I think you need check after turn off power supply, how long time will the board fully powered off, no leak power remained.
Maybe in the passed 6 month, some hardware component is not working as before.
If there are remained leak power, the board is easy to boot failure.
Thank you Qiang Li.
We proceeded in the same way as you mentioned in your previous reply. It is found that the board is not completely powering OFF within 2 seconds. There is some leak power remains. But the interesting fact was that a few number of boards fail to boot while the others will work properly without any issue. So we did some X-ray analysis on BGA components (mainly on PMIC) and it found that there is some minor contact degradation for some of the BGA balls. This can happen due to several reasons over a period of time.
We replaced those components and is working fine now. But it is not desirable to replace the PMIC in every 6month. So we optimized the input power supply for minimizing the leak power and optimized the board layout. Ensured the assembly process with recommended temperature profile and board finish made as ENIG.
With these modifications we hope we can override this problem.
Since this is happening after a few months only, we need to see the result after a few months of time.
Hi Naushad,
We are facing exactly the same issue you described. Could you please let us know if the solution of board relayout helped overcome the problem?
Thanks
Karthik
Hi Karthik, if you power off the board for log enough time, then power on the board again, will you get the boot failure issue?
Thanks Li for answering the post.
Typical scenario is : a functional board @ customer site on a Soft/ Hard reboot gets locked up. But the funny thing, if we it power off and repower it say giving few seconds, then it seems to always work. Our biggest issue is to be unable to answer why suddenly unavailability of the device and ways to solve it.
It hasn't occured in the lab conditions even after running a script to reboot the board for days together continuously and it remains unclear for us if this is a software or a hardware issue till date.
Any pointer to incrementally know the root cause will be of a great help.
I think you can check the followed two things:
1. Power sequence when the issue happens.
2. Boot mode sampled from boot pin. This can be checked from Reset Control and Source Register (RCSR) with Realview ICE when boot up failure happens.