Forum Discussion

Shweta_24's avatar
Shweta_24
Icon for New Contributor rankNew Contributor
2 years ago

MAX-10 IBIS model issue

Hello team,

I am working on IBIS model validation of MAX10

I have observed that the temperature ranges in IBIS file for CMOS and TTL models are same.

Ideally, For CMOS temperature which causes the slow, weak conditions in the "MIN" column

and temperature which causes the fast, strong condition in the "MAX" column.

For bipolar, exactly opposite, "MIN" column contains Highest value and "MAX" column contains lowest value.

This issue will impact the simulation results while using slow, typ, max corner cases.

I have attached below screenshot of both models temperature range

Please help me to solve this issue.

Thank you in advance,

Shweta Chendake

Board Debug

6 Replies

  • Mengzhe's avatar
    Mengzhe
    Icon for Occasional Contributor rankOccasional Contributor
    2 years ago

    Hi Shweta,

    The present IBIS model of MAX10 has no update plan and the temperature characteristic is correct.

    Please feel free to use it.


    BR,

    Mengzhe




  • Mengzhe's avatar
    Mengzhe
    Icon for Occasional Contributor rankOccasional Contributor
    2 years ago

    Hi,


    Thanks for your advice.

    In my opinion, the difference between LVTTL and LVCMOS model is the voltage leveling difference, not meaning the difference on physical die. As you can change the output mode from LVTTL to LVCMOS in a specific GPIO pin through Quartus project, however, the structure of gate in the I/O bank has not been changed.

    So, don't worry about they have different electric characteristic. The IBIS model just reflect the I-V curves under different situation.


    BR,

    Mengzhe


    • Shweta_24's avatar
      Shweta_24
      Icon for New Contributor rankNew Contributor
      1 year ago

      Hello Mengzhe,

      I am using MAX10 IBIS model from that model- lvttl_rio_d4s2 in this model

      the power clamp table exhibits stair-stepping .

      According to IBIS specification ,There should be no stair stepping of any I-V tables, with flat sections and abrupt jumps. This is caused by
      insufficient significant digits in the table current columns

      can you please check attached snapshot below

      Regards,

      shweta chendake

  • Shweta_24's avatar
    Shweta_24
    Icon for New Contributor rankNew Contributor
    2 years ago

    Hello Mengzhe,

    Thank you for your reply.

    In IBIS, Temperature keyword is there to indicate what temperature was used when the data was extracted from spice simulation or lab measurement.

    It is up to the model maker to select correct number for it.

    The meaning of the IBIS temperature is simply that the model's data was generated at these temperatures(typ/min/max).

    So, my concern is regarding that temperature range it should not be same for both CMOS and TTL models.

    In my opinion Temperature range should be updated in IBIS file for TTL models.

    Thank you,

    Best regards,

    shweta chendake

  • Mengzhe's avatar
    Mengzhe
    Icon for Occasional Contributor rankOccasional Contributor
    2 years ago

    Hi,


    I think your opinion is right on the different I-V characteristic of TTL and CMOS under different temperature.

    However, IBIS model of FPGA is just a behavior model, showing I/O I-V characteristic. You can treat it as a black box, not mean there is real CMOS or TTL component. So it does not totally follow TTL and CMOS characteristics under different temperature.

    Also, you can refer to this article: How do IBIS models relate to device speed grades? (intel.com)


    Best Regards,

    Mengzhe


  • Shweta_24's avatar
    Shweta_24
    Icon for New Contributor rankNew Contributor
    2 years ago

    Hello Mengzhe,

    Hope you are doing well.

    Can we get IBIS model of MAX10 updated with correct temperature?

    Thank you,

    Shweta