Thermocouple Readings

[koolchip]

I'm using a NI USB-6210 with SignalExpress to take temperature readings from a type K thermocouple. The readings are all over the place. If I add more channels to aquire from, the readings get within a 2 degree C range, which is close to acceptable. Still seems like a large swing for a thermocouple device that is sitting in a room close to room termperature. Is there anything I'm doing wrong? I don't see anyother way I can hook things up.

Thanks

Chip

[DFGray]

Since the USB-6210 is not specifically designed to read thermocouples, there are a plethora of things that could be going wrong.  Without your project and a connection diagram, we can only guess.  Here are a few things to consider:

1. The resolution of the 6210 is inadequate and you are just getting random noise.  Check the specs and run through the thermocouple calibration code to check this.
2. You are taking data too fast for the 6210 to properly settle between readings.
3. You are getting voltaic effects between the thermocouple and the connector to the 6210.
4. Your connections are loose.

All of these can be worked around.  Please post your SignalExpress project and your connection diagram.  You can create a connection diagram with almost any drawing package (OpenOffice and Inkscape are free) or hand draw it and use a document scanner to digitize it.  Please post at least your project.

[koolchip]

I hope this can help you help me.

Thanks

Chip

[DFGray]

Looking at a specification chart for the type K thermocouple, it appears that the response is about 40 microvolts per degree C (note that iron/copper-nickel is a type J, not a type K, but responses are similar).  Looking at the specs for the NI USB-6210, if set to the lowest gain, it has a resolution of 4.8 microvolts and random noise of 12 microvolts.  It should work OK for you.  Your connection diagram is correct, but you list a type J, not type K thermocouple.  Change the thermocouple type to type J (assuming you have an iron/copper-nickel).  Also verify that the thermocouple itself is well bonded at the tip.  Verify that your connections to the device are solid.

If this doesn't work, let us know.

[koolchip]

I checked the connections and they are solid. I also checked readings with type J, K settings and different thermocouples. The thermocouple gets the proper readings on another device. Still getting the same readings.

Thanks

Chip

[Jordan F]

Hi Chip,

I noticed that you have not received any response on your last post so I wanted to provide some more help if you still are having trouble.  One thing that I noticed was that you are using a differential connection with your thermocouple (AI0-AI8, AI1-AI9).  It is recommended that you use a single ended ground referenced configuration (RSE) when you are measuring a floating signal such as a thermocouple.  This could be what is causing the erratic behavior on your measurement.  I attached a jpg of the different wiring configurations.  

Also, what is the other device that you are using that gets a proper reading with the thermocouple?  What are the settings and how do you have it connected?

Field Wiring and Noise Considerations for Analog Signals

http://zone.ni.com/devzone/cda/tut/p/id/3344

[koolchip]

The other device is a hydra bucket. The researcher has decided to go ahead and purchase a hyrda bucket so I no longer need help on this issue. The USB-6210 has worked great for the temperature and humidity readings I'm taking via voltage. Thanks for everyone's help on this issue.

Chip