• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 33 subscribers
  • Views 981 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

DS92LV18: CAT5 Cable length

Mark Guastaferro
Genius 11645 points
Part Number: DS92LV18
Other Parts Discussed in Thread: DS25BR110

System designer is trying to estimate the maximum length CAT5 cable that can be used with a DS92LV18 SERDES device on each end operating in full duplex. The cable would have LVDS data on two different twisted pairs, running in opposite directions. The SERDES devices would be operating at 64 MHz.

Looking at the SERDES Design Guide (SNLA202a), it shows a table for data rate vs cable length for the DS92LV18 on Page 7.  The assumption is this is for data running on one twisted pair.  How much would the length degrade with a second twisted pair running data in the reverse direction?

The Design Note also recommends using shielded pairs to minimize crosstalk when mulitple pairs in the same cable are used. Does each pair need to be indivudually shielded, or is it good enough to have only one pair shileded?   

  • 0
    TI__Mastermind 43795 points

    Mark,

    I would recommend individual shields in the cable if possible.

    Running high speed signals in a CAT5 cable in opposite directions requires some care.  The 1.2 Gbps serialized signals are above the specified range for normal CAT5 cables.  The RJ45 connectors are not helpful at the higher rates either.  I would use a pinout on the cable to keep these signal very far away from each other.  Something like pins 1-2 and 7-8 on the RJ45.

    With cable and connector sensitivity, it is hard to estimate the max cable length impact.  I would say cut the distance by 50% in this case.

    Regards,

    Lee  

  • 0 in reply to Lee Sledjeski
    TI__Genius 11645 points
    Hi Lee,
    Thanks for the quick response.
    Would CAT6 cabling help since it has better BW and crosstalk characteristics? I suppose the RJ545 connector is still a challenge.

    How about the affect of adding a buffer with pre-emphasis? This could help mitigate the cable BW limitations, but I suspect it would not help the crosstalk concerns (might actually make it worse).
  • 0 in reply to Mark Guastaferro
    TI__Mastermind 43795 points

    Hi Mark,

    I agree.  The CAT6 cable will reduce the overall loss slightly vs. CAT5e.  The RJ45 connector is still the limiting factor.  Using a pinout which spreads the Rx and Tx signals apart on the connector will help.

    I would suggest trying the DS25BR110 equalizer if the signals are attenuated.  It can work well even at 1.2 Gbps data.

    Regards,

    Lee