• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 24 subscribers
  • Views 1133 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-Q1 Level Translator skew between the outputs,clarification Required

sathiya seelan sekar
Prodigy 20 points
Other Parts Discussed in Thread: SN74AVC16T245-Q1, LSF0108

Hi,

     We are using the SN74AVC16T245-Q1 level translator(16 IOs)  to convert the 1.8V to 2.5V.

Q1. Will this device suitable to be used in GMII (Gigabit Ethernet interface)?

MAC side is 1.8V IO,the PHY chip is 2.5V IO. Will this level translator suits this application?

Q2. Will there be any skew among the 16 outputs? 

The Tphl & Tplh are given some range values about 0.5 to 7.5 ns. I want to be clarified that the output1 to 16 will have same propagation delay values or it varies from port to port?

Please clarify us for the above two queries.

Thanks,

S.Sathiya Seelan

s.sathiya@mobiveil.com

  • 0
    TI__Genius 14890 points

    Sathiya,

     

    I've assigned an engineer to look at this post and get back to you. Thanks

  • 0 in reply to Alexander Scheuermann
    Prodigy 20 points

    Hi,

          Any update on my query?

    -Sathiya Seelan

  • 0 in reply to sathiya seelan sekar
    TI__Genius 14890 points

    Sathiya,

     

    I have contacted the group responsible for this device again. If it does not get answered in a few days, post again and I will escalate the request.

  • 0 in reply to Alexander Scheuermann
    TI__Mastermind 20580 points

    Hi Sathiya,

    I've just looked at this and have worked on Ethernet in the past:

    A1:  I need to get some more info on this as your Q2 directly influences the answer.

    A2: This is a very valid question and the datasheet doesn't answer it.  GMII interface will allow some skew between it but I will have to see if there is any data on this device to support channel to channel skew which will be critical in addition to the absolute delay through the translator.  In parallel I'm asking our GigE PHY expert if there is anything else we should consider.

    - Scott

  • 0 in reply to Scott Monroe
    TI__Mastermind 20580 points

    Hi Sathiya,

    A1:  It may work.  We have no detailed information on channel to channel skew.  The TI Ethernet team has been using LVDS serialiers in some of their boards between the MAC and PHY for GigE.  They know the cummulative effect of the 2 LVDS devices are: 4.4ns total delay max and 3.7ns max skew device to device (multiple devices are needed to get enough channels for the GMII).  They have no issues with that amount of skew and delay.  The SN74AVC16T245-Q1 is slower at a specified delay of 7.5ns max for the 1.8V to 2.5V translation. We have no detailed channel to channel skew data but it will not be more than 7.5ns, so the device may work.  Do you require an AEC -Q100 device?     Alternatively there are other devices in the SN74AVC family with lower propogation delays and alternative devices like the SN74AVCHxTxxx family or LSF0108 (8 channel shifter) where the prop delays are much faster so they would keep the delay and skew under the amounts we know work from the LVDS solution.   I'm waiting on the Ethernet team if they can find if there is an actual maximum amount of allowable delay specified for GMII. 

    A2:  there is no detailed characterization on the channel to channel skew or device to device skew for the SN74AVC16T245-Q1.

    - Scott

  • 0 in reply to Scott Monroe
    TI__Mastermind 20580 points

    Hi Sathiya,

    The Ethernet team got back to me and they didn't find a specific spec on delay or skew, however since they know it works they recommend staying in the range they gave since they know it works at least on the MAC and PHY combo they are using.   The SN74AVC16T245-Q1 may well work, maybe you could ask your PHY vendor the max delay they could afford.  Above I gave some other alternatives if not.

    - Scott