Type T thermocouples are commonly used in industrial measurement applications due to their accuracy relative to other thermocouple types, low cost, and the ready availability of measurement equipment. Type T thermocouples are very effective when used in differential measurements, as there is no cold junction compensation necessary for the connections to the measurement equipment. Type T’s published accuracy specifications result in its frequent use in low-temperature applications. An examination of over 250 samples from a number of manufacturers has been completed for this investigation. Samples were compared to a standard platinum resistance thermometer (SPRT) at the LN2 boiling point along with four other standardized measurement points using a characterized ice point reference, low-thermal EMF scanner, and an 8.5 digit multimeter, and the data were compiled and analyzed. The test points were approximately −196 °C, −75 °C, 0 °C, + 100 °C, and + 200 °C. These data show an anomaly in the conformance to the reference functions where the reference functions meet at zero. Additionally, in the temperature region between −100 °C and −200 °C, a positive offset of up to 5.4 °C exists between the reference function equations published in the United States in ASTM E230-06 for the nitrogen point and the measured response of the actual wire. This paper also examines the historical and technological reasons for this anomaly in the US reference function. The study concludes that Type T thermocouples typically do not conform to the ASTM E230-06 published reference function describing their performance when used to measure temperature in the range of −100 °C to −200 °C.
