I am try to get a thermocouple by AD7124-4 application again.


I met AD7124 IC two years ago.

I knew great works well. I made 10 boards on same circuit as bellow.

when I turn on the board and check a coldjunction temperature(RT1), the 5 borads temperature is different.

the 5 boards are about 24 degree and the other boards are 30 degree. the part is used same.

what do i do wrong? 

my code is like this:

1) reset data send ( 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff)

2) channel-0 setting ( 0x09, 0x10, 0xA6)      // AINP=AIN5,     AINM= AIN6

3) config-0 setting ( 0x19, 0x01, 0xEF)      // unipolar, ref_bufp

4) filter-0 setting ( 0x21, 0xE6, 0x07, 0xFF)

5) IOUT setting (0x03, 0x00, 0x12, 0x0B)

6) when i read, (0x01, 0x01, 0x84); // ADC controll setting

                         (0x42, 0x00, 0x00, 0x00, 0x00) // read

 i send number 6) messages on period.

please, check my setting codes and help me.

thanks for reading.

  • Hi Brandon,

    I have more question.

    Is there a cold junction compensation for thermocouple on this block diagram as follows:

    when I heat the AD7124-4 from 20 celcius to 40 celcius, the themocouple temperature is increasingly higher. 

    I found this chapter in AD7124-4 datasheet.

    1st question:

    If I compensate the cold junction for thermocouple, do i get thermocouple data through AIN1 and AIN3, cod junction data through AIN5 and AIN6?

    2nd question:

    If I get compensation temperature,

    is the equeation right as follows:

    compensation temperature = thermocouple temperature + cold junction temperature

    thank you for reading.

  • Hi Hahmin,

    Sorry for the delayed response, I was out on vacation.

    For question 1, the thermocouple data would be read back through channels AIN1 and AIN0 in the drawing shown above.  The cold junction temperature would be back from AIN4 and AIN5 (as shown above)

    For question 2, the formula for reading back the temperature compensated data is much more complex than that.  Thermocouples are very non-linear, so you must use polynomial equations to get an answer.  We have a nice video that explains this (Thermocouple 101: Compensating for Nonlinearity - YouTube  )  and you'll also want to visit the National Institute of Standards website for the tables and formulas.