I guess you meant USB-6210... All depends on the accuracy level you want... You will meet 2 limiting factors :
- input noise of the DAQ board : M Series boards in general have fairly good performances, so without signal conditionning you can already reach measurements with noise lower than 1°C with the most sentitive input range (we've done this on one project for a customer). But for this you have to be carefull in your wiring, sampling frequency (highest possible), signal averaging, etc. To get better you will definitely need additional electronics (preamplifier and an antialiasing Low-Pass filter with cutoff not higher than half of your actual sampling frequency). You should find VIs to convert the measured voltage into temperature in the numeric palette (Scaling).
- cold junction temperature compensation : thermocouple as such not only measure temperature at their "sensing end" (the thermocouple junction), but are also influenced by the junction of the thermocouple wires (made of some specific materials depending on TC type) with copper at the connector block on the DAQ board side. To compensate for this parasitic part of the signal, you should measure the connector block temperature (e.g. with an RTD, so you actually need one more Ain chanel) and correct the measured total voltage for the expected thermocouple voltage at this temp. All this is done automatically when you use NI signal conditioning specifically made for thermocouple connection. But if you have a "reasonnably accurate guess" (with respect to your accuracy needs) of your connector block temperature, the V to T conversion VIs let you specify its value. At first order (that is neglecting the thermocouple non linearity), the overall error will be equal to the error in your temperature estimation for the connector block. Not really suitable for accurate (1° or less) measurements, but OK if you measure an oven at high temp +/- 10°C...
Hope this helps.
Vincent
