• State TI Thinks Resolved
  • Locked Locked
  • Replies 16 replies
  • Answers 2 answers
  • Subscribers 32 subscribers
  • Views 438 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TMP102: read error data

Tommy Tzeng
Genius 4805 points
Part Number: TMP102

Hi Sir,

may we know picture 2 waveform have problem?

picture 1:

is the normal driving timing of power on device 112.78mS (bule is VIN, green is clock, yellow is data)

picture 2:

the 2 picture is that data initial read by the abnormal sensor(tmp102) at P1 87.9u/s is 74, and the abnormal temperature is 116℃ (green is data, yellow is clock)

  • 0
    TI__Expert 8970 points

    Hi Tommy,

    Thank you posting and for the included oscilloscope images.

    In the first image you attached it appears that Vin exceeds the absolute maximum rated supply voltage. Can you confirm that Vin is within the recommended operating condition of 1.4–3.6 V? This will then help us debug the read error more accurately.

    Best regards,

    Nicole

  • 0 in reply to Nicole Khoury
    Genius 4805 points

    Hi Nicole,

    thanks for you reply,

    Figure 1 VIN 5v is the SATA power, when connect to the thermal sensor, there is also 5v buck that convert to 3.3v for TMP102(meet SPEC power supply)

  • 0 in reply to Tommy Tzeng
    TI__Expert 8970 points

    Hi Tommy,

    Thank you for the clarification. 

    Tommy Tzeng said:
    the 2 picture is that data initial read by the abnormal sensor(tmp102)

    If it is only the initial read that is abnormal it's possible that you are reading the temperature register before the next measurement has completed. Setting the temperature register to Extended mode (EM=1 in the Configuration register) and then reading too soon could cause a read of the previous default 12 bit measurement. 

    Best regards,

    Nicole

  • 0 in reply to Nicole Khoury
    Genius 4805 points

    Hi Nicole,

    got it, if the reading time is too fast in extended mode,

    at this waveform, how to make sure we setting EM1 now?

    how to set EM1 mode( adjust from hardware or software?) could provide relevant suggestion for SOP? maybe the rise/fall/hold time?how to adjust?

    thanks for your help and looking forward reply.

  • 0 in reply to Tommy Tzeng
    TI__Expert 8970 points

    Hi Tommy,

    Extended mode is set with bit 4 in the second byte of the Configuration register. Please see section 7.5.3 of the datasheet for the Configuration register format. To set the Temperature register to Extended mode, set the Configuration register EM bit = 1. For more information on Normal and Extended mode, refer to section 7.4.2 of the datasheet.

    For device timing requirements, see section 6.6 of the device datasheet:

    Best regards,

    Nicole

  • 0 in reply to Nicole Khoury
    Genius 4805 points

    Hi Nicole,

    thanks for your great help.

    customer want to know the following:

    1. is there any suggestion this situation will not happen again after a few millisecons(ms)?

    2. when EM=1, amy we know the relationship between extended Mode and FW timing sequence on milliseconds(ms)?

    is this related the internal ADC conversion time?

  • 0 in reply to Tommy Tzeng
    Prodigy 50 points

    Hi Nicole,

    Good day. Two more questions as below:

    1. Is the Thermal sensor reported the waveform on device 112.78ms normal or abnormal?

    2. If it’s abnormal, TI estimates the risk high or low? Any DPPM record can be referenced?

  • 0 in reply to Frank Fu
    TI__Expert 8970 points

    Hi Tommy and Frank,

    Please see my comments below.

    Tommy Tzeng said:
    1. is there any suggestion this situation will not happen again after a few millisecons(ms)?

    My understanding is that this issue is being seen on the first read operation. If all the timing requirements are being met, I would not expect there to be an issue.

    Tommy Tzeng said:

    2. when EM=1, amy we know the relationship between extended Mode and FW timing sequence on milliseconds(ms)?

    is this related the internal ADC conversion time?

    The timing requirements seen in section 6.6 are the same for both Extended mode and Normal mode. Additionally, the typical and maximum conversion times for the device can be seen in section 6.5 of the datasheet.

    Frank Fu said:

    1. Is the Thermal sensor reported the waveform on device 112.78ms normal or abnormal?

    2. If it’s abnormal, TI estimates the risk high or low? Any DPPM record can be referenced?

    I will need to double check the device startup behavior in the lab before responding, and I should have more insight on this tomorrow. Additionally, in the first waveform picture it seems that Vin does not start at 0V. Is there a reason why there may be this initial voltage on the bus (see below)?

    Best regards,

    Nicole

  • 0 in reply to Nicole Khoury
    TI__Expert 8970 points

    Hi Frank,

    Thank you for your patience.

    How many devices has this been observed on? Is this issue only being seen in the first read operation, or is it being seen in subsequent reads?  Additionally, if possible please wait 35 ms after start up and before initiating communication for us to further debug this issue.

    Best regards,

    Nicole

  • 0 in reply to Nicole Khoury
    Prodigy 50 points

    Hi Nicole,

    Now we found abnormal Qty 180/1900pcs (9%)

    This issue has being seen in the first read operation after about 112.78ms from power turn on, it's strange because large then 35ms.

    In the first waveform picture it seems that Vin does not start at 0V. Is there a reason why there may be this initial voltage on the bus? It's right of 0.5V start with Vin, because SSD around the capacitor, for thermal sensor it's OK from 0V.

  • 0 in reply to Frank Fu
    Prodigy 50 points

    Now RD is trying to set EM=1 mode, if it work, it still needs to PASS serious verification which takes some weeks.
    For a short term solution, could TI provides the production of default EM=1 to us, or supplies the IC after "trim"?

  • 0 in reply to Frank Fu
    TI__Expert 8970 points

    Hi Frank,

    Can you confirm device behavior when waiting 35 ms from power on to begin communication, instead of waiting 112.78ms? Additionally, we don't have production of devices that default to Extended mode.

    Best regards,

    Nicole

  • 0 in reply to Nicole Khoury
    Prodigy 50 points

    OK, we are trying two ways as below now:

    1. EM=1

    2. Waiting 35 ms from power on to begin communication

    Last week we sent 2 OK & 2 NG Samples to TI, do you get these samples? Please help to update the analysis schedule.

  • 0 in reply to Frank Fu
    TI__Expert 8970 points

    Hi Frank,

    I have sent you a message regarding this.

    Best regards,

    Nicole

  • 0 in reply to Nicole Khoury
    Prodigy 50 points

    Sorry, but I don't see the message. Can you send me message again?

    Or you can send to my e-mail: frank_fu@phison.com

    Thank you.

  • 0 in reply to Frank Fu
    TI__Expert 8970 points

    Hi Frank,

    I'm closing this thread as we will continue the discussion offline.

    Best regards,

    Nicole