• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 32 subscribers
  • Views 1390 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.

MBIST algorithms, and testing the test.....

Phil Eagleton
Intellectual 335 points


Hi All,

I am running the MBIST on the memories in the TMS470M on startup, and guess what, they pass! Does anyone know if there is an application note, or further documentation about the algorithms selected? Can you select more than 1 algorithm at a time? are some better than others? 

In a similar vein, is it possible to test that the MBIST is working? is it possible to inject a problem somewhere and show that it is detected?

thanks,

Phil

  • 0
    TI__Intellectual 1480 points

    phil,

    You can select the algorithms  by configuring bits 23-16 in the MSTGCR

    Below is the description:

    23-16 MBIST_ALGSEL Selects different Algorithm for MBIST

    00000000 No Algorithm selected.

    00000001 Checkerboard has been selected.

    00000010 March 13N with background of all 0s and all 1s has been selected.

    00000100 March 11N with background of hex 5 and As has been selected.

    00001000 March 13N with backgrounds of hex 3 and Cs, hex 0F and F0s, and 69 and 96s has been selected.

    00010000 PMOS Open Address Decode has been selected.

    00100000 No Algorithm selected.

    01000000 No Algorithm selected.

    10000000 No Algorithm selected

     

    Each bit from 16 to 23  enables its corresponding to a algorithm as described above .  my recommendation would be to run all algorithms mentioned above except for the PMOS Open Address Decode algorithm for the best tradeoff between run time and coverage but you can still run all of them bu programming 00011111

    The is no hardware measure for failure injection for MBIST based self test.

    We plan to include this in the next publcly avaliable version of the TRM .

     

    Regards,

    Hercules forum support

  • 0 in reply to Hari Pendurty
    Intellectual 335 points

    Is there any documentation showing the run time / coverage trade off for the different algorithms? I need to get "enough" coverage, but I need to know more about the implementation on chip to decide when enough is enough! 

    When is the next rev of the TRM expected?

    thanks,

    Phil

  • 0 in reply to Phil Eagleton
    TI__Expert 7650 points

    Phil,

    Sorry for the delay in responding to your post.  Assuming you are still interested in these answers, I've notified our expert to follow up.