Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

K type thermocouple malfunction at 800-1000K

  1. Sep 26, 2013 #1
    Hey Everybody,

    So we have a PID heating system with K-type thermocouples running through labview. We have a chassis that has 6 DAQ cards for Tc inputs (2 daq cards), flowcontroller in/out, an output for the heating elements for the PID control, and another input for a pressure transducer.

    Anyway, the thermocouples give nice reading at most temperatures near room temperature. And the PID system seems to work, as I can control temperature pretty closely in the 300K-800K range. However, when I get to 800K+, some (not all) of the thermocouple readings go kind of nuts. They might randomly spike up to 1200K in a second, and then come down to 400K, etc. If I turn off the heating, after enough time passes and things have cooled back down, they might start behaving normally again.

    I have all the DAQ cards coms wired together and then sent together to ground. Is this more likely to be a wiring/grounding issue or is it likely some sort of programming/labview issue? I don't suspect the problem is with the Tc's, as they are pretty new. I have no background at all in circuits, so I kind of think it is most likely something I am doing wrong, but I have to admit, when someone says "It's probably a grounding issue," I don't really understand what exactly that means or how I should ground it differently.

    Sorry if this is a really basic question, but I guess this is what happens when a mathematician tries to like, actually do stuff.


  2. jcsd
  3. Sep 26, 2013 #2


    User Avatar
    Science Advisor
    Gold Member

    Does your DAQ cards have thermocouples wired directly into it or is there a buffer which is commonly called a transmitter between the thermocouples and the DAQ? I would not wire any commons together at the DAQ. What are you measuring? Obviously something that gets hot, but how is it heated? Electrically?
  4. Sep 26, 2013 #3


    User Avatar
    Science Advisor
    Gold Member

    Sounds to me like the thermocouple Nickel to Chromium weld is opening. I would silver solder the bad ones and give them a try. If you want to run a test of temperature vs t/c voltage, here's a type K temperature (ÂșC) vs. voltage (millvolts) table.

    "www.omega.com/temperature/Z/pdf/z204-206.pdf" [Broken]
    Last edited by a moderator: May 6, 2017
  5. Sep 30, 2013 #4
    Thanks for the replies. I definitely think it was bad wiring the Tc common to ground. They are
    working better now. But still, not perfectly. Now, instead of freaking out at the 800K range,
    they start behaving badly at the 1100K range, and become useless at 1200K. I would really
    like to go up to 1200K.

    To answer your question, I am using electrical heating elements from Omega. There are two
    57-Volt heaters wired in series, and then plugged directly into the wall/PID system. The PID
    system is controlled by labview through a different set of DAQ cards, but connected to the
    same chassis as the TC Daq cards.

    I'm starting to think that there may be a programming issue as well, because I noticed that in
    the labview program, when I turn off the "start PID" programming loop, the TCs go back to
    giving accurate temperatures, even at the high range.

    Here is a picture of everything behaving nicely at 1000, and then heating to 1100, and
    starting to bug out a bit:


    Here is a picture of everything working completely terribly as I approach 1200K, but
    then when I turn off the 'start PID' switch, the TCs acting a lot better as the temp comes down.


    Thanks again for all the help!

    EDIT: yikes! Sorry they are huge. If someone knows how to fix that, let me know. Thank you!
    Last edited: Sep 30, 2013
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook