Temperature measurement using USB 6351

[ncm]

I am using a T thermocouple and I am using the DAQ, USB 6351.  What would be the most accurate LabVIEW program just to measure the air temperature?  When I use Acq Thermocouple Signal.vi it runs through many samples and does not seem even close to accurate (I am getting readings between 16 and 21 C and inside this building it is closer to the low to mid 20's).  Any suggestions?  And can the USB 6351 give me temperature readings that are +/- 3C accuracy?

[0utlaw]

Hello ncm,

We'll need more information about your setup before offering specific advice (To start with- what measurements do you expect?), but in general you'll want to average multiple samples when acquiring from a thermocouple- they are generally quite noisy, particularly when used to measure air temperature.  Air currents are rarely as static as you think they are.

Some things to keep in mind:

-Cold junction compensation (manual vs. automatic)

-Thermocouple accuracy

-DAQ Analog input range (smaller range will give you higher resolution, which is important when measuring small signals such as the uV differential coming from your thermocouple)

The X-series USB DAQ has a pretty wide range of capabilities and should be able to acquire high-accuracy signals from just about any basic sensor with proper conditioning and a suitable task configuration.  I would posit that your limiting factor is going to be the accuracy of the thermocouple itself. +/- 3 degrees C should absolutely be achievable.

Here are a few things I'd recommend reading through:

Making a Thermocouple Measurement With NI LabVIEW

KnowledgeBase: How Do I Reduce Noise in My Thermocouple Measurements?

Edit- This is the LabVIEW discussion forum, so you may also want to consider posting your hardware-specific questions here:

NI Multifunction DAQ forum

https://forums.ni.com/t5/Multifunction-DAQ/bd-p/250

Regards,

[ncm]

Thank you for the link for the DAQ forum, that will be helpful.  As for the LabVIEW portion, I attached a screen shot of the temperature example that I am using.  I have the T thermocouple hooked up to the USB 6351 to pin 23 and 24 (which is AI.6 +/-)  The major issue that I am running into is that the temperature seems good for a few seconds when running this program but then drops off unexpectedly (see attached)  I am not sure the reasoning for this  I need to know how much data I should feel comfortable that is accurate so that I can average the signal correctly.