• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 17 subscribers
  • Views 647 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.

SN74AVC16T245 for 125 MHz GMII

Andrew Toth
Prodigy 70 points
Other Parts Discussed in Thread: SN74AVC4T245, SN74AVC16T245, LSF0108

Hello,

I'm trying to set up an interface from a legacy FPGA which only has 3.3V I/O to an ASIC which requires 2.5V for a gigabit ethernet GMII interface. This will be 8/10 bits running at 125MHz. It uses 8b/10b encoding so I believe that only 8 bits are actual data but I will need to provide the other 2 bits as control codes and I will need to send a TX clock as well.

At first I was considering using a couple SN74AVC4T245 parts but I became concerned about too much skew between the different parts so I began researching wider parts. It seems like the SN74AVC16T245 and other wider parts might be switching too slowly for what I'm trying to do. 

I saw elsewhere in the forums a recommendation to use the LSF0108 part for higher speed applications so my question is what would the skew look like from one part to another? (That question would apply to my original idea of using several of the SN74AVC4T245  parts as well)

Thanks,

-Andrew

  • 0
    TI__Mastermind 46480 points

    The LSF0108 parts are basically switches so the Skew would be very small. The max frequency may not reach 125Mhz if you are translating up. the speed will be determined by the pullup resistor.

    The AVCxT245 will be buffer type translators so they would have greater skew but is would be in the range of 3 400ps within the same part so the 16T245 might be best bet. That way you avoid part to part skew.

  • 0 in reply to Chris Cockrill
    Prodigy 70 points

    Will the 16T245 part switch fast enough to support 125MHz data / clock signals? I understand some of that will depend on the load... the -Q1 automotive variant of the part lists some datarates but this one does not.  Is there a fundamental difference in the parts or just an estimate on the -Q1 datasheet?

    -Andrew

  • 0 in reply to Andrew Toth
    TI__Mastermind 46480 points

    the automotive parts are the same parts just tested differently. These the lvcXt parts dont belong to me but the best way to check max frequency is to use IBIS model into your load conditions. This will give you rise and fall times so you can calculate speeds.

    I will move this to the translator forum for further help.