• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 83 subscribers
  • Views 812 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

C2000WARE: f2837xs_diagnostic_library - March13n Test patterns

Peter Berger
Prodigy 210 points
Part Number: C2000WARE

Hi ,

I am using march13n test from C2000Ware_1_00_01_00,

I am trying to understand the affect of test patterns on the coverage and if i need to run the test with all four patterns,

i found this:

from C2000 CPU Memory Built-In Self-Test spracb9.pdf.

This seems to indicate i need to run all four patterns, this would make quadruple the time it will take me to finish a complete run of the test,

Please explain the need why we need these patterns..

  • 0
    TI__Expert 7985 points

    Hi Peter,

    The test patterns are required for ECC and Parity.  You can take a look at section 2, entitled: System Challenges to Memory Validation (in SPRACB9.pdf)  to understand the impact of running these test patterns while the application is active.  Please let us know if this helps you or you need further assistance.

    Regards,

    Krishna  

  • 0 in reply to Krishna Allam
    Prodigy 210 points

    Hi Krishna,

    Thank you for your answer,

    I meant something a little different from what you answered,

    i will try asking it from another angle,

    if i use only one RAM Test Pattern - will it be enough to detect all possible ECC and Parity errors?

    Thank you,

    Peter Berger

  • 0 in reply to Peter Berger
    TI__Mastermind 32995 points
    Hi Peter,

    To clarify all four test patterns are provided to get maximum coverage for worst case speed paths.

    To answer your questions, you only need to use one pattern. Executing the march algorithm with any of these provided test patterns will provide you will 100% stuck-at-bit (SA) coverage of the RAM being tested including the data and EDAC bits.

    You do not need to use all 4 test patterns. One will test for all stuck-at-bits. We provide the 4 different test patterns to aid in testing worst case speed paths of the RAM bit cells. This isn't required, but is additional.

    Also, please see www.ti.com/.../spracb9.pdf

    Hope this helps,
    sal