I don't expect any of the orginal posters to respond to this, but for others who found this by searching the discussion board...
I had a look at markwysong's VIs (since I also need to convert Type C thermocouple voltages to temperatures) - they don't work very well. The general idea is correct, but above 200degC the values are way off (maybe correct values weren't intended).
Anyway, I think that the polynomial coefficients given (extracted from 2nd and 3rd order polynomial fits) just aren't adequate. Here's what you need to do (step-by-step).
1. Find type C thermocouple voltage-temperature data (I used one from Omega http://www.omega.com/temperature/Z/zsection.asp - under 'thermocouple reference data'). Incidentally, this website gives Type C coefficients, but for temperature to voltage conversion. Voltage-temperature conversion coefficients are different (something I didn't realise for a while - d'oh)! It's the inverse polynomial coefficents that we want (which aren't given for Type C) - see note under point 2.
2. Using the General Polynomial Fit VI, generate coefficients for whatever order polynomial you want. I input data for the temperature range 0-1400degC (in 50 degree intervals) and got the following coefficients from a 6th order polynomial fit (can't remember which algorithm, but it doesn't seem to make much difference):
These coefficients reproduce the table's data much more accurately, and allow conversions from voltage to temperature. Note that the temperature data must be wired to 'x values' and voltages to 'y values' in this case. Wiring oppositely will give coefficients for temperature to voltage conversion.
Hope this helps some fellow Type C-er!