Thermocouple Noise and Calibration SCB-68

[Jeramyk]

I'm measuring two Thermocouples on an SCB-68. Measuring voltage directly and converting or using MAX I get very steady but inaccurate measurments (they read about 8degC below known values).

If I short a channel and read it, I find that there is an offset in the micro volt range. Measuring this and subtracting it from the TC readings I find that my average readings are now pretty accurate but very noisy (+/- 1degC).

I also notice that if I increase the sampling rate the offset changes (which results in a change in the TC readings).

Is there a better way to eliminate the offset or do I have to tolerate the noise?

[Sean_C.]

Hello Jeramy,

Because thermocouples produce such small signals (mV range), it is best to apply some sort of signal conditioning before measuring the signal. Gain should be applied to amplify the signal and make better use of the range of the ADC. Hardware filtering helps remove noise from the signal. Please see the following tutorial for more information on thermocouples:
http://zone.ni.com/devzone/conceptd.nsf/webmain/806355D26B7378B1862567EE0003A842

National Instruments offers a variety of signal conditioning products. Please see the following Knowledge Base for information on these products as well as the pros and cons of each:
http://digital.ni.com/public.nsf/websearch/F0196F5B6CD8F60286256D4B00728B78?OpenDocument

To correct the offset, I would run a self-calibration on the DAQ device. You could also reduce the noise in your measurements by oversampling and averaging the signal. That way you don't have to buy new hardware.

Let me know if this does not help.

Regards,
Sean C.