• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 150 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.

UCD90160A: [Consultation] UCD90160 black box function check (UCD90160 vs UCD90160A)

Paul Lei
Prodigy 160 points
Part Number: UCD90160A
Other Parts Discussed in Thread: UCD90160

Hi TI service team,

I have one thing about black box (0xeah/0xebh/0xech) want to consult with you.

I've found that black box (18 fault log entries) is really easy to get full in it after power cycle sequencer UCD90160(A).

It's really hard for user to debug with such event history...

Do you have any application suggestions to avoid this case ? or should we clear black box logs first when power on in the beginning ?

Thanks a lot.

Best Regards,

Paul Lei

  • 0
    Prodigy 160 points

    Hi Yihe,

    I have one another question about black box logs in 0xECh, it's because that we cannot differentiate order in 0xEAh.

    So if we need to check fault logs through 0xECh, how can I interpret to match with GUI results ?

    Thanks.

    Best Regards,

    Paul Lei

  • 0 in reply to Paul Lei
    TI__Expert 6530 points

    Hi, Paul

    Regarding faults logs being full: you can choose to log only faults that are relevant, and choose fault-logging scheme as described in below screenshot.

    Regarding how to interpret logged-fault-details (EBh and ECh), please refer to UCD90XX PMBus Commands References of how bit fields are formatted.

    Regards

    Anne Ngo

    Texas Instruments

  • 0 in reply to Anne Ngo77
    Prodigy 160 points

    Hi Anne,

    I know that setup in UCD90160(A), but all GPIs or MONx pins are useful in my design.

    However, if I activate all of them in SEQ configuration, it will false trigger when power off UCD90160(A), and black box will record many trash logs after that...

    Or you suggest that we can use FIFO to reserve some of the real fault logs in boxes ?

    For ECh/EBh, I have referred this document already...but I cannot interpret it to match with results from GUI.

    How can I use ECh/EBh to interpret logged faults result from EAh ?

    You can see my results as below show:

    Thanks.

    Best Regards,

    Paul Lei

  • 0 in reply to Paul Lei
    TI__Expert 6530 points

    Hi, Paul,

    You cannot use EBh/ECh to interpret EAh. EAh reports a history of all faults that have ever been reported and logged into
    nonvolatile memory until clear, whereas ECh reports what are in faults-log. For example: If FIFO mode is disable and faults-log is full, EAh reports the fault, but that fault is not logged; or if FIFO mode is enabled, EAh reports the fault, but that fault is no longer in the faults-log since it is replaced.

    Unfortunately, we have limited entries in faults-log => choose what to log, and logging scheme to best suit your application.

    Regards

    Anne Ngo

    Texas Instruments

  • 0 in reply to Anne Ngo77
    Prodigy 160 points

    Hi Anne,

    However, I need method to interpret fault logs sequence in black box, if it can be read with 0xEAh for all history, I cannot differentiate them with every each fault logs (for example: power off sequencer or trigger one rail and trigger fault response to turn off other rails).

    I knew FIFO function if we activated it in black box (reserve half of history, 9 events or all records will be replaced, 18 events).

    However, I think we need use 0xEBh and 0xECh to further check event history in sequence rather than check 0xEAh for overall logs.

    I don't think we should use 0xEAh to check all event history in period, it's really for user to debug in system.

    Can you help to check with my actual fault logged detail with 0xEBh and 0xECh registers (how can I check with fault value, byte9 and byte10 with below screenshot to match with GUI results) which is read by i2c command as below ?

    and how to interpret to match with GUI results (LOGGED FAULT DETAIL) at that moment ?

    Is there any interpreter tool file to import fault logs for explanation (0xEAh/0xEBh/0xECh) ?

    Thanks.

    Best Regards,

    Paul Lei

  • +1 in reply to Paul Lei
    TI__Expert 6530 points

    Hi, Paul

    Unfortunately there is no tool available to interpret fault log but the document.

    Below are example of how to interpret fault-log entries for Fault #1 and Fault #2. For Voltage Fault value (bytes [9:8]), the value is in Linear 16 format, and the exponent is from VOUT_MODE (20h) register.

    Regards

    Anne Ngo

    Texas Instruments

  • 0 in reply to Anne Ngo77
    Prodigy 160 points

    Hi Anne,

    Got it, I can understand what's your expressing in previous captions.

    I think that it's okay for user to further check fault logs in sequence with 0xEBh and 0xECh, not just using 0xEAh to check for all fault logs.

    That's would be helpful to my system if we need this feature for history examination.

    Thanks for your explanation!!!

    Best Regards,

    Paul Lei