• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 33 subscribers
  • Views 96 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.

DS90UB925Q-Q1: TX Buffer usage recommendation for long cable distance

Kiyoun Ryu
Prodigy 10 points
Part Number: DS90UB925Q-Q1
Other Parts Discussed in Thread: DS25BR100, ALP

Tool/software:

Hello TI Support Team,

I am a hardware design engineer at LIG Nex1, currently working on a system using the FPD-Link III devices: DS90UB925QSQE/NOPB (serializer) and DS90UB926QSQE/NOPB (deserializer).

We have tested our system using various cables and observed the following issues when no TX buffer was used between the serializer and the cable:

1. The system's performance varied depending on the cable type and length.
2. When using cables longer than approximately 3 meters, the deserializer frequently lost lock.

I am aware that FPD-Link III devices are typically designed for point-to-point connections, and that external LVDS drivers or redrivers are generally not recommended. However, due to the lock instability over longer cables, I would like to ask the following:

---

Questions:

  1. Can a TX buffer (such as an LVDS redriver or repeater) be safely used between the DS90UB925QSQE and the cable without violating signal integrity or protocol timing?
  2. If so, which TI-recommended devices (e.g., DS25BR100 or others) are appropriate for use with FPD-Link III, especially at cable lengths over 3m?
  3. If using a TX buffer is not recommended or supported, what is the best practice to ensure reliable link lock over long cable distances (3m+), especially in noisy environments?
  4. Are there any official application notes or design guides specifically for FPD-Link III systems regarding signal extension or cable driver strategies?
  5. The datasheet mentions that AC-coupled STP cables can support transmission up to 10 meters — are there any specific design or layout considerations we should follow when building such cables?

We would greatly appreciate your guidance and any technical documentation you can recommend for robust long-distance FPD-Link III transmission.

Thank you in advance for your support.

Best regards,
Kiyoun Ryu
Hardware Design Engineer
LIG Nex1

  • 0
    TI__Mastermind 19845 points

    Hi Kiyoun,

    Kiyoun Ryu said:
    Can a TX buffer (such as an LVDS redriver or repeater) be safely used between the DS90UB925QSQE and the cable without violating signal integrity or protocol timing?
    Kiyoun Ryu said:
    If so, which TI-recommended devices (e.g., DS25BR100 or others) are appropriate for use with FPD-Link III, especially at cable lengths over 3m?
    Kiyoun Ryu said:
    If using a TX buffer is not recommended or supported, what is the best practice to ensure reliable link lock over long cable distances (3m+), especially in noisy environments?

    While it is not recommended to put any redrivers or repeaters for the FPD-Link connection between the device TX and RX of the SER and DES, there are other recommendations for the configuration that can be used to optimize the signal integrity.

    For example, the STP configurations of 92x devices often recommend a CMC, please see the example diagram below:

    As long as the paired CMC is also at the DES side, this should improve signal integrity.

    There are a few layout guidelines that are available for the FPD3 devices under section 10, but the option above is also available to help improve the channel integrity even over different lengths.

    Another option for verifying link integrity at different EQ levels (there is an adaptive equalizer on the 926 for this scenario), is to use the ALP application with Margin Analysis to determine if there is sufficient margin at different lengths. Please see the example of the environment below:

    References:

    As long as FPD-Link components over the link are meeting signal integrity, and 92x/94x channel specifications, lock should be stable under these conditions.

    Please let me know if you would like to request any further documentation from here.

    Best,

    Miguel