• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 33 subscribers
  • Views 651 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.

SN65LVCP114: MUX multiple 10G-KR lanes?

AnBer
Mastermind 29420 points
Part Number: SN65LVCP114
Other Parts Discussed in Thread: DS100MB203, DS100MB201, HD3SS3412

Team,

The use case is to MUX one 10G Ethernet source (4Tx 10G-KR lanes and 4Rx 10G-KR lanes) to 2 boards over a back-plane.
What would be the most cost efficient solution to do this?

SN65LVCP114 could be an option but it is quite expensive as you would need 4 SN65LVCP114.

-What would be the most cost effective option? May be DS100B201 or DS100B203?
-What are the main pros/cons with both options compared to SN65LVCP114?
-Are there additional options to the TI web selection swicth/MUX table below?
http://www.ti.com/interface/pcle-sas-sata/retimers-redrivers-and-switches-muxes/products.html#p1694=Switch/MUX

Thanks in advance,

A.

  • 0
    TI__Mastermind 43795 points

    Hi,

    The SN65LVCP114 is twice the density of the DS100MB203, so only one device would be needed.  Two DS100MB203 devices would be required to create a quad 1:2 2:1 function.

    I would not recommend the DS100MB201 it is an older device.

    The only other option I can think of would be a passive mux like the HD3SS3412.  This is definately the cheapest option, but the passive switch will limit the reach of the 10G-KR interface.

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    TI__Mastermind 29420 points

    Hi Lee,

    Thanks for the answer.

    Is there a way to use webench to do a quick simulation on both SN65LVCP114 and HD3SS3412 and compare what is realistic in term of distance reach on the backplane?
    I can see that SN65LVCP114 is in webench but not HD3SS3412.

    Best regards,

    Anthony

  • 0 in reply to Lee Sledjeski
    Prodigy 20 points
    Hello Lee,

    just to give you some technical info.
    Worst case trace length we use in simulations is 8.5 inch on the central switching board, 12 inch on the back plane and 5 inch on the interface card to connect a phy to a switch.
    The mux we are looking for will sit in this path on the IFM board and can be close to the back plane connector. So for an active component the trace length will be +/- 22 inch. In between are 2 connectors (e.g. an Examax and Tinman). Assume PCB material ME888.
    In this setup I doubt if a passive mux is an option.
    Can you confirm that my initial feeling to use an SN65LVCP114 will provide a working solution ?
    We also need only a single device (as it is a quad 1:2 2:1 mux) correct ?

    Some other questions:
    - This component has an 'active' status, so it can be safely used in new designs ?
    - How long is this component already in production ? (and how long will it be continuate ?) (need at least 10 year)
  • 0 in reply to marc moeskops
    TI__Mastermind 43795 points

    Hi Anthony,

    The HD3SS3412 is not currently supported in WEBENCH.  The distance/attenuation of the 10G-KR link does make the passive mux a risk in this application.  I would agree the SN65LVCP114 is a good single device solution.

    The SN65LVCP114 can safely be used for new designs.  It has been in production for approximately 6 years.  It is impossible to guarantee 10 year avialability for a product, but this type of function will be viable for many years to come.  Similar devices have been in production for > 20 years.

    Regards,

    Lee