I am working on an experiment where I need to measure the Delta T across a heat ex-changer. I have two independent T Type thermo couple, one at the inlet and the other at the outlet. Both tied to an Agilent data logger. I normally put the data into an excel sheet and generate Delta T = T1 - T2. The problem is, I am dealing with delta T values as low as 0.2 C. Since a general T Type thermo couple has an uncertainty of 0.5, my results may not be accepted I decided to make a differential thermo couple, [like "M" shape] one junction at T1 and other at T2. The wire coming out now is the two copper wires, I believe that I can no longer use the two Cu wires in the data logger and select T type thermocouple. Thus I have selected the channel as a voltage measure. My question is will this arrangement improve the accuracy? How can I determine that there is an improvement or what is the reason for the improvement if it does improve. Second question is how to convert the voltage measured to Delta T? Should I use the eight order polynomial or should I use the seeback coefficient(41 microV/C) or should I use the standard tables? What is the difference in each of these method? Last question, how are these thermocouples calibrated in the lab environment? what is the general and easy procedure to calibrate thermocouple?