• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 45 subscribers
  • Views 1826 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1118: cold-junction compensation for thermocouple test

Hollis Liu
Intellectual 2595 points
Part Number: ADS1118

Hi,

This is Hollis. My customer wants to know why do the cold-junction compensation as below?

Why not convert the Vtc to Ttc directly, then get Tactual by adding Ttc and Tcj? Is there any difference in theory with the below? Thanks.

  • 0
    TI__Mastermind 23275 points
    Hi Hollis,

    Bryan and I discussed about your question already last week.

    We tried to come up with a good example to show you why adding Ttc and Tcj will not yield a 100% accurate result.
    The reason why you have to go the route of converting Tcj to Vcj first is because of the nonlinearity of the thermocouple. If the thermocouple would have a linear response then you could simply add Ttc and Tcj. But this is not the case.

    Let's assume you measure the following TC voltage and CJ temperature:
    V(TC_measured) = 35.313mV
    T(CJ_measured) = 100°C

    Now if we use the method you are proposing this would yield the following result:
    T(TC_lookup) = 850°C (converting 35.313mV to temperature using a Type-K thermocouple lookup table)
    T(calculated) = T(TC_lookup) + T(CJ_measured) = 850°C + 100°C = 950°C

    When using the correct (or more accurate) method this would yield the following result:
    V(CJ_lookup) = 4.096mV (converting 100°C to a thermoelectric voltage using a Type-K thermocouple lookup table)
    V(TC_calculated) = V(TC_measured) + V(CJ_lookup) = 35.313mV + 4.096mV = 39.409mV
    T(TC_calculated_lookup) = 952.x°C (converting 39.409mV to temperature using a Type-K thermocouple lookup table)

    As you can see both methods yield a slightly different result.

    Regards,
  • 0 in reply to Joachim Wuerker
    TI__Intellectual 2595 points

    Joachim,

    thanks for your reply.

    i see there will be some difference between 2 method from your good example. And my question is why the method in datasheet is the right way? is there any standardization organization having some standard process for this?

    Thanks.

  • 0 in reply to Hollis Liu
    TI__Mastermind 23275 points
    Hi Hollis,

    I had an application note from Omega in the past which described the theory behind it in a nice way. Unfortunately I cannot find it anymore.
    If you do a little research on the web you will find according explanations.

    As this is all based on physics, there is no organization that needs to implement a process for it.

    Regards,
  • 0 in reply to Joachim Wuerker
    TI__Mastermind 23275 points

    Please have a look at the following document from Omega (p. 23):

    www.omega.com/.../z021-032.pdf

    Regards,

  • 0 in reply to Joachim Wuerker
    TI__Guru** 113765 points

    Hi Hollis,

    I would like to emphasize what Joachim said at the very beginning.  Temperature is linear, but the thermocouple is not linear throughout the same temperature region.  Adding two temperatures is not the same as adding the voltages and then converting voltage to temperature.

    Along with the conversion calculation, the cold-junction area must be accurately measured and isolated from drift sources.  As an example, if the ADS1118 is far from the junction of where the TC enters the board, then you may not have an accurate cold-junction measurement.  Another example is when air currents flow unevenly across the cold-junction.  So board layout is very important for accuracy of the cold-junction measurement.

    Best regards,

    Bob B

  • 0 in reply to Joachim Wuerker
    TI__Intellectual 2595 points
    got it. thanks Joachim