or better over the temperature range of –50 to +60 degrees C.

The Device Under Test (DUT) is placed in the thermal chamber, and is connected via its serial interface to a PC running Digiquartz Interactive software (DQI).  The unit is queried to ensure that it is configured with default calibration coefficients.  The reference thermistor probe is inserted into the thermal mass, and is read by a Hart Scientific Model 1560 Thermometer Readout.  The traceable accuracy of the thermistor probe and thermometer readout is +/- 0.016 degrees C over the range of –50 to +60 degrees C, based on a multi-point calibration performed at the Hart Scientific metrology lab. 

Calibration data are acquired at –50, -25, 0, 25, and 60° C.  At each temperature, the indicated temperature from the DUT and the reference thermistor are monitored until thermal equilibrium was reached.   Thermal equilibrium is typically reached after soaking the DUT at the desired temperature for two hours.  Once the DUT temperature has stabilized, several DUT and reference thermistor temperature measurements are taken, averaged, and recorded.

At each temperature, the equation T = E + Fx + Gx² is solved for x, where T is the indicated MET3A temperature, and E, F, and G are the default MET3A coefficients.  The resulting values of x are fitted to the corresponding reference thermistor temperature measurements using a quadratic model, yielding a new set of corrected coefficients.

The DUT is configured with the corrected coefficients. Conformance data are acquired at    –50, -25, 0, 25, and 60° C.  Data are typically acquired over three complete temperature loops.  At each temperature, the indicated temperature from the DUT and the reference thermistor are monitored until thermal equilibrium is reached.  Thermal equilibrium is typically reached after soaking the DUT at the desired temperature for two hours.  Then, several DUT and reference thermistor temperature measurements are taken and averaged, and the averages are recorded.  Temperature error at each test temperature is calculated by subtracting the reference thermistor temperature from the indicated DUT temperature.

Figures 1–3 show the results of temperature calibrations performed on three MET3A prototypes. Conformance data were typically acquired over three complete temperature loops.  In all three cases, system temperature conformance to a highly accurate and traceable temperature reference probe are well within +/- 0.1° C over the range of –50 to +60 °C.  Refer to Figures 1–3.

Figure 1. SN 73354

Figure 2. SN 74906

Figure 3. SN 73382

©2004 Paroscientific, Inc.