• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 97 subscribers
  • Views 892 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.

AM1808: Can't load uboot

Ray Dennison
Prodigy 20 points

Part Number: AM1808

We are currently in production with a processor board using AM1808. We have a fairly high fall out rate (~5%) of boards where we can't load uboot. The boards come up in bootme as expected, We have 2-way serial communication with bootme. All supplies and clocks look right. But it fails to talk to flash when we try to flash uboot. We have tried replacing the flash chip but that did not help. I was wondering if anyone has experienced similar issues and what they did to fix it. Thanks.

  • 0
    TI__Guru* 85925 points
    Hi Ray
    Can you elaborate further.
    1) What boot modes are you using for primary boot (when you say boot me is it just UART?) vs uboot. Flash as is NAND flash or something else?
    2) 5% failure - what board quantities are you talking
    3) Since you have been in production , any change in hardware BOM or software , recently?
    4) Do you have waveforms on what is going on between good board vs bad board when uboot fails?
    5) You did a flash swap experiments - any experiments around swapping AM1808s?
    6) Have you also confirmed the your flash media, clock and voltages are correct?
    7) Do you have any simpler diagnostic test to confirm that flash is ok, and not failing something in linux/uboot etc? Hardware vs software issue?


    Regards
    Mukul
  • 0 in reply to Mukul Bhatnagar
    Prodigy 20 points

    Hi Mukul,

    Thanks for the quick response. You will have to forgive some of my ignorance. This board was designed by a consulting firm that we no longer do business with. We are pretty much just running a procedure provided by the subcontractor. You may need to help me out as far as some of the details. I would be happy to send you the schematic, though I hesitate to post it to a public forum.

    Mukul Bhatnagar said:
    Hi Ray

    Can you elaborate further.
    1) What boot modes are you using for primary boot (when you say boot me is it just UART?) vs uboot. Flash as is NAND flash or something else?

    we start with bootme via UART. the board uses NAND flash. Specifically

    Micron Technology MT29F4G08ABADAWP:D

    That part is end of life.. we have also had good results using a substitute from ISSI

    2) 5% failure - what board quantities are you talking

    ~5000


    3) Since you have been in production , any change in hardware BOM or software , recently?

    This has been an ongoing issue since the start of production. Over the time period there have been die revision on the AM1808, but I don't see any correlation to the failures

    4) Do you have waveforms on what is going on between good board vs bad board when uboot fails?

    No. Let me know what you need and I will try and get them for you.

    5) You did a flash swap experiments - any experiments around swapping AM1808s?

    We have not done that yet. We have no in-house BGA capability. Our experience using subcontractors to swap BGA has been poor. But we can send some out and see

    6) Have you also confirmed the your flash media, clock and voltages are correct?

    Yes. To the extent of our limited knowledge. All regulators on the board are producing the correct voltages. Crystal clocks are running at specified frequency.

    7) Do you have any simpler diagnostic test to confirm that flash is ok, and not failing something in linux/uboot etc? Hardware vs software issue?

    In manufacturing, the procedure is to use bootme to flash uboot via UART, then we use uboot to load kernel and jffs2 via Ethernet. Since we are failing at the uboot step I don't know what other diagnostics I could run. I don't know of any any way to load anything else onto the board.  I am no expert on bootme. If there are other things we can try with bootme, let me know. It seems like a hardware issue since we perform the same procedure on good boards all the time.

    Regards
    Mukul

  • 0 in reply to Ray Dennison
    TI__Guru* 85925 points

    Hi Ray

    Ok, thanks for additional details.  Minor comments on all 

    1) Ok the Micron NAND flash part you are using, is supported by the RBL and was also on the older EVMs, 

    2) Ok, so about 250 boards failing. You mentioned that this has been a recurring problem, so is 250/5000 the current batch failure or cumulative?

    3) Yes, I do not see how AM1808 silicon rev should impact this. However if the failure % has increased recently, it would be good to understand if there is something that changed recently in the hardware or software design that is making the problem worst?

    4) It would be good to see the activity on the NAND signals and chip select on a good vs bad board. 

    5) Yes understood that replacing the processor maybe non trivial. Need to see what other options you have for isolation 

    6) Is it possible to flash something simpler than uboot via UART, say a led blink to help you figure out on what stage the NAND is failing ?

    Not sure if the following utilities help

    Do you have the ability to connect to JTAG etc on the board? Any way to see if the content written in the flash is as expected , to see if the failures are while writing to the NAND or reading from the NAND?

    If the timings are  somewhere in your system, it maybe good to see if you can reduce the speed of the process or relax the setup , hold , strobe timings for NAND to see if helps?

    Regards

    Mukul