Thermocouple noise

[EJ347]

I am attempting to setup a thermocouple array for a project at school, I am using a USB 6211. What i have found is there is large magnitude noise in a 50-60 hz range.

My vi is currently set up using the daq mx assistant and takes three differential voltage inputs. I would like to continue using the daq mx assistant because it allows for cold junction compensation and easy calibration. As of now the three thermocouples have been working correctly, however they deviate from constant values substantially, and will drift from the correct value by as much as 60 degrees or so and come back, the speed of the shift is dependant upon the sample rate, or so i believe from trial and error. I removed most of the deviation from a constant value of short periods of time with a low pass (3rd order i believe) butterworth filter at 1hz. The temperature only deviates about .5 degrees which is acceptable because i only need to know the temperature change, not the actual temperature. the climb and fall problem seems to still be present.

My questions are what can i do to eliminate the climbing and falling voltage, I was thinking that i could average the peak values to eliminate it, i would like to know how this could be done, or any other conditioning that could be useful to solve this problem. The second issue is that my signal is filtered after the thermocouple daq mx task is completed so calibration is innefective because the noise is not controlled prior to the clibration phase. I tried to right click and convert to VI but all that it is replaced with is a signal aquisition diagram (or so I think) and none of the thermocouple cjc or calibration parameters. Can I aquire the signal filter it and assign it to a channel that the daq mx assistant can use as an input somehow?

any help is greatly appreciated

[Jake_G]

Hey EJ,

There are many different options for signal conditioning.  There is some information about signal conditioning on the page Signal Conditioning Fundamentals for Computer-Based Data Acquisition Systems.  Also, with all that drift, I wonder if your system is grounded properly.  To make sure, you can check out the Field Wiring and Noise Considerations for Analog Signals page. 

Regards,

Jake