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.

SN75LVPE802: PCIE GEN3 eye diagram test RMS jitter fail

Expert 3420 points

Replies: 4

Views: 138

Part Number: SN75LVPE802

Hi team,

My customer has the following problems, which are described in Chinese. I am not familiar with eye diagram test. I translate the questions from Chinese to English. If there are descriptions that confuse you, please point out.

1、U.2-1  without redriver : eye diagram test PCIE GEN3 RMS jitter fail,lan0-lan3 all fail,3.2~3.6ps, Shorter route, Intel spec less than 3ps

U.2-2  with redriver : eye diagram test PCIE GEN3 RMS jitter fail,lan0-lan3 all fail,3.7~4.2ps, Longer route

2、PCIE  GEN3 clock test:About jitter part test pass

3、Use the PCIe X4 and U.2 that passed the test to cross verify the clock and date eye diagram test:

U.2 clock + PCIE X4 date: RMS jitter pass.2 date  + PCIE X4 clock: RMS jitter  fail

4、Adjust DEW,DE,EQ, eye diagram changes, RMS jitter still fail

Help to see the relationship between RMS jitter fail and date, thank you!

attachments:

3730.pac-2000mb-q370 U.2.pdf

3173.SAK4 U.2 GEN3 lane0 data.pdf

SAK4 v100_layout guide-PCIE.xlsx

PCIe Compliance Report for U.2 1.pdf

PCIe Compliance Report for U.2 2.pdf

U.2-2 GEN3 lane0 data.pdf

Best Regards,
Amy Luo

Best Regards,
Amy Luo

4 Replies

  • redriver may help ISI or Dj, but it will not help RJ

  • In reply to Brian Zhou:

    Hi Brian,

    Thank you for your reply.

    What does RJ have to do with? Could you give some advice? Thank you in advance.

    Best Regards,
    Amy Luo

    Best Regards,
    Amy Luo

  • In reply to Amy Luo:

    RJ is random jitter, it's from you system , like power noise. So you need decoupling cap on each power supply.

  • In reply to Brian Zhou:

    Hi,

    The Rj RMS failure is not typically associated with a redriver.  The redriver is designed to reduce or eliminate ISI jitter due to trace/cable attenuation.  It will have very little impact on the overall Rj RMS number.  I would expect typical measurements with and without the redriver to have similar Rj RMS numbers.

    The redrivers have a ferrite bead connected to the supply.  You could try a couple things to see if there is anything noise related on the redriver supply.

    1. Add a 10-22uF capacitor after the FB to improve the local decoupling and repeat the Rj test.

    2. Shorting the FB out and repeat the Rj test

    Do either of these adjustments change the results? 

    Regards,

    Lee  

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.