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.

TMS320C6713B: Reading errorenous parameters from EMIF flash at the start up (TMS320C6713B)

Part Number: TMS320C6713B

Hello,

 

TMS32C6713BGDPA200 DSP is used as DSP in some of the cards. This DSP stores boot code and some of the system parameters on EMIF flash (INTEL TE28F320C3TD90). Some of erroreneous DSP cards, at the start up (boot), some of the system parameters is read false and after some time later (I guess 730ms) system parameters starts to be read true. I checked the writings on the DSP. All of the errorenous cards have the same DSP symbolizations which is C20-9AC04HW . I am not sure the problem's source is C20-9AC04HW but it is suspicious. So, Is C20-9AC04HW marked DSP different than others when we consider data storage, addressing, boot or etc ? Do you have any experiance ?

Thank you

  • Hello Mustafa,

    The C6713B has limited design support. Let me forward your inquiry to our HW team and see if they can answer your questions.

    Regards,

    Jianzhong

  • OK, you can forward this inquiry to HW team.

    Thank you

  • Mustafa

    As mentioned, this device has limited support, but I will see if we can provide some suggestions. 

    Can you provide the details of the symbolization marked on the  parts that can read the flash?   A picture of both devices may be helpful.

    --Paul 

  • Hello,

    Thank you for your response.

    I disjointed the erroreneous DSP(C20-9AC04HW) with (C20-0BP1608) DSP. With the new DSP (C20-0BP1608), the card worked. (parameters, read correctly from the EMIF flash at once)

    I share with you erroreneous DSP, working DSP and EMIF Flash's photos.

    Thank you.

  • Mustafa

    Thank you for providing the pictures. 

    The C20-9AC04HW versus C20-0BP1608 is manufacturing information, with "C20" identifying the silicon revision.   This information will be different for different production lots. Despite these device being from different lots, they will have undergone the same testing to ensure they meet the datasheet specifications. 

    All silicon devices will have some variation and it is possible that your interface timing is marginal. I suggest you re-check the interface timing to ensure that it meets the datasheet specifications. 

    If possible, it would be good to instrument the interface to help determine the reason for the bad reads.

    Has an processor A-B swap be done to determine where the fault lies?  It is possible that the board/flash is cause of the issue.

    --Paul 

     

  • Hello Paul,

    Thank you for your message.

    I replaced flash with ST M28W320FCT70N6E which is a replacement for Intel TE28F320C3TD90.  Same problem still continues. So i think, the problem is not caused by the flash.

    I wrote a new code to DSP, which only checks (i removed other codes) the flash pattern at the start-up.(I wrote to flash a pattern before the test.) Code verifies the pattern succesfully. So, The source of problem may depend on code.(Code which has lots of functions). Or Code+DSP LOTC20-9AC04HW at the same time.

    Thank you.

  • OK, sounds like you may identified the source of this issue.  I'll close this thread for now. 

    --Paul