Campbell Scientific CR10 CR10 Measurement and Control - Page 181

SECTION 13. CR1O MEASUREMENTS THERMOCOUPLE POLYNOMIALS - Voltage to Temperature Conversion NBS Monograph 125 gives high order polynomials for computing the output voltage of a given thermocouple type over a broad range of temperatures. In order to speed processing and accommodate the CR10's math and storage capabilities, 4 separate 6th order polynomials are used to convert from volts to temperature over the range covered by each thermocouple type. Table 13.4-2 gives error limits for the thermocouple linearization f unctions. TABLE 13.4-2. Limits of Error on CR10 Thermocouple Output Linearization (Relative to NBS Standards) TC Type Range'C Limits of Error oC -270to 400 -270lo -200 -200 to -100 -100to 100 100 to 400 x18 @ -27O t 0.08 t 0.001 t 0.015 -150to 760 -100 to 300 t 0.008 t 0.002 E -240to 1000 -240 to -130 -130 to 2OO 200 to 1000 t 0.4 t 0.005 t 0.02 K -50to 1372 -50 to 950 950 to 1372 t 0.01 t 0.04 REFERENCE JUNCTION COMPENSATION. Temperature to Voltage The polynomials used for reference junction co mpensation (convertin g ref e rence temperatu re to equivalent TC output voltage) do not cover the entire thermocouple range. Substantial errors will result if the reference junction temperature is outside of the calibrated range. The ranges covered by these calibrations include the CR10 environmental operating range, so there is no problem when the CR10 is used as the reference junction. External reference junction boxes, however, must also be within these temperature ranges. Temperature difference measurements made outside of the reference temperature range should be made by obtaining the actual temperatures referenced to a junction within the reference temperature range and subtracting. Table 13.4-3 gives the reference temperature ranges covered and the limits of error in the linearizations within these ranges. Two sources of error arise when the reference temperature is out of range. The most significant error is in the calculated compensation voltage; however, error is also created in the temperature difference calculated from the thermocouple output. For example, suppose the reference temperature for a measurement on a type T thermocouple is 300"C. The compensation voltage calculated by the CR10 corresponds to a temperature of 272.6C,a-27.4C error. The type T thermocouple with the measuring junction at 290'C and reference at 300"C would output 578.7 pV; using the reference temperature of 272.6"C, the CR10 calculates a temperature difference of -10.2'C, a -Q.2"C error. The temperature calculated by the CR10 would be 262.4"C,27.6'C low. TABLE 13.4-3. Reference Temperature Gompensation Range and Linearization Error Relative to NBS Standards TC Type Range'C Limits of Error oC T -100 to 100 J -150 to 296 E -150 to 206 K -50 to 100 t 0.001 t 0.005 t 0.005 t 0.01 ERROR SUMMARY The magnitude of the errors described in the previous sections illustrate that the greatest sources of error in a thermocouple temperature measurement are likely to be due to the limits of error on the thermocouple wire and in the reference temperature determined with the built-in thermistor. Errors in the thermocouple and reference temperature polynomials are extremely small, and error in the voltage measurement is negligible. To illustrate the relative magnitude of these erors in the environmental range, we willtake a worst case situation where all errors are maximum and additive. A temperature of 45oC is measured with a type T (copper-constantan) thermocouple, using the t2.5 mV range. The nominal accuracy on this range is 2.5 pV (0.1% of 2.5 mV), which at 45'C changes the temperature by 0.06oC. The RTD is 25"C but is indicating 25.3"C, and the terminalthat the thermocouple is connected to is 0.3'C cooler than the RTD. 13-15