• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 33 subscribers
  • Views 949 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.

DP83620 receives but won't send

Alvaro GM
Prodigy 120 points
Other Parts Discussed in Thread: DP83620

Hi,

I'm using a DP83620 in fiber mode. I've tried this out with several SFPs from different providers and the result is always the same.

I can receive messages sent to me (for example, UDP broadcast messages, as these do not require any ARP communication), but I can't get anything out of my board.

I've checked SFPs and cables and everything is alright. The only thing I can think of that may be wrong is the matching network between TX from the PHY and the SFP.

However, based on datasheet from Avago product HFBR-57E5APZ, which is quite nice, and that of DP83620, I can't see what I'm doing wrong. I've tried changing termination resistors, from single ended 50Ohms to LVPECL and basically everything that has crossed my mind.


LEDs do blink when trying to send, so it is my understanding that the DP83620 has proper communication with the FPGA it is attached to.

Any ideas?

  • 0
    TI__Mastermind 23675 points
    Hi Alvaro,

    Can you send me your schematic of the PHY and fiber transceiver?
    The DP83620 uses a current mode line driver and therefore must have capacitive isolation between the PHY and fiber transceiver.
    Most fiber transceivers use PECL driver so this might be why you are having issues.

    Please see page 63 in the DP83620 datasheet for a reference.

    Kind regards,
    Ross
  • 0 in reply to Ross Pimentel
    Prodigy 120 points

    DP83620-sch.pdf

    Hi,

    I've attached PDF file with schematic.

    At the center of the page there are the schematics for both TX and RX lines. RX works as is, while TX does not.

    We've also tried different configurations on both sides of the line with the following results:

    R18=NO MOUNT, R20=NO MOUNT, R24=NO MOUNT, R28=NO MOUNT
    R17=NO MOUNT, R19=NO MOUNT, R23=NO MOUNT, R27=NO MOUNT
    C22=0hm, C23=0ohm

    No link detected

    With all the following configurations the same thing happens: Link detected, receives data but can't send. TX led, however, flashes with every packet that I try to send.

    R18=130, R20=80, R24=130, R28=80
    R17=NO MOUNT, R19=NO MOUNT, R23=NO MOUNT, R27=NO MOUNT
    C22=0hm, C23=0ohm

    R18=130, R20=80, R24=130, R28=80
    R17=130, R19=80, R23=130, R27=80
    C22=0hm, C23=0ohm

    R18=130, R20=80, R24=130, R28=80
    R17=130, R19=80, R23=130, R27=80
    C22=100nF, C23=100nF

    R18=130, R20=80, R24=130, R28=80
    R17=50, R19=NO MOUNT, R23=50, R27=NO MOUNT
    C22=100nF, C23=100nF

    R18=130, R20=80, R24=130, R28=80
    R17=50, R19=NO MOUNT, R23=50, R27=NO MOUNT
    C22=0ohm, C23=0ohm

    This covers all the options we've seen on DP83620 datasheet page 63, suggestions made by Avago for its SFP, and basically every other option that we were able to implement with the same layout.

    Thanks!

  • 0 in reply to Alvaro GM
    TI__Mastermind 23675 points
    Hi Alvaro,

    Thank you for the details above.
    This is not meant to be insulting, but have you made sure that the TX Disable pin on the SFP is not HIGH or Floating?
    In the transceiver datasheet, this pin needs to be tied to ground for operation.

    Also, can you see NRZI signaling on the TX pins when you probe with an oscilloscope? Can you send me the image?
    Is pin 9 connected to ground?

    Kind regards,
    Ross
  • 0 in reply to Ross Pimentel
    Prodigy 120 points

    Hi Ross

    Don't worry, it wouldn't be the first time a simple mistake is the cause of very strange behaviours. However, I can confirm that TX Disable is tied to LOW, so this isn't the issue this time.

    In fact, onto very good news, when trying to capture the data you asked me for, we realized that PIN TXD3 wasn't making proper contact! This board wasn't professionally assembled and it wasn't perfect. After several minutes worth of work under the microscope and solder we've been able to make it contact -barely-, but enough to make it finally work!

    Our final configuration for analog configuration is

    R18=130, R20=80, R24=130, R28=80
    R17=130, R19=80, R23=130, R27=80
    C22=100nF, C23=100nF

    Thanks a lot!

  • 0 in reply to Alvaro GM
    TI__Mastermind 23675 points

    Hi Alvaro,

    This is great news! I am glad that you found the issue. 

    Kind regards,

    Ross