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.

MSP430FR2032: MSP430FR2032

Part Number: MSP430FR2032

Hi, 

I have few MSP430FR2032 with on chip 30C calibration at higher error. The data stored at address 1A1Ah and 1A1Bh are relatively larger than normal unit causing temperature reading becomes 23C when actual temperature should be 30C, May I know could we re calibrate it and overwrite data into 1A1Ah and 1A1Bh?

best regards

monglong

  • Hi Mong Long Ang,

    You can re calibrate the data and save it somewhere in the FRAM. But you can not overwrite the data in 1A1Ah or 1A1Bh.

    Regards,
    Ling
  • Hi Ling,

    it is too late for us to change the code. During prototyping, MSP430 initial temperature for 30C is ok for various datecode, so we directly call cal data in 1A1Ah and 1A1Bh. Only when comes to mass production, few parts do not calibrate properly and is cal data is different.

    Data in 1A1Ah and 1A1Bh is hard coded or can be programmed during production test? could you please confirm?

    best regards

    monglong

  • Hi Monglong,

    Yes, confirmed that data in 1A1Ah and 1A1Bh is hard coded.

    Regards,
    Ling
  • HI Monglong,

    Can you share some more details on this issue?
    How many devices show this increased data?

    Best regards
    Lukas
  • Monglong,

    can you please post the complete TLV data of the affected devices?
  • Hi Dietmar,

    We have 7 devices reported with much lower temperature reading which we know it is in correct. Production pass me a device to check and I made a comparison with a good device:

    ADC data for good device:
    measure at 30C 667
    Cal30 666
    Cal85 791
    Calculated temperature 30.44
    ADC data for bad device:
    measure at 30C 665
    Cal30 677
    Cal85 796
    Calculated temperature 24.45378151

    the bad device 30C calibration data is 677d, this makes actual 30C calculation to be 5C lower. I wish to find away to correct Cal30 register for that 7 devices instead of trash away, there is nothing wrong with it, probably just bad calibration.

    best regards

    monglong

  • Hi Monglong,

    thanks for providing detailed results. Indeed it's looking like not perfectly calibrated cal data.
    Is it possible to send the complete memory dump of the 7 failed parts to us. Based on this we can identify the production lots and check the production test data.

    As mentioned it is not possible after the devices left our production to change the data in the boot data or TLV section.
    The only thing I can propose is to store it somewhere else in FRAM and use this alternative cal value instead of the factor data but this was already proposed before.

    Best regards,
    Dietmar

  • Hi Dietmar,

    We now have total 13pcs with same problem and I think it is increasing for recent built unit. The device calibration data for all these units are about the same, 30 drgree C calibration data is too high. 

    May I know the 30 degree C calibration is done on wafer level or after packing? Could there be a problem with calibration process in TI factory? We start tracking datecode of the MSP430, possible datecode related problem? We would expect +/-3C error but we are getting error more than that.

    The footnote in  MSP430 datasheet comes with this: 

    (1) The temperature sensor offset can vary significantly. TI recommends a single-point calibration to minimize the offset error of the built-in
    temperature sensor.

    Question: is this necessary if there is 30C and 85C calibration were done by TI? How is a single point calibration actually works and help?

    best regards

    monglong 

  • Hi Monglong,

    Let me take over this case and check the details with the Systems team.

    I will come back to you soon.

    Lukas

**Attention** This is a public forum