• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 46 subscribers
  • Views 329 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.

DAC38j84 adaptative equalizer "adapt time"

Marc-Antoine SAVOYAT
Prodigy 190 points
Other Parts Discussed in Thread: DAC38J84

Hi,

I can read in the DAC38J84 datasheet that I have to "Allow sufficient time for the equalizer to adapt".

What is sufficient time?

Is there a register bit/value that I can check to be sure that equalizer is ready for JESD204B link synchronisation?

Can I reduce this "sufficient time" by sending PRBS pattern at startup?

  • 0
    TI__Genius 17290 points
    Hi Marc-Antoine,

    The equalizer lock time (with random data) is 6x10^6 / x [UI]. Where x is the CDR activity rate - defined by the CDR setting of the DAC - and UI is the bit rate of the link. For a CDR setting of 000 (default) the activity rate is 83%, so the lock time is 6x10^6/83 = 72.3 kUI. If you're running at 10 Gbps, the lock time is 72.3k / 10G = 7.23 us. If the raw BER performance of the link is 10^-12, then the probability of having an error occur during the 72.3 kbit lock time is 0.00000723%.

    Overall, I think you're better off not doing anything special for the equalizer. The serdes link will work on startup before the adaptive algorithm optimizes the equalizer settings - just the BER performance will be optimized after adaptation. The exception to this could be a really lossy channel (10 inches on FR4?) where the intersymbol interference will dominate the BER performance, but transmitter pre/demphasize would likely be used in this case as well to provide sufficient initial equalization.

    Regards,
    Matt Guibord
  • 0 in reply to Matt Guibord
    Prodigy 190 points

    Hi Mattt,

    Thank you for this update. From the experiments I made, the link quality is very good in my case, so it seems that I don't really have to care about the equalizer.

    Regards,

    Marc-Antoine