• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 45 subscribers
  • Views 377 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.

ADS124S08: Split power rails & Internal temperature sensor

Paul White
Prodigy 20 points
Part Number: ADS124S08

Posting this for a coworker who is having all kinds of IT issues trying to get an account going here ...

We are using ADS124S08IPBSR A/D converters prototyped on a PCB to measure temperatures from either PT1000 RTDs or Type-J Thermocouples. Two ADCs share a common SPI port. The Start/~Sync line is kept low. The external reference input is used as a low-side for the RTDs and the internal reference used for the T/Cs. The T/C are of the grounded sheath type – thus the need for split power rails. The good news is that we’ve been mostly successful in reading both types of sensors & the values returned are making sense. But, there are a few issues that could use clarification:

 

  1. Split power supplies of nominal + & - 2.5v are used. Not knowing what do with AINCOM here, it’s unused & tied to ground with a 33Ω resistor in case I need to tie it to a different level or short or leave open.
  3. The internal temperature sensor has been configured to be ON with a PGA gain of 4. The value being returned is around 1B8800H with the PCB nearby measured at 77.8 degF. I have no idea if this value is valid or how to scale things - especially with REFCOM at -2.5v and REFOUT being at around ground level. I’ve tried to use the ADS1x4S0x Design Calculator, but I don’t even know if it’s a valid tool for the internal temp. I’ll need to know the PCB temp in order to compensate for the thermocouple applications.

 

 

 

  • 0
    TI__Guru** 113765 points

    Hi Paul,

    You can pass the E2E thread link to your coworker and they can respond to this thread once they are able to get to E2E.

    Paul White said:
    • Split power supplies of nominal + & - 2.5v are used. Not knowing what do with AINCOM here, it’s unused & tied to ground with a 33Ω resistor in case I need to tie it to a different level or short or leave open.

    The AINCOM pin is the same as any other analog input pin.  In hind site we should have called this pin AIN12 to lessen confusion.

    Paul White said:
    The internal temperature sensor has been configured to be ON with a PGA gain of 4. The value being returned is around 1B8800H with the PCB nearby measured at 77.8 degF. I have no idea if this value is valid or how to scale things - especially with REFCOM at -2.5v and REFOUT being at around ground level. I’ve tried to use the ADS1x4S0x Design Calculator, but I don’t even know if it’s a valid tool for the internal temp. I’ll need to know the PCB temp in order to compensate for the thermocouple applications.

    The internal temperature sensor is quite precise, but not necessarily accurate unless you adjust by calibration.  Also, it measures the internal die temperature which may or may not be that accurate relative to the cold-junction of your thermocouple.

    In most cases I've seen, other sensor types are used for the cold-junction measurement which in turn is connected to measurement inputs on the ADS124S08.  This could be a chip RTD, or a thermistor in the cold-junction.

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Prodigy 20 points

    My coworker still doesn't have his IT issues resolved, so he has asked for more clarification ...

    1. Given split-supply design, what should be done with the unused AINCOM pin?
      1. Open circuit.
      2. Tied to ground.
      3. Tied to a different voltage.
      4. Terminated in a resistance to ground of X ohms.
    2. Please give the formula for the internal temperature sensor in degrees. Our ADCs are located within less than an inch of the T/C connectors. Well within our accuracy requirements. Especially since each PCB will be calibrated at several temperature points.
      1. So, given a PGA gain of “X”, what is the die temperature in degC as a function of A/D counts returned.
  • +1 in reply to Paul White
    TI__Guru** 113765 points

    Hi Paul,

    Paul White said:
    Given split-supply design, what should be done with the unused AINCOM pin?
    1. Open circuit.
    2. Tied to ground.
    3. Tied to a different voltage.
    4. Terminated in a resistance to ground of X ohms.

    The information regarding Unused Inputs and Outputs is discussed in the datasheet in section 10.1.5.

    Paul White said:
    Please give the formula for the internal temperature sensor in degrees. Our ADCs are located within less than an inch of the T/C connectors. Well within our accuracy requirements. Especially since each PCB will be calibrated at several temperature points.
    1. So, given a PGA gain of “X”, what is the die temperature in degC as a function of A/D counts returned.

    This is also in the datasheet but I realize this takes a little more digging to figure out.  The ADS124S08 measures relative to the reference, which in this case is the internal reference.  The internal reference is a voltage reference so the outcome in codes relates proportionally to the reference voltage.

    The typical values for the internal temperature sensor are given in the Electrical Characteristics table as shown as follows:

    Figure 36 in the datasheet shows that the response is linear throughout the operational temperature of the ADC.  The discussion on the operation of the internal temperature sensor is given in section 9.3.9.1 of the datasheet.

    As mentioned, if the PGA is enabled the maximum gain that can be applied is 4 (so gains of 1, 2 and 4 are possible).  The ADC counts work in the same way as any other voltage measurement.  This is given in equation 11 in section 9.5.2.

    So for the internal temperature sensor the range for +FS is 2.5V/Gain.  If we use a gain of 4, for example, the +FS is 2.5/4 V = 625mV.  So the value of 1 count is 625mV / 2^23 = 74.5nV.

    To illustrate I will work backwards, so 129mV (25 deg C from the table) should have a count equivalent of 129mV / 74.5nV = 1731544 counts.

    Now lets calculate something in the right computational direction.  Lets say the counts returned by the ADC are 1895443.  To compute the voltage you would get the ADC_Counts * 74.5nV = 141.21mV.  This is equal to 25 deg C plus some additional value.  141.21mV - 129mV = 12.21mV.  So the temperature is 25 deg C + 12.21mV / 403uV/ deg C (which is the coefficient from the table).   So this comes out to be 25 + 30.3 deg C = 55.3 deg C.

    You can see that the values match up pretty well with Figure 36.

    Best regards,

    Bob B

  • +1 in reply to Bob Benjamin
    Prodigy 20 points

    Relaying the last message ...

    Perfect ... exactly what he was looking for!  Now just going to see if the internal temp sensor will work for what we need.

    We appreciate your help!