• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 45 subscribers
  • Views 717 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: Low temperature strange behaviour

Bolesław Rohm
Prodigy 60 points
Part Number: ADS124S08

Dear All,

Sometime ago I have asked on this forum for an advice regarding the ADC selection - I have chosen ADS124S08. 

Bunch of information regarding the application:

  • 4 load cells with full bridge configuration,
  • Supply voltage of load cells is connected as REFP0 and REFN0,
  • STM32 is being used to communicate with ADS,

Operation:

  • Wait 1s for PCB boot, after power apply,
  • Set ADS registers as I wish,
  • Wait for 100ms,
    • Start the conversion on channel 1, wait on DRDY,
    • Read the value, reconfigure for next channel, wait on DRDY, 
    • ... continues, 

SPI is running correctly, registers are written also accordingly. Read values are correct, registers settings are also correct.

Issue:

I have run some simple temperature tests on this PCB, separating the PCB and load cells in the climate chamber, or combining them. At certain temperature, after the initialisation it seems that PGA settings are not correctly written - read values are 100 times smaller (around). After heating the area specifically and reset of power supply, ADS runs with correct settings of PGA. 

Voltages at rails and reset signals, raise properly nothing abnormal measured.

Did anyone, encountered similar issues? Any temperature dependence observed? Shall I reinitialise the ADS periodically? Is there any sensitive circuit (RESET, AVDD, etc.) which could be affected by value temperature change of the components?  

  • 0
    TI__Guru** 113765 points
    Hi user4229703,

    The ADS124S08 has been thoroughly evaluated for temperatures from -50 deg C to +125 deg C. Some additional information would be helpful, such as how cold was the actual temperature you were using? What are the configuration register settings for the ADS124S08? What is the SCLK frequency you are using? What are the AVDD, DVDD and IOVDD voltages you are using? And do you read back the register contents to verify the write register values sent to the ADS124S08?

    I've seen issue in the past with frost that may build up on the PCB that causes issues. I've also seen problems where the micro has different operating temperatures than the ADC.

    Best regards,
    Bob B
  • 0 in reply to Bob Benjamin
    Prodigy 60 points
    Hi Bob,

    Thank You for Your answer.

    Additional information:
    - actual temperature was -20 degrees Celsius,
    - SCLK frequency is 2.25 MHz,
    - IOVDD and DVDD are on 3V rail (same as microcontroller),
    - AVDD is on 5V rail,
    - REFN0 and REFP0 are sensing 4V rail,
    - registers' values
    PGA_ADDR_MASK, ADS_PGA_ENABLED | ADS_GAIN_128 // PGA enabled, 128
    DATARATE_ADDR_MASK, ADS_CONVMODE_SS | ADS_DR_2000 //Single Conv, 2000
    REF_ADDR_MASK, ADS_REFN_BYP_DISABLE | ADS_REFP_BYP_DISABLE | ADS_REFSEL_P0 // /ref P and ref N buffers
    IDACMAG_ADDR_MASK, 0x00 //IMAG
    DACMUX_ADDR_MASK, 0x00 // IDAC off
    VBIAS_ADDR_MASK,0x00 // Vbias off

    Currently the software is not reading back the register values after writing. I will add this piece of code with checking the register content against written settings.

    Frost is not an issue in this case and microcontroller is capable of handling -20 degrees Celsius.
  • 0 in reply to Bolesław Rohm
    TI__Guru** 113765 points
    Hi Boleslaw,

    Outside of the gain being incorrect at cold temperatures, have you noticed if the output data rate is actually at 2ksps or is it the default rate of 20sps?

    Best regards,
    Bob B
  • 0 in reply to Bob Benjamin
    Prodigy 60 points

    Hi Bob,

    I have edited the code, so it will write the registers and readout the content in the next step as long as they don't match. It helped - no issues with wrong configuration so far.

    Currently,  I am struggling with temperature drift of the IC, I might need to change the package from TQFP to VQFN for better temperature performance and probably change the reference LDO.

    Thank You for Your help,

    Kind regards,

    Boleslaw

  • 0 in reply to Bolesław Rohm
    TI__Guru** 113765 points
    Hi Boleslaw,

    It appears that you are using a ratiometric measurement where the excitation and the reference use the same voltage source. This should minimize drift error except for gain error and offset of the PGA. You can see the typical response in Figures 19 and 20 of the ADS124S08 datasheet. To compensate for offset drift you can issue periodic SFOCAL (self offset calibration). You could also enable the Global Chop which will also cancel the offset, but this will also decrease the overall throughput as two measurements are required for averaging.

    Best regards,
    Bob B
  • 0 in reply to Bob Benjamin
    Prodigy 60 points

    Hi Bob,

    Yes, correct, I am using ratiometric measurement and I have chosen ADS124S08 exactly for this purpose. 

    I have tried using Global Chop at the beginning to cancel the offset, but the data rate was insufficient. But still, I will try to compare the results with and without the chopping. 

    Regarding the SFOCAL - I am doing it at startup but later on I am not using it. How often would You recommend to run it, assuming 1 degree per 5 minutes temperature  increase (at the beginning), for 45 minutes. Once per minute? Once per 5 minutes? 

    Kind regards,

    Boleslaw

  • 0 in reply to Bolesław Rohm
    TI__Guru** 113765 points
    Hi Boleslaw,

    Telling you when or how often to issue the SFOCAL is difficult to answer. Basically this is dependent on the temperature range of the system and how often there is a large or significant change in the device temperature and the effect it has on the measurement. You could monitor the ADS124S08 temperature and if the temperature varies an x (you would need to determine how much temperature change affects the measurement) number of degrees from the last calibration you run the calibration again.

    Best regards,
    Bob B