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: Destination host unreachable/Request timed Out

Part Number: DP83869HM
Other Parts Discussed in Thread: DP83869

Hello,

We are using DP83869HM for media conversion between 100BaseTx and 100BaseFx. We are trying to establish LAN connection between source and destination and vice versa. The hardware connection is as below. Please refer the block diagram.

Source (PC-A) -> DP83869HM (100BaseTx) -> DP83869HM (100BaseFx) -> Destination (PC-B)

 Destination (PC-B) -> DP83869HM (100BaseTx) -> DP8369HM (100BaseFx) -> Source (PC-A) 

We tried pinging from the source and monitor through command prompt. Initially we were getting ping response frequently (Ref image 1). But later we are only getting "Destination host unreachable/Request time out" response (Ref Image 2). 

Please let us know what the reason for this could be and how can we debug this scenario?

Regards,

Rajesh

  • Hi Rajesh,

    Please send me a register dump from the PHY including register 0x00-0x1f, 0xC00-0xC08, and 0x6E. I can have a look at your schematic as well if you can share it.

    Thanks,

    David

  • Hello David,

    Thanks for reply!!!

    Please find the register dump from the PHY including register 0x00-0x1f, 0xC00-0xC08, and 0x6E. 

     

    Schematic:

    1321.Ethernet Section.pdf

    Regards,

    Rajesh

  • Hi Rajesh,

    Register 0x1 shows link is down. Has it ever been up? If so, please send me a register dump in this case. 

    Thanks,

    David

  • Hello David,

    Thanks for reply!!!

    Yes, it has been up, whenever I connect board to board test that time link is up, below is result

    0x1:

    Regards,

    Rajesh

  • Hi Rajesh,

    Please send me the same register dump in the link up scenario.

    Thanks,

    David

  • Hello David,

    Thanks for reply!!!

    PFA of same register dump in the link up scenario. I did the compression of (non-operational and operational) those register values are the changes.

    1.0001 = 796D

    2.0005 = CD81

    3.0006 = 006F

    4.0008 = 4806

    5.0011 = 7C02

    6.0013 = 1C44

    0x00-0x1f:

     

    00xC00-0xC08, and 0x6E:

    Regards

    Rajesh 

  • Hi Rajesh,

    I will go through the schematic and register dump and get back to you next week. 

    Thanks,

    David

  • Hello David,

    Any update on Schematic and register dump?

    Regards,

    Rajesh R

  • Hi Rajesh,

    It looks like there is a difference in link partner between the two cases, or that a cable is not installed at all in the fail case. Register 0x6 shows us the link partner is not auto-negotiation able. Register 0x8 shows there is no acknowledgement from the link partner. The register dump looks at if no link partner is connected at all. Is anything about the setup different between these two cases?

    I see there are several comments on the schematic. Were these changes implemented? Please send me an updated schematic if so.

    I noticed in that the magnetics are not what we recommend in the datasheet. The center taps should not be connected nor tied to VDDIO, and the capacitor values are different. This is likely contributing to the issue. 

    Thanks,

    David

  • Hello David,

    Thanks for replying!!!

    1.It looks like there is a difference in link partner between the two cases, or that a cable is not installed at all in the fail case. Register 0x6 shows us the link partner is not auto-negotiation able. Register 0x8 shows there is no acknowledgement from the link partner. The register dump looks at if no link partner is connected at all. Is anything about the setup different between these two cases?

    ANS:  The configuration is set to forced 100mbps media converter as per the data sheet. Therefore, auto negotiation was not selected as per datasheet. 

    ANS:  Link partner does come up at certain times when we had successful ping. This means setup is okay. 

    2. I see there are several comments on the schematic. Were these changes implemented? Please send me an updated schematic if so.

    ANS: Below is the Updated schematic 

    SCHEMATIC1 _ 07-PHY SECTION.pdf

    3. I noticed in that the magnetics are not what we recommend in the datasheet. The center taps should not be connected nor tied to VDDIO, and the capacitor values are different. This is likely contributing to the issue. 

    ANS: Already in TI forums one person using same RJ45 connector (7499111121A) with 1000base-TX to 1000base FX it's working fine in our schematic we used 100baseTX to 100base FX with same RJ45 connector (7499111121A). Below is the 1000base-TX to 1000base FX sch. FYR Media_Converter_1000BaseTX to 1000BaseFX.sch.pdf

    Regards,

    Rajesh R

  • Hi Rajesh,

    All the register dumps you have sent show auto-negotiation is enabled in the DP83869 (register 0x0000 has value 0x3100). If you force the speed on the link partner side while leaving auto negotiation on on the DP83869 side, they will not link up. Please leave auto negotiation on on both sides of the link. 

    What E2E thread are you seeing that approves the 7499111121A magnetics? We do not recommend center tap connected magnetics, as shown in figure 111 of the datasheet.

    Thanks,

    David

  • Hello David,

    Thanks for replying!!!

     In current design we used (Part: no: 7499111121A) RJ45 connector. Please confirm this connector is an issue? and below is the connector datasheet & image.

    7242.7499111121A_RJ45_Connector_datasheet.pdf

    Kindly confirm ASAP

    Regards,

    Rajesh 

  • Hi Rajesh,

    We do not recommend center tap connected magnetics as shown in figure 111 of the datasheet, so I cannot recommend the 7499111121A part. This may be the cause of link issues you are seeing. Please use one that meets the requirements listed in section 10.2.1.2.2 and figure 111.

    Thanks,

    David