• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 47 subscribers
  • Views 309 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.

OMAPL137HT AEMIF timings Temperature characterization and Max AEMIF Module clock

MAlanis
Prodigy 245 points


Hi,

1) The OMAPL137HT notes specifically for AEMIF timings, that "40°C to 125°C unless otherwise noted."  However the device is being sold/marketed at 175C, so what are the implications of the device AC timings when running at up to 175C?  

2) Also, it is not clear from the OMAPL137HT datasheet what the maximum AEMIF module clock supported speed is for ASYNC (non SDRAM) mode is?  The regular OMAPL137 Datasheet Table 5-24 lists it as 10ns (implying a 100Mhz module clock), however subnote #1 of the same table, mentions  that the EMIF "EMA_CLK is selected either as SYSCLK3 or the PLL output clock divided by 4.5. As an example, when SYSCLK3 is selected and set to 100MHz, E=10ns"   This implies that the EMIFA clock could be operated from the PLL0/4.5, which could yield 133Mhz.  Can someone sort this out please?

Regards,

Miguel

  • 0
    TI__Expert 7470 points

    Hi,

    MAlanis said:
    2) Also, it is not clear from the OMAPL137HT datasheet what the maximum AEMIF module clock supported speed is for ASYNC (non SDRAM) mode is? The regular OMAPL137 Datasheet Table 5-24 lists it as 10ns (implying a 100Mhz module clock), however subnote #1 of the same table, mentions that the EMIF "EMA_CLK is selected either as SYSCLK3 or the PLL output clock divided by 4.5. As an example, when SYSCLK3 is selected and set to 100MHz, E=10ns" This implies that the EMIFA clock could be operated from the PLL0/4.5, which could yield 133Mhz. Can someone sort this out please?
    The EMIFA frequency is limited to 100MHz for both ASYNC and SYNC operation. Yes, the settings can yield a higher operating frequency but that would be running the device out of spec. You can use the div-by-4.5 clock output to drive EMIFA if you configure output of PLL0 to provide 450MHz or lower. Regards, Sunil Kamath

  • 0
    TI__Guru 62930 points

    Sunil, Thanks for answering the second question.

    Miguel,

    The EMIF and other peripherals are functionally test and characterized to the extended temperature range.  Unfortunately not all timing parameters are testable at extreme temp range.  In these cases where we cannot test every parameter, we are noting the fully characterized parameters temperature range.

    A good design practice for high temperature products would be to design conservatively to provide guardband for these timing parameters. 

    Regards,

    Wade

  • 0 in reply to Wade Vonbergen
    Prodigy 245 points

    Sunil and Wade,

    Thanks for your help and clarifications.

    regards,

    miguel

  • 0 in reply to Wade Vonbergen
    Prodigy 245 points

    Wade,

    Just a followup  on this question, what are the temp ranges when you say, 'Extended' and 'Extreme'?

    The Datasheet defines the following grades.

    A = -40°C to 105°C (Extended Grade)

    T = -40°C to 125°C (Automotive Grade)
    H = -55°C to 175°C (High Temp Grade)

    Regards,

    Miguel

  • 0 in reply to MAlanis
    TI__Guru 62930 points

    I apologize for using the vague terminology.

    In this case, I am referring to the high temp grade H parts for both.  Meaning, we can functionally verify the peripherals and logic.  However we cannot verify all timing parameter that are listed in the datasheet.  Thus the notation in the specification.

    Regards,

    Wade

  • 0 in reply to Wade Vonbergen
    Prodigy 245 points

    Wade,

    Thanks.  The (-40 to 125C) notation exists in mostly all the Peripherals timing parameters with exception of the GPIO module.  As a developer it is important to understand what kind of degrading factor on the AC Timing parameters occurs at the High Temperature since it will directly dictate what kind of parts the peripheral interfaces to and to see if they will meet the AC Timing requirements else they will fail to meet timings at the High Temperature.  In the cases where the AC timings directly affect individual data throughput performance, such as a Memory Interface, this could affect both interface performance and potential incompatibility to meet degraded AC timings. Therefore the importance to have this information on the datasheet.
    Are there any plans in the coming future to provide this AC timing degradation at >125C?

    It is not enough to say   "A good design practice for high temperature products would be to design conservatively to provide guardband for these timing parameters. "
    How do you determine what 'design conservatively to provide guardband' is?


    I hope you understand the concern here.

    regards,
    Miguel