• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 32 subscribers
  • Views 636 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.

DS100MB203: Please help analyze the use of SFP and DS100MB203

user5888214
Prodigy 30 points
Part Number: DS100MB203
Other Parts Discussed in Thread: DS125MB203

Hi Sir：

        In our applications，transmit four SFP+ signals to QSFP+ , the schematic diagram such as attachments. Between SFP+ and QSFP+, use two DS100MB203 to change the signal path. We’ve designed the circuit card, but the signal only run up to 4.25Gbps. When it work in 10.3125Gbps, there is no link.

         I have the query as:

  1. We use two DS100MB203 in tandem, are we using MB203 in here correctly? What attention should be paid to this use.
  2. I know，there is a need for AC coupling capacitors between MB203, but, between SFP+ and DS100Mb203，is there a need for AC coupled capacitors? At the same time, Between QSFP+ and DS100Mb203, is it necessary? When I was testing, with or without capacitors, it can pass through 4.25Gbps, but 10.3125Gbps won’t make it.
  3. Currently, I’m in pin mode, working voltage is 3.3v, MODE try using either “0” or “1”. EQ_D1=0,EQ_D0=0,EQ_S1=0,EQ_S0=0.

DEM_D1=0,DEM_D0=0,DEM_S1=0,DEM_S0=0.

Is this a good setting?

  1. Previous question mentioned channel attenuation, I didn’t think about it before. On my board, differential line length between SFP+ and DS 100MB203 is about 1 inch. The length of two MB203 is about 7.2inch. What are the requirements for the length of the signal line?

 

Look forward to your reply. Thank you.

   

  • 0
    TI__Mastermind 43795 points

    Hi,

    Tuning multiple mux-buffers in a signal path can be difficult.  Starting with minimal EQ settings is a good idea.

    Usually there are AC coupling capacitors on the SFP and QSFP cards, so there is no need to add another AC coupling capacitor between the (Q)SFP+ cage and the MB203.

    I would try to increase the EQ by one step for the MB203 Rx pins which are receiving signals from the previous MB203

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Prodigy 30 points
    Thank you for your reply. There are also a few uncertainties to look forward to your answer.
    （1）On the circuit board, is there a minimum length of the optical module interface to the MB203 differential line? I learned from the previous question that 2 inch is a good fit，Is that so？
    （2）You said "Tuning multiple mux-buffers in a signal path can be difficult"。In that case, how to Control the Line length on the Circuit Board between two MB203，and，are there any special requirements for signal attenuation between two MB203？
    （3）In addition，the test waveform given in the datasheet is on the FR4 plate, whether the 10G signal needs a special high speed plate？
    （4）About the evaluation board，why not take the nearest route, but go round and round？
    Thank you!
    xiaosong.hu
  • 0 in reply to user5888214
    TI__Mastermind 43795 points

    Hi,

    1. A 2 inch length to the optical interface is a good fit.

    2. I would recommend a minimum distance of ~ 5 inches to account for the minimum DS125MB203 input equalization setting.

    3. For shorter distances like this FR4 is preferred.  The FR4 provides some attenuation to help reduce effects from any reflections on the channel.

    4. In most cases on our evaluation boards we try to equalize the trace lengths.

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Prodigy 30 points
    Thank you very much！I think your suggestion will be very helpful in our design.