Board unstable??

Discussion in 'General Electronics Chat' started by cameo, Jul 16, 2007.

  1. cameo

    Thread Starter New Member

    Jul 9, 2007
    8
    0
    Hi guys,

    Any embedded-hardware experts out there to help me ??

    I hav a board built for a personal media player application.
    it consists:
    A Mips Processor
    NOR Flash
    2 Samsung 64Mb RAM
    RS232
    Audio Decoder etc.

    When i turn the board on,it starts playing the videos on an interfaced LCD Display.But it plays for some time and hangs.I need to again restart the system.
    The same board with the same design was built last year and is in production now.But now when we assembled it again these probs occurs.So i am sure that the software is working,the circuit design is ok and the board designing is also through.Apart from these what could be the problem.

    let me know,if want me to explain more.

    Cheeers!
     
  2. nanovate

    Distinguished Member

    May 7, 2007
    665
    1
    I'd look at thermal issues, and verify the BOM.

    I have a few questions:
    Has the Bill of Materials changed?
    Are any comonents getting hot?
    Can you take an old working board and compare supply currents and node voltages with one of the new boards?
    Does it hang at the same spot every time?
    Could you post the schematic?
     
  3. cameo

    Thread Starter New Member

    Jul 9, 2007
    8
    0
    Thermal issues..can u pls explain?

    BOM is not changed.I have checked it thoroughly.
    The same components are assembled.
    No components are getting heated.
    I first tried to get an old board to cross check,but not available as of now.
    Some videos hang at a particular spot.
    Most of the other videos hangs randomly.
    i dont mind posting the scematics.I suspecp the hardware.
    I'll do it if really you need it.

    one difference between the old board and the new stuff is,in the old board the RAM was a leaded BGA chip.Now i have un-leaded BGA chip.That doesnt make any problem i guess.
    Then last time the components were machine assembled.Now Except the BGA chips,rest everything are assembled manually here.Will poor assembling create such troubles.??
    I guess i am giving lot of confusions to you.But i am confused like hell.
    I think the problem is really silly and i dont know when i'll laugh at it.


    thnks
     
  4. nanovate

    Distinguished Member

    May 7, 2007
    665
    1
    What is the part number for this? Usually BGA is a ball grid array that doesn't have leads -- it has solder balls and a different pinout.

    Is the PCB exactly the same physically -- same design files or was it updated? If you are moving highspeed signals on the PCB then any changes can affect the performance.

    Assembly can cause sneaky problems-- it is not always the assembler's fault. One time I had two resistors side by side and accidently reversed their silkscreen designators on the PCB so R10's label was next to R11 and vice versa. The board was sent out for hand assembling and when it came back it took me a couple of hours to figure out why one of the voltages was wrong.
     
  5. John Luciani

    Active Member

    Apr 3, 2007
    477
    0
    Do you mean the new BGA is lead-free and the old one wasn't?

    Has the assembly house changed their soldering process? Are they capable of performing a lead-free process? Has the PCB vendor changed there finishing process?

    (* jcl *)
     
  6. cameo

    Thread Starter New Member

    Jul 9, 2007
    8
    0
    oh sorry.."lead-free" not un-leaded.
    I got even confused with these things.
    yea.The old one was leaded, and now its lead-free.
    Will that be a problem...? No..i guess.that sounds odd.
    For your other questions i have to check up with the pcb vendor.But everything is done here in India.The previous one was made in China.So there might be a difference in the assembling and manufacturing.
    But the design ,the board,the components(not the same,but the values) are the same.
    I do not doubt on the component values or the board design or the misplacing of the components.

    problem in brief = {Processor-Ram-Other components-LCD display-The video hangs} ????
     
  7. John Luciani

    Active Member

    Apr 3, 2007
    477
    0
    Especially if the required process was not communicated properly.

    A couple of other notes ---

    * Lead-free solders require higher soldering temperatures.

    * Board materials and plating may have to change. Do
    a web search for information on the "tin whisker" problem.

    * The higher solder temperature profiles can exacerbate problems with
    moisture. Some parts manufacturers already require a "bake out" for
    certain parts.

    (* jcl *)
     
  8. cameo

    Thread Starter New Member

    Jul 9, 2007
    8
    0
    thank you.
    I just started reading about tin whisker.That sounds interesting.
    Have to make a detailed study on the assembling techniques.

    By the way,Do you think this "tin whisker" problem will create such issues in my board.Like the video hanging,board unstability etc..???
    what do you say?
     
  9. nanovate

    Distinguished Member

    May 7, 2007
    665
    1
    Tin whiskering can cause a lot of problems. But if you recently assembled these boards I'd be surprised to find that there is whiskering.
     
  10. John Luciani

    Active Member

    Apr 3, 2007
    477
    0
    I don't know. I would have thought that tin-whisker problem was a longer term
    problem but maybe there are environmental issues that accelerate it.
    I would think that the moisture issue and the pre-bake of components
    would more likely show up as a short term problem.

    It would be prudent to review the process and process controls that
    are in place at your vendor. I would ask for the process documentation and
    the type of measurements that they use to ensure that their process is in
    control. I would not mention that you are having a quality problem until *after*
    you have received the documentation ;-)

    (* jcl *)
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Are you using caps next to your Vcc pins? Those can vary widely in value from batch to batch. You might be running "on the edge", and any slight glitch might be enough to cause a crash.

    It's been a few years, but I recall reading about high density RAM being sensitive to alpha particle radiation (basically, rays from space). The individual memory bit storage locations were so small, that an alpha particle could actually toggle the state of the RAM. Lead was used inside the chip carrier to shield the RAM against these alpha particles - it only took a very thin layer of Pb to fix it. If you have any sheet SN10 or SN63 solder around, you could try using some of that.

    And yes, whiskering is more of a long-term problem. The crystalline "whisker" of tin would grow until the power was removed. This caused the failure of an orbiting satellite built by Hughes Space & Communications a couple of decades ago, when a whisker shorted a critical system.
     
  12. nanovate

    Distinguished Member

    May 7, 2007
    665
    1
    Lead solder also can emit alpha particles ;) But it usually when using high lead content solder

    You are at higher risk for SEU at higher altitude. It is fairly random and could cause intermittent failures.
     
Loading...