• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 447 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.

MSPM0L1105: internal temperature sensor - trim value

Bernhard_87
Prodigy 160 points
Part Number: MSPM0L1105
Other Parts Discussed in Thread: MSP430FR5994

Hello all,

is there a issue with the trim value (TEMP_SENSE0.DATA) of the MSPM0L1105 known?

I am using chips from the production sample and the value of TEMP_SENSE0.DATA seems to be wrong:

  • When I do the calculation with the chip trim value (TEMP_SENSE0.DATA = 392) and the measured 12-bit value 1058 (at 2,5V reference value), then I calculate the temperature result  -268,3°C.
  • If I use the example trim value from the reference manual (SLAU847C – OCTOBER 2022 – REVISED MAY 2023 => TEMP_SENSE0.DATA = 1857), then I get 18°C.

I assume the value at the chip (e.g. TEMP_SENSE0.DATA = 392) can not be valid. The result with the example trim value matches the current room temperature better.

As a hint, I got the information that TI had a similar issue before. The MSP430FR5994 errata document SLAZ681O contains a ADC67 with an invalid temperature sensor calibration.

Regards

  • 0
    TI__Guru 70506 points

    Hi Bernhard_87

    To the best of my knowledge, there hasn't been any issues reported with temperature calibration constants.  I did find in the TRM that the factory constants are derived using 1.4vref.  I noticed in your description that you are using 2.5vref?

    Bernhard_87 said:
    measured 12-bit value 1058 (at 2,5V reference value),

  • 0 in reply to Dennis Lehman
    Prodigy 160 points

    Yes, for the measurement of the ADC channels (including channel 11 for the temperature sensor) the 2.5Vref is used. I use the formulas from the reference manual. The parameters Factory time temperature (30°C) and Temperature coefficient (-1,75mV/°C) are from the datasheet.

    The calculating of the Vtrim value is done with the 1.4Vref value. => Vtrim = 1.4V/4096*(TEMP_SENSE0.DATA-0.5)

    TEMP_SENSE0.DATA = 392 => Vtrim = 0.133V

    Vsample = (2.5Vref/4096)*(1058-0.5) = 0.645V

    Tsample = (1/(-0,00175 V/°C)) * (0.645V - 0.133V) + 30°C = -263°C

    Hint: The example from the reference manual use the value:

    TEMP_SENSE0.DATA = 1857 (ADC result code based on 12-bit mode and a 1.4-V reference)

    If I use this value (1857) for the calculation, then I get 23°C.

    This value is in the expected range, but not the previous value, which use the trim value from the chip.

  • 0 in reply to Bernhard_87
    TI__Guru 70506 points

    Hi Bernhard,

    I'm checking with our Systems team on this.

  • +1 in reply to Dennis Lehman
    TI__Guru 70506 points

    Hi Bernhard,

    It is bit 24 of the device id at location 0x41C40008.  

    There is an errata in the works and the work around will be to perform a single point calibration on those units where bit 24 = 0.

  • 0 in reply to Dennis Lehman
    Prodigy 160 points

    The bit is cleared at our controllers. For now, knowing that there is an errata is enough.

  • 0 in reply to Bernhard_87
    TI__Guru 70506 points

    Hi Bernhard,

    Ok.  You can see from my picture above mine is as well.  If that answers your question, I'll go ahead and mark this post resolved.