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.

DP83869HM: Circuit connects with Commercial media converter, not with another circuit using DP83869HM

Part Number: DP83869HM
Other Parts Discussed in Thread: DP83869

Tool/software:

Hi

We have designed a circuit which includes a DP83869 operating in media converter mode.

The SFP is an FS.com generic SFP module

We are configuring it using an external microcontroller communicating over the MDIO interface

When the optical side is connected to a commercial media converter, the link comes up OK and we can send data from one PC to another through the link.
When we connect two of our circuits together, the link fails to come up 99 times out of 100.

Our feeling is that this is an autonegotiation issue but we'd appreciate any advice or assistance to help us resolve the issue.

We have run through the troubleshooting guide and the schematic review and didn't see any major errors (except the signal detect line, which we failed to pull high,)

We do have a microcontroller connected to the MDIO interface and use this for initial setup.Single-Board-WOLF-Media_Converter.pdf

  • Hello,

    Could you please provide a block diagram of the working and non-working condition? This can provide better clarity on what link is failing. There are two links for Media Converter Mode; copper and fiber. When one of these fail, the entire signal chain falters.

    Sincerely,

    Gerome


  • Thank you Gerome.

    I have attached an image showing the working and non-working connections.

    In addition, a link between two 10GTek media converters works most reliably and has quickest time to re-establish link after the optical side is broken.




  • Hello,

    Thank you for clarifying that the FX (not copper) link on the 2x Media Converter signal chain is failing here. I would like to establish a baseline understanding of your setup. Could you please comment on the following questions:

    1) Could you please provide a register dump of the following; 0x0-0x1F, 0x6E, 0x6F, 0xC00-0xC01, 0xC10, 0xC30? Please note that any registers not within [0x0, 0x1F] block are extended registers and thus require extended access to read appropriately.

    2) Do you have EVMs available, or is this only observed on your custom design? If possible, I would like to evaluate this on the DP83869EVMs to also establish correlation to behavior that can be recreated in TI lab.

    3) Have you read SNLA479? This is a recently released, streamlined debugging app note on FX debugging which I believe can help.

    Sincerely,

    Gerome

  • Thank you Gerome

    We do not have EVM modules.

    We have further noticed that the behaviour of the two boards we have had made for testing is different, indicating there may be further issues. One board establishes a connection with the commercial media converter more reliably and is quicker for LED[0..2] to illuminate while establishing. Burn boards have the same firmware.

    I will get you the register dumps, but it won't be for a couple of days.

  • Hi Wez,

    Thanks for the update. Looking forward to your register dumps.

    Sincerely,

    Gerome