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.

DS280BR820: Adjust Limiting (FIR) Mode at IBIS-AMI

Part Number: DS280BR820

Hello team,

I want to evaluate DS280BR820 at Adjust Limiting (FIR) Mode.

According to Programmer's Guide, we can adjust 3 kinds of parameters [pre,main,post].

At first, I actually evaluated DS280BR820 at Adjust Limiting (FIR) Mode and adjust these all 3 parameters by using EVM. 

example setting) [pre,main,post] = [-5,20,-5]

After this, I tried to evaluate DS280 by using simlation (IBIS-AMI) to correlate with these actual results but I could not correlate with IBIS-AMI results.

How can I use Adjust Limiting (FIR) Mode at IBIS-AMI by adjusting 3 kinds of parameters [pre,main,post]?

Sincerely yours,

  • Hi,

    Are you able to share both your EVM testing and also your redriver simulation results? It's hard for us to provide informed feedback on your issues until we can review your observations.

    Thanks,

    Rodrigo Natal

  • Hello, Natal

    Thank you for kindly reply.

    I share you the results about EVK testing and simulation. Also I describe measurement conditions at each evaluations.

    <EVK testing>

    Bitrate:18Gbps PRBS31

    *CTLE settings
    [EQ1,EQ2,EQbw] = [7,7,3], EQ DCgain = Low,

    *FIR settings
    FIR tap [pre,main,post] = [5,20,5], Vod = 3

    <Simlation(IBIS-AMI)>

    Sim conditon:ADS simulation (Channel Sim)

    Bitrate:18Gbps PRBS31  Number of bits:10000

    DS280 AMI model Parameters
    *CTLE settings
    [EQ1,EQ2,EQbw] = [7,7,3], EQ DCgain = Low, LTI_mode = 1

    *FIR settings
    FIR tap [pre,main,post] = [5,20,5]
    Vod = 3, sel_fir = 1, LTI_mode = 1

    ↓Sim result and EVK result.


    A sharper waveform compared with the result of EVK was obtained as a result of Sim.

    I could not the reason of this difference.

    Please let me know if there is any missing information about my evaluations.

    Sincerely yours,

  • Thank you for sharing more details regarding your EVM test setup and simulation settings.

    I went back and performed some simulation analysis using the DS280BR8x0 example project with data rate=18Gbps. See below. I did not see any anomalies with my simulated results. Everything looks good and as expected. I would speculate that your unusual simulation observations are an artifact of either your simulation test point and/or channel configuration. Do review my results below and revise your redriver simulation setup as needed.

    Test bench

    I. Baseline case: sel_fir=0, VOD=3, [EQ1,EQ2,EQbw] = [7,7,3], EQ_gain = 0, LTI_mode = 1

    II. sel_fir=1, C0=30, Cp1=0, Cn1=0, [EQ1,EQ2,EQbw] = [7,7,3], EQ_gain = 0, LTI_mode = 1

    (To test out the FIR EQ function, i replaced the 2-inch FR4 output trace channel in my test bench with 5-inch FR4 then tuned the FIR taps to optimize the eye opening

    III. (5-inch FR4 redriver output channel) sel_fir=1, C0=30, Cp1=0, Cn1=0, [EQ1,EQ2,EQbw] = [7,7,3], EQ_gain = 0, LTI_mode = 1

    IV. (5-inch FR4 redriver output channel) sel_fir=1, C0=28, Cp1=2, Cn1=0, [EQ1,EQ2,EQbw] = [7,7,3], EQ_gain = 0, LTI_mode = 1

    IV. (5-inch FR4 redriver output channel) sel_fir=1, C0=25, Cp1=5, Cn1=0, [EQ1,EQ2,EQbw] = [7,7,3], EQ_gain = 0, LTI_mode = 1

    Thanks,

    Rodrigo

  • Hello Natal,

    Sorry for the late reply.

    Thank you for some simulate examples.
    I tried to simulate by using 2 parameters [main,post] or [main,pre] like your examples, butI could not get similar results like your results.

    When I used "PKG_ZBL_RX0.s4p" and "PKG_ZBL_TX0.s4p", I got strange waveform like these below figures.
    I reviewed whether there were any mistakes and wrong settings but I could not resolved.
    Do you know the causes about these results? (I did not use PKG at first simulation)

    <Sim Parameters>
    Tx:Electro 100ohm, Exclude load unchecked
    Rx:Electro 100ohm, Exclude load unchecked

    Redriver AMI:
    *CTLE settings
    [EQ1,EQ2,EQbw] = [7,7,3], EQ DCgain = Low, LTI_mode = 1

    *FIR settings
    FIR tap [pre,main,post] = [0,25,5]
    Vod = 3, sel_fir = 1, LTI_mode = 1

    Also,I would like to use 3 parameters [pre,main,post] and know how to use these parameters.

    Could you show me some Sim examples by using 3 parameters [pre,main,post] or could you teach me settings?
    Or, Can not I use 3 parameters same time?

    Sincerely yours,

  • Your results seem unusual. I would speculate there is a simulation setup issue on your side. I'd recommend to reproduce my simulation test bench exactly and see whether you are then able to observe results like mine. Using the same exact simulation setup should facilitate debugging. Note that the FR4 test channels I used in my simulation are included as part of the data files in the downloaded TI IBIS-AMI files package.

    Thanks,

    Rodrigo Natal

  • one thing. do double check to ensure that you are using the correct port orientation for all of your s-parameter elements.

    Thanks,

    Rodrigo

  • Hello Natal,

    Thank you for the easy-to-understand explanations.

    After reviewing the settings, I found the differences about PKG ports.
    And I got similar results as your previous results after changing to the correct ports.
    I am so sorry for the lack of confirmation.

    I resolved the port problem, but could you tell me another question about using 3 parameters [pre,main,post] I wrote at the bottom part of the previous post?

    I would like to use 3 parameters [pre,main,post] at ADS simlation, so I would like to know how to use 3 parameters [pre,main,post] at ADS simlation.
    Can not I use 3 parameters same time in ADS simlation?

    Sincerely yours,

  • Related to: I would like to use 3 parameters [pre,main,post] at ADS simlation, so I would like to know how to use 3 parameters [pre,main,post] at ADS simlation.Can not I use 3 parameters same time in ADS simlation?

    • The redriver IBIS-AMI model indeed includes parameters allowing adjustment of its FIR taps when selecting the limiting mode.
    • On ADS you would go to the Tx Parameters.
      • sel_fir must be set to 1 to enable the FIR-limiting mode
      • Then, the user can adjust C0 (main) Cp1 (post-cursor) and Cn1 (pre) parameters to desired values noting the following,
        • The FIR coefficients absolute sum cannot exceed 31, 31 corresponds to the largest voltage output differential (VOD) supported
        • A constant VOD level is achieved as long as coefficient sum is kept constant. It the user wants to for example apply some post cursor then the main tap should be reduced by same amount post-cursor is increased to maintain a constant VOD level

    Thanks,

    Rodrigo Natal