Because of the Thanksgiving holiday in the U.S., TI E2E™ design support forum responses may be delayed from November 25 through December 2. Thank you for your patience.

  • State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 32 subscribers
  • Views 207 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.

DS125DF1610: Time to finish Adap-1

Chris Zhao1
Expert 1240 points
Part Number: DS125DF1610

Hi expert,

I'm using Adapt-1 mode in my design, CTLE only mode, based on the FOM, DS125DF1610 will go through the default 32 group values(0x40-0x5f), then select a final value to use. 

can you tell me how long this Adapt-1 mode will take to get this final value, or in other words, how many bits will be used to calculate the FOM of one value in 0x40-0x5f. 

I think it's pretty fast, need exact info from your side.

Thanks

Chris

  • 0
    TI__Mastermind 19452 points

    Hi Chris,

    The typical lock time in adapt mode 1 is ~74ms. This assumes 20 inches of channel loss, with the HEO/VEO lock monitor enabled.

    As the lock time is application dependent, I recommend measuring the amount of time from a valid input signal being present to when the output sends retimed data.

    Thank you,

    Evan

  • 0 in reply to Evan Mayhew
    Expert 1240 points

    Thanks, I can make a measurement to find out.

    In my board, a IC will send data to DS125DF1610, while the initial data pattern is like below:

    send 1010 pattern for 2us

    -> stop 1us

    -> send 1010 pattern about 3us

    -> stop 1us

    -> send 1010 pattern for 40us

    -> normal data pattern

    Will this kind of initial pattern have influence on EQ of Adapt-1?

    Thanks

    Chris

  • 0 in reply to Chris Zhao1
    TI__Mastermind 19452 points

    Hi Chris,

    I don't expect the duration of this initial pattern to have an effect on the EQ.

    To confirm this, the HEO/VEO on the TX side of the retimer can be compared after lock with and without the initial pattern present.

    Thank you,

    Evan

  • 0 in reply to Evan Mayhew
    Expert 1240 points

    Hi Evan,

    Thanks for your quick response.

    Unfortunately, I can't turn on/off the initial pattern.

    Regards

    Chris

  • 0 in reply to Chris Zhao1
    TI__Mastermind 19452 points

    Hi Chris,

    Although it seems unlikely the initial pattern will have an effect on EQ during adaptation, another option is to set a constant EQ value manually with adapt mode 0 relative to the amount of insertion loss.

    Thank you,

    Evan

  • 0 in reply to Evan Mayhew
    Expert 1240 points

    Hi Evan,

    For now I'm using Adapt-1, and the insertion loss on my board is about 5-6dB@10Gbps, I'm using 10.3125Gbps, 

    Most of the EQ results are:

    1) 0-0-0-0, 0.29dB@2.5GHz

    2) 0-0-0-1, 3.69dB@5.5GHz

    3) 0-0-1-0, 4.06dB@6GHz

    Sometimes there will be other value, like 0-0-0-2, 1-0-0-0, which value should I use, kind of confuse.

    Thanks

    Chris

  • +1 in reply to Chris Zhao1
    TI__Mastermind 19452 points

    Hi Chris,

    The adapted EQ value may have some variance each time lock is acquired. One strategy is to note the HEO/VEO in response to each EQ value, and use the value with the best performance. Typically, matching the EQ value to your expected insertion loss is sufficient.

    Thank you,

    Evan