• State
  • Locked Locked
  • Replies 8 replies
  • Subscribers 99 subscribers
  • Views 258 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.

J784S4XEVM: J784S4XEVM Rev.E5 Random Shutoff

Semih Iseri
Prodigy 10 points

Part Number: J784S4XEVM
Other Parts Discussed in Thread: USB2ANY

Hello,

I am experiencing a fault where the EVM shuts off semi-randomly after a while and it works just fine until it happens. What I also noticed is that the MCUIO_3V3 and VSYS_IO_3V3 LEDs flicker a few times just before shutoff.

The shutoff happens after ~15-45 minutes after a cold bootup (after sufficient "cool-off" time), but happens either immediately (a few seconds) or in a few minutes if I cycle power after it shuts off once.

This all sounds like something is overheating but I didn't spot anything unusually hot with a thermal camera. What we also tried is to replace the SD card from a working board to rule out anything software related and cross checked that the switch positions are the same.

Here is a video of the LED behavior (reduced to last 10 seconds, and you can also see the fan spinning down);



I can re-upload the video in slow-mo if the flicker behavior isn't obvious but they flicker very quickly and turn off.

What do you suggest for diving deeper on the issue? I do have an oscilloscope and I can probe the board if it helps.

  • 0
    TI__Guru** 124715 points

    Hi Semih,

    Please expect a delay in response due to a TI Holiday today. 

    regards

    Suman

  • 0
    TI__Guru 56965 points

    Once a shutdown occurs, can you confirm which LEDs are still on and which are shut off (on previously)?

    What software are you running?  Does the shutdown occur when simply booting/setting at Linux prompt or only when custom software is running?

    Is this on single EVM or behave same on multiple EVMs?

  • 0 in reply to Robert Eschler
    Prodigy 10 points

    Hi Robert, 

    Thank you for the response. Here are the answers to your questions;

    • The situation with the LEDs is as follows
      • Always stays on: PWR_LED (LD4), LD10, VSYS_3V3 (LD7)
      • Starts on, flickers and turns off: MCUIO_3V3 (LD8) and VSYS_IO_3V3 (LD9)
    • Software-wise, it was running QNX 8.0 at the time of the video. It just booted up and was sitting idle. But I confirm that it shuts off with a known working SD card from a different EVM, or even without an SD card inserted.
    • This situation happens only on this EVM. We replaced the power source with a known working one to rule it out too.

  • 0 in reply to Semih Iseri
    TI__Guru 56965 points

    Yes - it sounds like the PMIC is detecting an error (possibly over-current or thermal issue) and is shutting down.  The blinking LEDs could be the PMIC re-trying to start the processor, but error/fault is being encountered.  Eventually the PMIC reaches its max number of re-tries, and enters safe-state (all outputs are OFF).  You mentioned this is the only EVM that has this behavior - what is the history of this EVM?  Has it worked for a period of time and recently started failing?

  • 0 in reply to Robert Eschler
    Prodigy 10 points

    This EVM has been running for a few months (sporadically, not always-on) but we didn't observe such behaviour earlier. It only started doing it last week.

    Regarding the characteristics of the error, I hypothesise that it has something to do with temperature raising the overall leakage currents and that triggering the overcurrent protection in PMIC eventually*. But what would you suggest for ruling out any individual component failure? I didn't see any component particularly hot based on thermal camera imagery but would it be possible to see the current consumption of the individual blocks? Maybe I can try to figure out the offending component and apply active cooling there.

  • 0 in reply to Semih Iseri
    TI__Guru 56965 points

    Yes - leakage will increase as temperature rises.  However the PMIC should be able to support worst-case temp/leakage without error.  Do you have other EVMs that are NOT experiencing this issue (is this isolated to a single EVM)?

    It could be something is failing (drawing excess current), but it doesn't trip the PMIC threshold until more current from processor.  Without reading PMIC registers and seeing specifically which errors are captured it difficult to know.

  • 0 in reply to Robert Eschler
    Prodigy 10 points

    Do you have any recommendation on how to read the PMIC? I understand that it shuts down the power and sets a fault register but it also shuts down the power to the entire J784S4 so no MCU is left to read it.

  • 0 in reply to Semih Iseri
    TI__Guru 56965 points

    Do you have access to a USB2ANY or Aardvark - or some other I2C reader?  There is a header on the EVM that can be enabled to read PMIC registers when the output supplies (processor) is shut off.  See page 79 of the EVM schematic.  Header is J37, have to toggle dip switch SW16 (position 3, net name EVM_PDN_TEST)