Trying to use thermocouple to measure voltage

[kkofori]

I have a two thermocouple that I use to measure temperature difference between two points, the two thermocouples are connected to two different channels on the keysight daq 973a , bear in mind that they are K type thermocoupoles. I have set those two channels to measure voltage between the same two points by subtracting the voltage from point thermocouples1 and thermocouple 2. the temperature readings are accurate but the voltage are widely different from the actual value. I am using this method to check the seebeck coefficient of substances. is there something wrong with trying to read voltage from thermocouples I dont want to introduce any probe to measure voltage as it will increase errors. Channels are 101 and 109

[RTSLVU]

A couple things come to mind....

For a K type thermocouple 300/°C is equal to 12.209 mV. This is why things like cold junction compensation is so important.

You also have a race condition in your code. You configure 101 and 109 as thermocouples then as voltage measurements. Then in your loop it's a race as to what one comes first. I am surprised this works at all. Are you using the "EZ" temperature and voltage VIs?

You should use two channels for thermocouples and two different channels for voltage measurements.

Also in my experience if there is voltage on a thermocouple. Such as measuring a MOSFET tab that is live voltage on it's tab in a switcher. This voltage can cause the Keysight to see it as an open thermocouple and effect the temperature accuracy.   

[altenbach]

A picture with messy code is impossible to troubleshoot.(overlapping elements and no way to tell what the control values are. Do you know that index array is resizable? Why do you use shift registers if you never even look at the content???)

---

@kkofori wrote:

Trying to use thermocouple to measure voltage

---

A thermocouple does not measure a voltage, it generates a voltage.

---

@kkofori wrote:

 but the voltage are widely different from the actual value. 

---

What you measure are the "actual values". Do they differ from what you expect?

Why are you measuring voltage differences if you want to measure temperature differences? 

Please attach your VI and more details what you are trying to do here.

[kkofori]

I am trying to measure the seebeck coefficient of semiconductor thin films that is the change in temperature vs the change in voltage. so I heat one side up and keep the other side cold and I use two thermocouple to measure thot and tcold. I then use labview to subtract thot-tcold. I also use the same channels and same thermocouple to do vhot-vcold. to get the voltage difference. However using an actual  voltmeter to measure the voltage instead of doing this maths gets me a different value .and you are right I am still a novice at labview sorry for the messy code.  So basically I export the data of voltage rise and temperature rise and use matlab to fit a linear curve of the voltage rise over temperature rise. I have noticed that no matter the material I keep getting the seebeck coefficeint of 41 microvolts/k which the same as the K type thermocouple.  I am using instrument drivers from Keysight daq973. What a I notice is that the voltage generatedby the thermocouples is different from me using a voltmeter to measure voltage across the sample.

[kkofori]

I am not using EZ voltage or temperature I am using 'get temperature measure' and get voltage measurements. My question is can two thermocouples be used as voltage probes when you connect them to a daq. Or I should reassses my the whole setup

[NIquist]

You need to rethink this. Remember the Seebeck effect occurs at the junction of two dissimilar metals, but they are still physically connected.  If you try to measure an external voltage with the tip of a thermocouple, it's still connected to your DMM ground though the other wire.  That's why you are just reading the thermocouple voltage.  At least that's my guess based on your description of your setup.

Is there some reason you don't just connect two more probes to measure the voltages on their own channels?  Your 973 DAQ con multiplex a LOT of signals.  🤗

[RTSLVU]

IMHO: You should reassess the whole setup

[RTSLVU]

---

@kkofori wrote:

I am not using EZ voltage or temperature I am using 'get temperature measure' and get voltage measurements. My question is can two thermocouples be used as voltage probes when you connect them to a daq. Or I should reassses my the whole setup

---

Also the reason I asked about the "EZ" vi is. A channel on the 97x series DAQ/Switch can only be configured for one type of measurement at a time. You configure two channels as both thermocouple and voltage measurement outside your loop. I am guessing only the thermocouple config is working. 

The "EZ" vi set gets around this by configuring the channel on every single scan then clearing it after the EZ measurement was taken.

Teh MUX card has 30 independant differential inputs so you have plenty to work with

[kkofori]

truth be told I am just reading thermocouple voltage and not measured voltage I was advised to do it this way. As an extra probe might bring error to the set up. My professor said it should be done with just the thermocouple probes and I am still unsure how thermocouple probe can also measure voltage and not just read generated thermocouple voltage

[kkofori]

really but when my code is running I can see it quickly flicker from celsius  to mV on the daq screen per each channel. Is this evidence that both is working or not?