Can I convert j-type thermocouple data to k-type?

[SAMullin]

Hello,

I'm using an NI 9213 to read thermocouple data.  The device was configured for j-type thermocouples, but k-type thermocouples were used to take a fair bit of data.  I should be able to convert the data directly with an exact equation if I knew how the temperatures are calculated from the raw voltage signal.  Would anyone happen to know the exact algorithm and constants to use?  

The settings were: read in degrees C, constant CJC at 25 C, max 100, min 0 (these are all defaulted).  I converted to code to k-type by wiring in a constant and now it reads properly. 

Other ideas I tried:

Re-calibrating my temperature stages is possible (but would take a huge amount of time).  I also thought about measuring the temperature while switching back and forth between k and j-type, but that would introduce a lot of noise.  I also tried to duplicate the vi and have one set at J and the other at K, but it wouldn't let me access the device from two vi's simultaneously...  

*EDIT* I also looked up the raw voltage signals from k and j-type thermocouples, but there was no way to match it to the data, and upon further reflection it wouldn't help anyway since the thermocouple was K-type and it was a matter of incorrect settings. 

[Sarah_Y]

Hi SAMullin,

You are correct, you will not be able to run 2 different VIs at the same time using the same hardware.  This will cause a reserved resource error.  This website describes briefly how thermocouple voltage is computed:

http://en.wikipedia.org/wiki/Thermocouple

You can also find tables online that have chats have already calculated temperature based on raw voltage.  Here is an example of one of those charts:

http://www.pyromation.com/downloads/data/emfj_f.pdf

Here is the website where I found the chart.  This site had tables written for both J and K types in Celsius and Fahrenheit:

http://www.temperatures.com/tctables.html

Using the formula from wikipedia you can derive some sort of algorithm or convert your data point by point.

[nathand]

I don't know if it's the same calculation that NI uses when reading thermocouples from DAQ devices, but NI provides a VI for converting voltage to temperature and vice versa for several thermocouple types.  Look at "Convert Thermocouple Reading.vi" under Mathematics->Numeric->Scaling.  If you look inside that VI there's a subVI for the reverse conversion.  I haven't had great luck using this VI for accurate temperature readings but I suspect that has more to do with trying to use a an analog input that isn't designed to reading the voltages in the range that a thermocouple generates than a problem with the VI.

[SAMullin]

Hi Sarah and nathand,

Thanks for the suggestions.  I ended up using a combination of both of your inputs.  ultimately I ended up opening up the LV files and seeing how the thermocouple values were calculated.  I ultimately ran into a problem that I couldn't open the library that contained info about how the CJC was calculated.  In order to account for this I used an arbitrary fudge factor to scale my data.  The revised estimates of my temperature calibration were that my control and sample thermocouples were in agreement to better than 1 degree C.  I did not believe this, so I re-calibrated my stage with the correct settings on the thermocouple reader.  In the re-calibrated run, the values were actually within 1 degree of the setpoint, so it looks like the back-calculation was correct, even without knowing how the CJC was calculated.

So, essentially that was the correct solution, though not precisely since I had to use a fudge factor.  The help was and is most appreciated.

Regards,

Scott

[Ryche]

I logged Ktype thermocouple data with the meter set for J type. I noticed my temps are lower but at room temp are similar.

What equation can I use to convert temperature readings?

Seems like a non linear equation? As the temp difference increases with temperature.

I do not have voltage readings...only temperature readings.

[nathand]

See my previous message in this thread. There's a VI that will go from temperature to voltage, and one that will go from voltage back to temperature, both of which have inputs for the thermocouple type. So you would wire in your voltages and set it to type J, then take that voltage and convert it back to temperature with the thermocouple type set to type K.

[Ryche]

Understand.

But I used a Fluke 54 thermometer and only have temperature readings to work with.

So if there is a formula I can apply to the temperature data in my spreadsheet, that would be helpful to convert it to a ktype reading.

Thank you for your reply. I hope this info helps.

[nathand]

I don't think you're understanding me. There is a VI that, given a temperature and a type of thermocouple, will provide the corresponding voltage. So, you can convert your temperature readings into voltages, without any information other than what you already have. Then, you can calculate a new temperature for the same voltage but a different type of thermocouple. The equations are all in the VIs I mentioned.

[Ryche]

Hi Nathan,

I do not know how to find the VI's.

I tried to find this ....do you have a link?

Appreciate you help!

Rcyhe

[nathand]

Did you look in the palette mentioned in my earlier message? It's here:

If you open the "Convert Thermocouple Reading" VI, you'll see that it contains subVIs for "Temperature to Volts" and "Volts to Temperature" which contain the scaling equations for various thermocouple types.