Real-Time Measurement and Control

cancel
Showing results for 
Search instead for 
Did you mean: 

Best way to calibrate Thermocouple and RTD sensors

Hi everyone, I'm using 3 thermocouples and 3 RTDs for my application. I'm using a compact rio 9039 to acquire data through NI 9211 and 9216 modules respectively and using Labview for the programming. I am also familiar with MATLAB. I already have a thermal which can be used as a calibration source. 

I want to know what is the best way to calibrate all my sensors? There is a lot of stuff I came across on the internet but none specific so please recommend the various ways I can do this.

I'm fairly new to this environment.

0 Kudos
Message 1 of 3
(2,667 Views)

Hi ashwinsn,

 

as you are using cRIO you can implement your calibration of thermocouples and RTD either on FPGA or using Scan Engine.

You can find shipping examples directly in LabVIEW.

NI 9211 (Scan Interface)

NI 9211 (FPGA Interface)

Converting NI 9211 Data (FPGA Interface)

 

Hope this helps.

Regards,

Patrik
CTA, CLA
Helping (sharing) is caring!

If the post was helpful - Kudo it.
If the post answered your question - Mark it as Solution.
0 Kudos
Message 2 of 3
(2,612 Views)

Are you referring to a NIST-traceable calibration?

 

The signal conditioning modules need to be calibrated using appropriate equipment (i.e. extremely accurate voltage sources and measurement) in a temperature and humidity controlled environment.  NI can perform these calibrations, as can many third-party calibration labs.  For temperature instruments such as thermocouples and RTDs, these are typically calibrated using dry block calibrators such as these:

 

https://us.flukecal.com/products/process-calibration-tools/temperature-calibrators/dry-block-calibra...

 

...though if you know your ambient pressure accurately, a two-point calibration within a particular limited range is possible using an ice water bath and boiling water.  Generally though, you want your instruments calibrated over the entire in-service measurement range, so that you never have to extrapolate.

 

The best way depends on your requirements.  More points is generally better, but much more time consuming.  You can perform a linear regression for a linear calibration factor, or implement a lookup table, or do a polynomial fit or something more complicated.  What are your requirements?

 

 

 

0 Kudos
Message 3 of 3
(2,483 Views)