• State TI Thinks Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 2 answers
  • Subscribers 33 subscribers
  • Views 193 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.

DP83640: DP83640 and SFP Modules

Nicolas SAGLIETO
Prodigy 10 points
Part Number: DP83640
Other Parts Discussed in Thread: DP83869

Tool/software:

Hello,

We are currently using the DP83640 Ethernet PHY in our design, and it is configured in fiber mode (100BASE-FX) using the FX_EN_Z pin.

This setup works as expected when using a fiber SFP module, which communicates via standard 100BASE-FX without auto-negotiation.

However, to our surprise, the exact same configuration also works with a copper SFP module (specifically the SFP-FB-GE-T from FS.com), even though this is an RJ45 copper module typically intended for 100BASE-TX or SGMII applications.

We were expecting the copper SFP to require SGMII or 100BASE-TX signaling, which should not be compatible with the DP83640 in fiber mode. Yet, the link comes up successfully and data is exchanged.

Questions:

  1. Is there any known behavior or tolerance in the DP83640 that allows it to communicate with copper SFPs in fiber mode?

  2. Could it be that the copper SFP module is internally switching to 100BASE-FX mode?

  3. Is this operation considered valid or reliable from a PHY perspective, or is it just a fortunate coincidence?

We are trying to determine if this setup is safe for production or if it relies on undefined behavior.

Thanks in advance for your support!

Best regards,

  • 0
    TI__Mastermind 39795 points

    Hello,

    DP83640 is an older device and support is solely limited to datasheet. If FX is a must for your application, our DP83822 or DP83869 devices would be more modern devices which we can support.

    Regarding your SFP question, it may be possible that there could be enough signal level compatibility (VOH/L with VIH/L) between the CU SFP module pins and the PHY which could explain this. As noticed in SNLA479, SFP modules can be backwards compatible in certain circumstances like above. Most likely this is a fortunate coincidence. To determine if PHY is in CU or FX mode, you may probe the MDI with a differential trace. 100Base-TX is signaling using MLT-3 while FX is pure 2-level differential. 

    Sincerely,

    Gerome

  • 0
    Prodigy 10 points

    Hi Gerome,

    Thank you very much for your detailed explanation!

    That makes perfect sense — we now better understand that this might be a case of electrical-level compatibility rather than intended support. We will proceed with differential probing on the MDI lines to verify whether the signaling is indeed 100BASE-FX or 100BASE-TX, as you suggested.

    Thanks again for your help!

    Best regards,