• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 244 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.

BQ3050: Watchdog cycle at low temperatures

zehu chi
Intellectual 445 points
Part Number: BQ3050
Other Parts Discussed in Thread: BQ8050,

Tool/software:

Dear Ti engineer

 If there will be a deviation in the wake-up cycle of the watchdog under low temperature conditions?

 Best wishes

  • 0
    Intellectual 445 points

    I'm sorry, let me add.

    I am referring to whether this problem exists in the BQ8050 chip.

  • 0 in reply to zehu chi
    TI__Mastermind 27321 points

    Hello,
    We are not aware of any deviation in the wake-up cycle of the watchdog under low temperature conditions.

    Can you please elaborate the issue? 

  • 0 in reply to Jose Couso
    Intellectual 445 points

    Hello,Jose

    Now we have a sleep setting for 1S wakeup to be used as a sleep timing, this timing is very accurate at room temperature, but at -20℃, the timing error can be -14% to -15%. So we're guessing that's maybe the reason. What do you think

     Best wishes

  • 0 in reply to Jose Couso
    Intellectual 445 points

    Hi  Couso:

    Could you please help to answer it when it is convenient for you? Thank you！

    Best Regards,

    zehu

  • 0 in reply to zehu chi
    TI__Mastermind 27321 points

    Hello,

    Watchdog timers in such ICs are typically driven by an internal clock source, such as an RC oscillator or a low-power crystal oscillator, which resets or wakes the system if it doesn’t receive a signal within a set period (in your case, a 1-second wake-up cycle). The accuracy of this timing depends heavily on the stability of the clock source across environmental conditions, particularly temperature.

    At room temperature, you note that the 1-second wake-up timing is "very accurate," but at -20℃, you observe a timing error of -14% to -15%, meaning the wake-up cycle is shorter than expected (e.g., ~0.85–0.86 seconds instead of 1 second). A negative deviation like this suggests the clock is running faster at lower temperatures, which is unusual for most RC oscillators, as they typically slow down (resulting in longer periods) when temperature decreases due to increased resistance and capacitance. However, certain oscillator designs or compensation mechanisms could lead to this behavior.

    Do you have access to BQ8050 Data Manual? Or BQ3050 DS?
    There should be info related to temperature drift in watchdog timer. 

    Possible workaround:
    • Use an external RTC or timer to handle the 1-second wake-up function, bypassing the BQ8050’s wake timer entirely.
    • This ensures high accuracy across the temperature range and is ideal for production systems where timing precision is critical.

    If power is a concern, choose a low-power RTC for the external timer solution (e.g., <1 µA). Shortening the wake interval increases power draw, so balance this with your system’s requirements.
    Note: If your system can tolerate a 150 ms error at -20℃, you might not need a fix.
  • 0 in reply to Jose Couso
    Intellectual 445 points

    Hello,

    Thank you very much for your earnest reply. I feel very honored

    According to the user manual, it looks like we set the programming step size is 7.8125 ms, but we don't see where the error range of this 7.8125ms is explained,where can I find it? 

    Best Regards,

    zehu

  • 0 in reply to zehu chi
    TI__Mastermind 27321 points

    Hello,

    The 7.8125 ms step size is a nominal value based on a 32-kHz clock, but its actual value depends on the oscillator’s frequency, which varies with temperature. The datasheet likely assumes the user will refer to the oscillator’s accuracy specs to determine the error range.

    • Look in the "Electrical Characteristics," "Timing Characteristics," or "Internal Oscillator Characteristics" sections for the 32-kHz oscillator’s frequency tolerance and temperature drift.
    • Check for footnotes or notes in Section  that reference the oscillator’s accuracy.
    Based on your data, the 7.8125 ms step size has a -14% to -15% error at -20℃ (i.e., it becomes ~6.64–6.68 ms), which corresponds to a ~17.6% increase in the oscillator’s frequency.


  • 0 in reply to Jose Couso
    Intellectual 445 points

    Hello,Jose

    I found the 32.768kHZ Oscillator perameters in the user manual, the 7.8125 ms step size has a -1.5% to +1.5% error at -20℃.It seems that the error in our test data is much large than in the user manual.

    Is the error of 32.768kHZ Oscillator we found correct?  Why is the difference so large?

    Best Regards,

    zehu

  • 0 in reply to zehu chi
    TI__Mastermind 27321 points

    Hello,

    Your observed 14–15% timing error corresponds to a 16.27–17.64% frequency error in the 32.768 kHz oscillator, which is consistent with your test data.

    The datasheet specifies a −1.5% to +1.5% frequency error at −20C, but your observed error (16.27–17.64%) is much larger. Possible reasons include environmental factors (e.g., voltage, humidity) or test setup issues.

    I would recommend to:
    • Verify test conditions (temperature, voltage, measurement accuracy).
    • Test additional units to check for consistency.
    • Implement a workaround (e.g., software compensation or external timer) to mitigate the issue in the meantime.