• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 35 subscribers
  • Views 425 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.

Does the pair of tlk1501s need the same clock exactly?

Tom Green
Prodigy 30 points

Other Parts Discussed in Thread: DS90CR288A, TLK1501

I have designed a project without FPGA!
It is supposed to work this way.
10 bits data                                  --tlk1501(A)--SFP--FIBER--SFP--tlk1501(B)--10bits data
60MHz(A) clock(input) is GTX_CLK --                                                            --60MHz clock(output) is RX_CLK
                                                              |                                             |
                                                              |                                             |
                                                              |                                             |
                                                              |                                             |
                                                       not cared                      60MHz(B) clock(input) is GTX_CLK
First , Only 60MHz(B) clock is supplied by crystal oscillator.and 60MHz(B) clock is supplied by DS90CR288A's clk output.
Then,when I work them this way as I planned, There are errors on rx_er.
when I discard the crystal oscillator and supply the clock of tlk1501(B) with tlk1501(A)'s clock(60MHz(A)),in anther word, the two tlk1501 share one clock,
they work defintely right!
So it is obviously I need two same clock source.Is it right? How's that possible?
who can give me some possible plans or chip suitable?

 

  • 0
    TI__Expert 7850 points

    Hello,

    Given your description of the setup there are two possible reasons why clock B may not be working.  The reference clocks on both sides of the link must be the same frequency   (within 200ppm).  Also, the reference clock on both sides must meet the GTX_CLK jitter requirement as stated in the data sheet(40ps p-p).  Since the clock from side A is working, it possible that the clock on side B does not meet the jitter requirement and/or is not exactly the same frequency as the frequency on side A.  Please measure the jitter of clock B and determine if it meets data sheet requirement. You can also put the same crystal oscillator on both sides to potentially resolve the issue.

     

    Thanks,

     

    Atul Patel

     

    Texas Instruments