• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 100 subscribers
  • Views 262 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.

Original question:

AM335x: IBIS and DDR_CK and DDR_DQS differential models

AM3352: Hyperlynx simulation cannot proceed due to lack of CK/CKn diff pair in IBIS model

Jesse Hawkins
Prodigy 60 points
Part Number: AM3352

I am attempting a DDR simulation of the am3352 using the sprm552c ibis model. The simulation fails with the following error:

Clock Net: DDR_CLK ---
Pin: U4.J7 (+) IBIS Signal: mt41k256.ibs [MT41K256] CK
Pin: U1.D2 (not differential) IBIS Signal: sprm552c.ibs [am335x_zcz] DDR_CK

Clock Net: DDR_CLKn ---
Pin: U4.K7 (-) IBIS Signal: mt41k256.ibs [MT41K256] CK#
Pin: U1.D1 (not differential) IBIS Signal: sprm552c.ibs [am335x_zcz] DDR_CKN

Meaning, the CK and CKn signals on the endpoint are differential as expected, but the DDR_CK on the AM3352 is single ended. I saw another forum post from years ago on this but there was no resolution. Can you provide a new Ibis model with differential signaling on the clock signals? Or is simulation in hyperlynx not possible with this setup?

Looking through the IBIS file, the only diff pairs in the design are the DQS pairs.

  • 0
    TI__Guru**** 168894 points

    Hello Jesse Hawkins

    Thank you for the query.

    Let me check with the simulation expert.

    Please expect delay in response due to the thanksgiving holidays.

    Could you have a look at the below thread and see if this helps

    (28) AM335x: IBIS and DDR_CK and DDR_DQS differential models - Processors forum - Processors - TI E2E support forums

    Regards,

    Sreenivasa

  • 0 in reply to Kallikuppa Sreenivasa
    TI__Guru**** 168894 points

    Hello Jesse Hawkins

    Below are the inputs i received from the simulation expert.

    After reviewing the comments in the other E2E thread, the AM3352 die contained two single ended output buffers for the CK/CKn as opposed to the DQSn/DQS which contains a true differential output buffer.

    That is the reason the CKn/CK is not specified in the “Diff Pins” section in the IBIS model so this is consistent with the model selection.

     The customer should be able to work with Mentor to allow the HyperLynx tool to interpret the CK/CKn as a differential pair. The below comment from the E2E states that a similar workaround was found earlier as well. Most EDA tools should be able to support this.

    Pls let me know if there is any progress from your side.

    Regards,

    Sreenivasa