• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 1 answer
  • Subscribers 32 subscribers
  • Views 326 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

DP83869HM: 100Mbps is working but 1Gbps is not working

DP83869HM: DP83869

sumathi m
Intellectual 905 points
Part Number: DP83869HM
Other Parts Discussed in Thread: DP83869, AM5748, SYSBIOS

Tool/software:

hi, we are using DP83869 in our design for a secured communication project.

we have issue with the device. the 1Gbps is not working for both copper and fiber interface.

below is the registers dump

DP83869 Ethernet PHY Register Dump
PHY Address : 00
/**********************************************************************
PHY REG 00 = 00001140
PHY REG 01 = 0000796d
PHY REG 02 = 00002000
PHY REG 03 = 0000a0f3
PHY REG 04 = 000001e1
PHY REG 05 = 0000cde1
PHY REG 06 = 0000006f
PHY REG 07 = 00002001
PHY REG 08 = 00004f0f
PHY REG 09 = 00000300
PHY REG 0a = 00003800
PHY REG 0b = 00000000
PHY REG 0c = 00000000
PHY REG 0d = 00000000
PHY REG 0e = 00000000
PHY REG 0f = 0000f000
PHY REG 10 = 00005048
PHY REG 11 = 0000bc02
PHY REG 12 = 00000000
PHY REG 13 = 00001c40
PHY REG 14 = 000029c7
PHY REG 15 = 00000000
PHY REG 16 = 00000000
PHY REG 17 = 00000040
PHY REG 18 = 00006150
PHY REG 19 = 00004444
PHY REG 1a = 00000002
PHY REG 1b = 00000000
PHY REG 1c = 00000000
PHY REG 1d = 00000000
PHY REG 1e = 00000012
PHY REG 1f = 00000000
PHY REG 25 = 0000cde1
PHY REG 2c = 00000000
PHY REG 2d = 00000000
PHY REG 2e = 00000000
PHY REG 31 = 0000ac02
PHY REG 32 = 00000000
PHY REG 33 = 00000000
PHY REG 34 = 000029c7
PHY REG 35 = 00000000
PHY REG 36 = 00000000
PHY REG 37 = 00000040
PHY REG 38 = 00006150
PHY REG 39 = 00004444
PHY REG 3a = 00000002
PHY REG 43 = 0000a0f3
PHY REG 4f = 0000f000
PHY REG 55 = 00000000
PHY REG 6e = 00000000
PHY REG 86 = 0000006d
PHY REG 134 = 000029c7
PHY REG 135 = 00000000
PHY REG 170 = 00005048
PHY REG 180 = 00001140
PHY REG 181 = 0000796d
PHY REG 182 = 00002000
PHY REG 183 = 0000a0f3
PHY REG 184 = 000001e1
PHY REG 185 = 0000cde1
PHY REG 190 = 00005048
PHY REG 191 = 0000ac02
PHY REG 192 = 00000000
PHY REG 193 = 00000000
PHY REG 194 = 000029c7
PHY REG 195 = 00000000
PHY REG 196 = 00000000
PHY REG 197 = 00000040
PHY REG 198 = 00006150
PHY REG 199 = 00004444
PHY REG 1a4 = 000001e1
PHY REG 1a5 = 0000cde1
PHY REG 1a6 = 0000006d
PHY REG 1df = 00000000
PHY REG 1e0 = 00001140
PHY REG 1ec = 00000000
PHY REG c00 = 00001140
PHY REG c01 = 0000796d
PHY REG c02 = 00002000
PHY REG c03 = 0000a0f3
PHY REG c04 = 000001e1
PHY REG c05 = 0000cde1
PHY REG c06 = 0000006d
PHY REG c07 = 00002001
PHY REG c08 = 00004f0f
PHY REG c10 = 00005048
PHY REG c18 = 00006150
PHY REG c19 = 00004444
/*************************************

this chip is interfaced to AM5748 RGMII 0 and 1 lanes.

Thanks,

Regards,

Sumathi

  • 0
    TI__Mastermind 22432 points

    Hi Sumathi,

    Thanks for sharing the register dump. Link is up while communication is failing, so I suspect the root cause is with RGMII delay settings:

    https://e2e.ti.com/support/interface-group/interface/f/interface-forum/1099945/faq-how-to-select-correct-rgmii-delay-mode-for-phy-and-mac

    Please refer to this FAQ and configure RGMII delay settings with registers 0x32[1:0] (enable shift/align) and 0x86[7:0] (change delay in shift mode).

    MAC signal timing requirements for Gigabit communication in RGMII are strict -  I recommend iterating through delay settings and capturing MAC signals to verify the timing requirements are satisfied.

    Thank you,

    Evan

  • 0
    Intellectual 905 points

    hi Evan,

    Thanks for the reply.

    we have already tried registers 0x32 and 0x86..but still there is no progress. we have bought DP83869EVB as well. in that also we are observing that the TRansmit from AM5748 at 1Gbps is happening but the receive is not ok. the AM5748 is not receiving the data. are we missing something in Processor side software drivers etc. pls let us know whether we can have teams call to discuss the issue.

    Regards,

    Sumathi

  • 0 in reply to sumathi m
    TI__Mastermind 22432 points

    Hi Sumathi,

    Can you share which values for 0x32 and 0x86 you have tested?

    I can also check schematic to confirm proper connections, but I suspect root cause is with delay setting.

    Please email me at e-mayhew@ti.com to schedule a call if needed.

    Thank you,

    Evan

  • 0 in reply to Evan Mayhew
    Intellectual 905 points

    hi Evan,

    Thank you for the support.

    below are the register values

    Register 0x86=0x77
    Register 0x32 = 0xEA

    Regards,

    Sumathi

  • 0 in reply to sumathi m
    TI__Mastermind 22432 points

    Hi Sumathi,

    If these are the only register values tested, I recommend iterating through other delay settings to confirm any difference in communication.

    Before changing these settings, can you confirm if communication is working at 10/100M? If lower speeds work, this isolates the issue to meeting RGMII timing requirements for Gigabit.

    In this case, please try these settings:

    0x32[1:0] = '11' (TX/RX align mode)

    0x32[1:0] = '00' (TX/RX shift mode)

    0x86[7:0] = 33h (TX/RX delay of 1ns in shift mode)

    0x86[7:0] = BBh (TX/RX delay of 3ns in shift mode)

    Thank you,

    Evan

  • 0 in reply to Evan Mayhew
    Intellectual 905 points

    hi Evan,

    Thank you for the support.

    below is the observation for 1Gbps,

    Board Side:

    Modified Iperf tool which would utilize both the CPU's.

    Performance mode is enabled in kernel instead of onDemand CPU mode.

     

    Host System Side :

    Ubuntu is used in console mode with modified iperf tool which will utilize all the available CPU cores.

     

     

    S.No

    Board

    Status

    Observation

    1

    001

    Transmit: 0% loss at 500Mbps

    Receive : 0% loss at 300Mbps

    At receiver, in few seconds (1 or 2 ) 0.02 to 0.0004% loss is seen

    2

    005

    Transmit: 0% loss at 500Mbps

    Receive : 0% loss at 300Mbps

    At receiver, in few seconds (1 or 2) 0.02 to 0.0004% loss is seen

    3

    003

    Transmit: 0% loss at 500Mbps

    Receive : 0% loss at 300Mbps

    At receiver,  in few seconds 0.02% loss is seen

    4

    004

    Transmit: 0% loss at 500Mbps

    Receive : 0% loss at 300Mbps

     

     

    Note:

    ​Above observations are taken across 9 runs by doing power cycle after 3 runs.

    Test interval is 100s.

     

    Summary:

    Ethernet Transmit and recevie is working upto 500Mbps and 300Mbps respectively with out any data loss. Above observations has to be verified on sysbios.

    Thanks,

    Regards,

    Sumathi

  • 0 in reply to sumathi m
    TI__Mastermind 22432 points

    Hi Sumathi,

    Thanks for confirming there is no data loss at lower speeds, this narrows the root cause to the RGMII delay settings.

    Please try the 0x32 / 0x86 settings mentioned above, and note any difference in throughput / errors.

    As the required delays are dependent on trace length, trial and error with the settings is typically required.

    Thank you,

    Evan

  • 0 in reply to Evan Mayhew
    Intellectual 905 points

    hi Evan, Thanks for the reply. we have already tried the settings..but its not helping. while i was trying to reply to you..its seems to be muted. kindly enable the reply.

    Regards,

    Sumathi

  • 0 in reply to sumathi m
    TI__Mastermind 22432 points

    Hi Sumathi,

    Strange, we can continue this discussion on email if reply on E2E is not working.

    Thank you,

    Evan

  • 0 in reply to Evan Mayhew
    Intellectual 905 points

    Thank you very much Evan!

    Regards,

    Sumathi

  • 0 in reply to sumathi m
    TI__Mastermind 22432 points

    Of course, closing this thread to handle over email.

    Thank you,

    Evan