• State
  • Locked Locked
  • Replies 42 replies
  • Subscribers 33 subscribers
  • Views 953 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.

DP83869HM: Wrong transmission with media converter mode (1000BASE-T to 1000BASE-X)

MARMONIER Cedric
Prodigy 20 points
Part Number: DP83869HM
Other Parts Discussed in Thread: DP83869

Tool/software:

Hello,

I am using the DP83869HM in media converter mode (1000BASE-T to 1000BASE-X).

Here is my application topology:

Wi-Fi Router <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X ---> DP83869HM <--- 1000BASE-T ---> Cisco Catalyst 3560-CX switch

On the switch side (Catalyst 3560-CX), I observe increasing "input errors" without any CRC errors or other error types (e.g., frame, overrun, ignored, etc.).

If I connect the Wi-Fi router **directly** to the switch, there are **no input errors**.

If I insert an **intermediate switch** between the Wi-Fi router and the first DP83869HM, the issue also disappears.

This makes me suspect a possible problem with **interpacket gap (IPG)** or **frame alignment** across the DP83869 media conversion.

**Questions:**
- Is there a way to validate or monitor IPG timing through the DP83869HM registers?
- Could improper IPG timing on the 1000BASE-T side propagate malformed frames through to the 1000BASE-X interface?
- Is there a register setting to **drop** or **filter** these frames before they reach the switch?
- Do you have any known compatibility issues with certain Wi-Fi routers or MACs?


**Exemple of the catalyst switch counter values:**
Switch#show interfaces gigabitEthernet0/1
GigabitEthernet0/1 is up, line protocol is up (connected)
  Hardware is Gigabit Ethernet, address is 247e.12ae.9a81 (bia 247e.12ae.9a81)
  MTU 1500 bytes, BW 1000000 Kbit/sec, DLY 10 usec,
     reliability 255/255, txload 1/255, rxload 1/255
  Encapsulation ARPA, loopback not set
  Keepalive set (10 sec)
  Full-duplex, 1000Mb/s, media type is 10/100/1000BaseTX
  input flow-control is off, output flow-control is unsupported
  ARP type: ARPA, ARP Timeout 04:00:00
  Last input 00:00:54, output 00:00:00, output hang never
  Last clearing of "show interface" counters 00:06:25
  Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
  Queueing strategy: fifo
  Output queue: 0/40 (size/max)
  5 minute input rate 1000 bits/sec, 1 packets/sec
  5 minute output rate 16000 bits/sec, 21 packets/sec
     819 packets input, 139032 bytes, 0 no buffer
     Received 624 broadcasts (210 multicasts)
     0 runts, 0 giants, 0 throttles
     23 input errors, 1 CRC, 0 frame, 0 overrun, 0 ignored
     0 watchdog, 210 multicast, 0 pause input
     0 input packets with dribble condition detected
     8024 packets output, 725317 bytes, 0 underruns
     0 output errors, 0 collisions, 0 interface resets
     0 unknown protocol drops
     0 babbles, 0 late collision, 0 deferred
     0 lost carrier, 0 no carrier, 0 pause output
     0 output buffer failures, 0 output buffers swapped out
Switch#

Thanks in advance for your support.

  • 0
    TI__Genius 14026 points

    Hi, 
    Could we look at the schematic of the DP83869HM connections?
    If possible, could we also get register dump from 0x00 to 0x1F and 0xC00 to 0xC30 on both PHYs when this issue happens?
    Also, are the fiber and copper link forced to the same speed? If there is a speed mismatch, this can cause packet loss.

    To answer your question:

    MARMONIER Cedric said:
    - Is there a way to validate or monitor IPG timing through the DP83869HM registers?

    There is none that I am aware of. 

    MARMONIER Cedric said:
    - Could improper IPG timing on the 1000BASE-T side propagate malformed frames through to the 1000BASE-X interface?

    Improper IPG timing will propagate malformed frames through the media converter. 


    MARMONIER Cedric said:
    - Is there a register setting to **drop** or **filter** these frames before they reach the switch?

    There is none that I am aware of. 


    MARMONIER Cedric said:
    Do you have any known compatibility issues with certain Wi-Fi routers or MACs?

    Customers have not yet reported any compatibility issues. 



    Please let me know. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    Schematic is basic : there are RJ45/Transformer in 1000BASE-TX side and SFP in 1000BASE-X side.


    ********************************************
    DP83869HM connected to wifi router:

    Dump registers 0x00 to 0xFF:
        |  .0  |  .1  |  .2  |  .3  |  .4  |  .5  |  .6  |  .7  |  .8  |  .9  |  .A  |  .B  |  .C  |  .D  |  .E  |  .F
        +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     0. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     1. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0018 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     2. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     3. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0018 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     4. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     5. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0018 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     6. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     7. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0019 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     8. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     9. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0019 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     A. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     B. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0019 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     C. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     D. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0019 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     E. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     F. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0019 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000

    Dump registers 0x130 to 0x1EF:
         | ..0  | ..1  | ..2  | ..3  | ..4  | ..5  | ..6  | ..7  | ..8  | ..9  | ..A  | ..B  | ..C  | ..D  | ..E  | ..F
         +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     13. | 0200 | 4A23 | 6520 | 4820 | 1000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     14. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     15. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     16. | 0000 | 000C | 080C | 0110 | 4489 | 00C3 | E115 | 0D4A | 0A96 | 0000 | 2A4A | 818A | 0800 | 0800 | 318C | 0095
     17. | 0C0F | 0000 | 0000 | 0000 | 000F | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     18. | 0752 | C850 | 5326 | A01E | E976 | 19CF | 31D7 | 0D9B | 0CA9 | 0014 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     19. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1A. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | F020 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1B. | 4058 | 0602 | CAC2 | 001E | 1C04 | 0016 | 0000 | 0000 | 0010 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1C. | 5C29 | A829 | 7E9E | F3C6 | 01C2 | 1C70 | 530B | F1C1 | 01C1 | B1C1 | 1600 | 1600 | 1000 | 1241 | 0555 | 1708
     1D. | 1313 | 1313 | 0904 | 0000 | 0000 | F500 | FFFF | FFFF | 0800 | 30D3 | F423 | 0044 | F3DF | 0000 | 0000 | 0044
     1E. | 417A | 0002 | 0025 | 0018 | 88F7 | 0000 | 0080 | 0012 | 0920 | 036D | 0000 | 4830 | 1FF5 | 0000 | 0035 | 0000

    Dump registers 0xC00 to 0xC3F:
         | ..0  | ..1  | ..2  | ..3  | ..4  | ..5  | ..6  | ..7  | ..8  | ..9  | ..A  | ..B  | ..C  | ..D  | ..E  | ..F
         +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     C0. | 1140 | 616D | 2000 | A0F1 | 0020 | 4020 | 0005 | 2001 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     C1. | 3148 | 0000 | 10B4 | 0400 | 0100 | 652A | 0029 | 0002 | 01FF | 0000 | 0000 | 0000 | 1000 | 0000 | 00DB | 0000
     C2. | 250A | 0000 | F423 | 0000 | 0044 | 0000 | F3DF | 0000 | 70D3 | 0000 | 0812 | 0000 | 4844 | 0000 | F423 | 0000
     C3. | 3056 | 0000 | 0001 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000


     

    ********************************************
    DP83869HM connected to Catalyst switch:

    Dump registers 0x00 to 0xFF:
        |  .0  |  .1  |  .2  |  .3  |  .4  |  .5  |  .6  |  .7  |  .8  |  .9  |  .A  |  .B  |  .C  |  .D  |  .E  |  .F
        +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     0. | 1140 | 796D | 2000 | A0F1 | 0001 | C001 | 006D | 2001 | 4806 | 0300 | 3800 | 0000 | 0000 | 401F | 0000 | F000
     1. | 5048 | AC02 | 0000 | 0000 | 29C7 | 0000 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     2. | 1140 | 796D | 2000 | A0F1 | 0001 | C001 | 006D | 2001 | 4806 | 0300 | 3800 | 0000 | 0000 | 401F | 0000 | F000
     3. | 5048 | AC02 | 0000 | 0000 | 29C7 | 0000 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     4. | 1140 | 796D | 2000 | A0F1 | 0001 | C001 | 006D | 2001 | 4806 | 0300 | 3800 | 0000 | 0000 | 401F | 0000 | F000
     5. | 5048 | AC02 | 0000 | 0000 | 29C7 | 0000 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     6. | 1140 | 796D | 2000 | A0F1 | 0001 | C001 | 006D | 2001 | 4806 | 0300 | 3800 | 0000 | 0000 | 401F | 0000 | F000
     7. | 5048 | AC02 | 0000 | 0000 | 29C7 | 0000 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     8. | 1140 | 796D | 2000 | A0F1 | 0001 | C001 | 006D | 2001 | 4806 | 0300 | 3800 | 0000 | 0000 | 401F | 0000 | F000
     9. | 5048 | AC02 | 0000 | 0000 | 29C7 | 0000 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     A. | 1140 | 796D | 2000 | A0F1 | 0001 | C001 | 006D | 2001 | 4806 | 0300 | 3800 | 0000 | 0000 | 401F | 0000 | F000
     B. | 5048 | AC02 | 0000 | 0000 | 29C7 | 0000 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     C. | 1140 | 796D | 2000 | A0F1 | 0001 | C001 | 006D | 2001 | 4806 | 0300 | 3800 | 0000 | 0000 | 401F | 0000 | F000
     D. | 5048 | AC02 | 0000 | 0000 | 29C7 | 0000 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     E. | 1140 | 796D | 2000 | A0F1 | 0001 | C001 | 006D | 2001 | 4806 | 0300 | 3800 | 0000 | 0000 | 401F | 0000 | F000
     F. | 5048 | AC02 | 0000 | 0000 | 29C7 | 0000 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000

    Dump registers 0x130 to 0x1EF:
         | ..0  | ..1  | ..2  | ..3  | ..4  | ..5  | ..6  | ..7  | ..8  | ..9  | ..A  | ..B  | ..C  | ..D  | ..E  | ..F
         +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     13. | 0200 | 4A23 | 6520 | 4820 | 1000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     14. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     15. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     16. | 0000 | 000C | 080C | 0110 | 4489 | 00C3 | E115 | 0D4A | 0A96 | 0000 | 2A4A | 818A | 0800 | 0800 | 318C | 0095
     17. | 0C0F | 0000 | 0000 | 0000 | 000F | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     18. | 0752 | C850 | 5326 | A01E | E976 | 19CF | 31D7 | 0D9B | 0CA9 | 0014 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     19. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1A. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | F020 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1B. | 4058 | 0602 | CAC2 | 001E | 1C04 | 0016 | 0000 | 0000 | 0010 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1C. | 5C29 | A829 | 7E9E | F3C6 | 01C2 | 1C70 | 530B | F1C1 | 01C1 | B1C1 | 1600 | 1600 | 1000 | 1241 | 0555 | 1708
     1D. | 1515 | 1515 | 0904 | 0000 | 0000 | F500 | FFFF | FFFF | 0800 | 30D3 | F423 | 0044 | F3DF | 0000 | 0000 | 0044
     1E. | 417A | 0002 | 0025 | 0018 | 88F7 | 0000 | 0080 | 0012 | 0920 | 036D | 0000 | 4830 | 1FF5 | 0000 | 0035 | 0000

    Dump registers 0xC00 to 0xC3F:
         | ..0  | ..1  | ..2  | ..3  | ..4  | ..5  | ..6  | ..7  | ..8  | ..9  | ..A  | ..B  | ..C  | ..D  | ..E  | ..F
         +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     C0. | 1140 | 614D | 2000 | A0F1 | 0020 | 0000 | 0004 | 2001 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     C1. | 3148 | 0000 | 10B4 | 0400 | 0100 | 652A | 0029 | 0002 | 01FF | 0000 | 0000 | 0000 | 1000 | 0000 | 00DB | 0000
     C2. | 250A | 0000 | F423 | 0000 | 0044 | 0000 | F3DF | 0000 | 70D3 | 0000 | 0812 | 0000 | 4844 | 0000 | F423 | 0000
     C3. | 3056 | 0000 | 0001 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000

     
     





  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    Could you check the register dump on 0x10 to 0x1F again? They seem to be identical to 0x00 to 0x0F register dump on both PHYs, and they cannot be identical. I am specifically looking for register data of 0x0015 which tracks any receive error count  as I try to isolate at which point the error is happening. 

    MARMONIER Cedric said:
    Schematic is basic : there are RJ45/Transformer in 1000BASE-TX side and SFP in 1000BASE-X side.

    We would like to ensure that all hardware configurations are according to the datasheet recommendation. If you would like to not post it on the public domain, you can accept my friend request and private message the schematic. 

    In addition, could you do a reverse loopback on the second PHY (PHY closer to the switch) at the copper side to see if there are any input errors at the wifi router from the returned packets? Reverse loopback will send the incoming packets back so we can verify where the error is happening. 

    Reverse loopback can be enabled by enabling 8h on the register 16h bit 5-2. In addition, auto-negotiation has to be disabled and forced speed on the copper side for reverse loopback to work. 




    Please let me know. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    Hi,

    The schematic :



    Note: I only have one board with debug mode. When I did the dumps, I did them one at a time (I swapped boards between the two dumps).

    I will redo dumps with more all registers and loopback mode...

    Why errors disappears when inserted a switch beetween Wi-Fi router and DP83869HM?

  • 0 in reply to J
    Prodigy 20 points

    My dump functions was wrong: I updated my second message with good dump values.

    To disable auteg, I must set reg 00h to 0140h?

  • 0 in reply to J
    Prodigy 20 points

    With reverse loopback there is no error:

    DP83869HM <--- 1000BASE-T ---> Cisco Catalyst 3560-CX switch <--- 1000BASE-T ---> Ethernet tester

    No erreur detected with packet sizes of 64 and 1518 (by default, Catalyst drop jumbo frames).

    (I haven't deactivated autoneg and it works: when I read the datasheet, it doesn't seem necessary to deactivate it with the Far-end loopback)

    I tested with Jumbo frame without Catalsyt:

    DP83869HM <--- 1000BASE-T ---> Ethernet tester

    No erreur detected with packet size of 10000.

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    Thank you for your detailed response. 

    MARMONIER Cedric said:
    If I insert an **intermediate switch** between the Wi-Fi router and the first DP83869HM, the issue also disappears.

    In this comment, did you mean the following?

    Wi-Fi Router <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X -----> Cisco Catalyst Switch


    Based on the register dump, it looks like there is a receive error on the 869 connected to the wifi router (15h). 
    I am looking to see if this error counter increments with the error on the copper side or the fiber side. Assuming the data goes from wifi router to the switch all the time, it seems like input error may be propagating from the copper side of the PHY connected to the wifi router. I will update you as soon as I get this data. 

    If you can, could you do a reverse loopback with both PHYs on since we know it works with one PHY?

    I am imagining a setup as such:
    Reverse loopback here
     DP83869HM <--- 1000BASE-X ---> DP83869HM <--- 1000BASE-T ---> Ethernet tester


    Lastly, does this issue occur when the PHYs are forced to be at 1000M on both copper and fiber? This can be done by writing 0x0140 to both 0h and C00h. 

    I apologize for the lack of guidance at the moment and please be patient as we figure out what is going on here. 

    Please let me know and I will keep you updated.

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    I have no error with those configurations :

    Wi-Fi Router <--- 1000BASE-T ---> Ethernet switch <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X -----> Cisco Catalyst Switch

    Ethernet tester (TX/RX) <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X -----> Cisco Catalyst Switch <--- 1000BASE-T ---> Ethernet tester (loopback) 

    Ethernet tester (TX/RX) <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X ----->  DP83869HM <--- 1000BASE-T ---> Cisco Catalyst Switch <--- 1000BASE-T ---> Ethernet tester (loopback) 

    Wi-Fi Router <--- 1000BASE-T ---> Cisco Catalyst Switch



    I have errors with those configurations :

    Wi-Fi Router <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X -----> Cisco Catalyst Switch

    Wi-Fi Router <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X -----> DP83869HM  <--- 1000BASE-T ---> Cisco Catalyst Switch

    I see error with an other switch that Catalyst.


    the issue occur when the PHYs are forced to be at 1000M on both copper and fiber.

    Wifi router is CISCO CW9166I-E.

    Switch error couters are InBadOctets, InFCSErr, InFragments.

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi,

    I discussed with the team and for media conversion autonegotiation has to be ON but only advertise 1Gbps on the copper side. Therefore, it is expected to see some errors when the copper is in forced mode due to compatibility issue with the link partner when 1G is forced. The default protocol for 1G Ethernet dictates that auto-negotiation is ON for all PHYs so forcing speed can cause issues in some link partners. 


    Based on the various configurations you have also tried, it also looks like this can be the compatibility issue between the router on DP83869HM. 

    Did the setup below give packet errors even when the auto-negotiation was on?

    Wi-Fi Router <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X -----> Cisco Catalyst Switch


    I am still trying to recreate the setup to see if the error goes up on copper or fiber side. I am aiming to have an update to you by tomorrow. 

    Please let me know and I will keep you updated. 

    Best,
    J

  • 0 in reply to J
    TI__Genius 14026 points

    Hi, 

    We see that the error counter goes up when there are symbol errors on the copper side. It seems like there's a connection issue between the DP83869HM and the wi-fi router and such error propagates to the Catalyst switch. 

    Could you see if the PHY only advertises 1000M on the auto-negotiation, there are still packet errors?

    Please let me know. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    Hi,

    I tryed to follow section 7.4.8.5 of https://www.ti.com/lit/ds/symlink/dp83869hm.pdf and I have the same result.

    Wifi router is CW9166E-I and there is the same problem.

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    I discussed with the team. 
    I am confident that this connection is the issue. 
    Could you do the PHY's on-chip reverse loopback mode to see if the router sees any packet errors? 

    Wi-Fi Router <--- 1000BASE-T ---> DP83869HM

    In addition, could you run your setup with after running these two scripts? Note that the first is the register address and the second is the write value. 

    These scripts should help with the link margin. 

    Lastly, if you use the other PHY to do the media conversion in the setup below, are there still errors? I am unsure if you have tried changing the media converting PHY. 

    Wi-Fi Router <--- 1000BASE-T ---> DP83869HM  (Replaced with the other PHY) <--- 1000BASE-X -----> Cisco Catalyst Switch

    Please let me know. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    I don't have access to management interface of Wifi router (the router mangement is available only with a remote service and I don't have access to it) : I can't do loopback test.

    You want I try the script in normal condition (without loopback mode)?

    The problem appears when I use an other board with a DP83869HM. I don't have a problem when I use an another media converter (but I don't konw if inside there is a switch component or a PHY component).

  • 0 in reply to J
    Prodigy 20 points

    There are always errors in using scripts and this configuration : 
    Wi-Fi Router <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X -----> Cisco Catalyst Switch

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    MARMONIER Cedric said:
    I don't have access to management interface of Wifi router (the router mangement is available only with a remote service and I don't have access to it) : I can't do loopback test.

    Understood. 

    MARMONIER Cedric said:
    There are always errors in using scripts and this configuration : 
    Wi-Fi Router <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-X -----> Cisco Catalyst Switch


    Has the number of errors on the switch decreased using the script? 
    Also, what is the cable length between the router and DP83869HM?

    In addition, could you send me the datasheet of the transformer to ensure that the transformer is within the recommended specification? I tried searching with the chip number on the schematic, but I couldn't find anything on Google. 

    Please let me know. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    Hi,

    Has the number of errors on the switch decreased using the script? 

    No.

    Also, what is the cable length between the router and DP83869HM?

    I tried tests with some cables (1m to 2m).

    The transformer : https://www.mouser.fr/ProductDetail/Bourns/SM51625EL?qs=1Fv0GIFcFssHdQX2RAfKzg%3D%3D

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    Thank you for the transformer information. I looked at the specification for the transformer and it does not fit our recommended specification. Typically, we recommend transformer within 2% tolerance but your transformer has 5% tolerance. 

    We recommend transformers like HX5008NL which has 1% tolerance. 

    Would you be able to replace the transformer to the recommended specification?
    Below is the recommended specification. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    Hi,

    It works the same way with a 2% transformer.

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    If that is the case, could you write 0x2053 to register 53h? This will lower the minimum IPG length threshold needed by the PHY. 

    Please let me know. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    Hi,

    With 0x2053 to register 53h: it works the same way

  • 0 in reply to MARMONIER Cedric
    Prodigy 20 points

    I did some tests : 

    Wi-Fi Router <--- 1000BASE-T ---> PoE injector (for Wi-Fi Router) <--- 1000BASE-T ---> SFP Copper to SGMII  <---> DP83869HM <--- 1000BASE-T -----> Cisco Catalyst Switch
    => There is no error in Catalyst counters (but I don't kow if there are dropped packets)


    Wi-Fi Router <--- 1000BASE-T ---> PoE injector (for Wi-Fi Router) <--- 1000BASE-T ---> DP83869HM <---> SFP SGMII to Copper <--- 1000BASE-T -----> Cisco Catalyst Switch
    => It is the same taht the previous test but in inverting input and output ports for DP83869HM. 
    => There are errors in Catalyst counters

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    53h register lowers the minimum IPG length threshold needed by the PHY. As this does not work, I am unsure what the issue is. 

    MARMONIER Cedric said:
    Wi-Fi Router <--- 1000BASE-T ---> PoE injector (for Wi-Fi Router) <--- 1000BASE-T ---> SFP Copper to SGMII  <---> DP83869HM <--- 1000BASE-T -----> Cisco Catalyst Switch
    => There is no error in Catalyst counters (but I don't kow if there are dropped packets)

    Was this test done with the register configuration I mentioned, 0x2053 to register 53h? If so, could this be a workaround for you?

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    The test was done with the configuration of 53h register. It can't be a workaround, I need to do 1000BASE-T to 1000BASE-X conversion. I've done these tests just to show different behaviours and perhaps find the cause.

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    I understand that this cannot be a workaround. 
    Could you remove ESD diodes for the PHY that connects to the wifi router to see if there are still any errors?
    In addition, is there a way to lower the throughput on the wifi router?
    Have you tried testing with other wifi router?

    I also have a test to propose based on the discussion with the team:

    Could we set the PHY in RGMII to copper mode and then do

    WIFI Router <---- 1000Base-T ----> DP83869HM in RGMII to Copper Mode

    and then do a reverse loopback test?
    In this test, could you read the register 15h so we can see if this is the PHY's problem in media converter mode OR if this is a connection issue between the PHY and the router. 

    Please let me know and keep me updated. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

     I have same results when I remove ESD diodes.

    I don't have error in 100Mbps mode :
    Wi-Fi Router <--- 100BASE-TX ---> DP83869HM <--- 100BASE-FX -----> Cisco Catalyst Switch


  • 0 in reply to J
    Prodigy 20 points

    For your information, the Ethernet port of the Wi-Fi terminal supports 100M/1000M/2.5G/5G speeds.
    Could the fact that it supports 2.5G and 5G cause any issues?
    Have you ever tested this type of Ethernet interface with the DP83869HM?

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    We have not yet had a case in which DP83869 has a problem with Ethernet switch that supports multi-gig Ethernet. 

    J said:
    Could we set the PHY in RGMII to copper mode and then do

    WIFI Router <---- 1000Base-T ----> DP83869HM in RGMII to Copper Mode

    and then do a reverse loopback test?
    In this test, could you read the register 15h so we can see if this is the PHY's problem in media converter mode OR if this is a connection issue between the PHY and the router. 

    Were you able to do this test also?

    In addition, after further discussion, could you try using a longer cable (10 to 50m) to see if the number of errors decrease?

    Lastly, I noticed that the register dump for 0x00 to 0x1F and 0x20 to 0x3F are identical. 

    MARMONIER Cedric said:
    Dump registers 0x00 to 0xFF:
        |  .0  |  .1  |  .2  |  .3  |  .4  |  .5  |  .6  |  .7  |  .8  |  .9  |  .A  |  .B  |  .C  |  .D  |  .E  |  .F
        +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     0. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     1. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0018 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000
     2. | 1140 | 796D | 2000 | A0F1 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     3. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0018 | 0000 | 0040 | 6B00 | 4000 | 0004 | 0000 | 0000 | 0000 | 0012 | 0000

    Based on this, I was wondering if you are accessing the extended registers properly. If not, this would mean that any register writes we have done in the past to fix this issue would not have been applied properly. Could you check if the extended registers are accessed correctly? Extended registers are accessed by below:



    Please keep me updated. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    Hi,

    My read and write functions in extended mode are correct; however, I thought they were supposed to be used for addresses above 0xFF and not 0x1F. Therefore, all the read and write actions I performed earlier in the address range from 0x20 to 0xFF are incorrect. I will correct this issue.

    Can you provide me with the exact initialization sequence for the component, along with the recommended corrections I should apply, to make it work in 1000BASE-X media converter mode, so I can redo the test?

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    HI, 

    Could you retry running your system with this script? This script is meant to improve the link margin.


    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    Before to run the script, can you confirm I do:

    // Set Operation Mode to 1000Base-T to 1000Base-X
    01DF   0044

    // Reset BMCR
    0000   1140

    // ***********
    // "Advertise 1000Base-T ability" and "Disable Advertise "100Base-T ability"
    0009   0300
    0014   FFFF
    // ***********
    // Disable Advertise "100Base-T ability"
    0004   0001
    // ***********

    // Reset PHY_CONTROL
    0010   5048

     // Reset FX_CTRL
    0C00   1140

    // Software Reset
    001F   4000

  • 0 in reply to J
    Prodigy 20 points

    For register 0053h, value to set isn't 2053h?

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 
    Sorry for the delayed response. 
    Yes, you are correct on the script above. There is no need to disable 100Base-T advertise, but it wouldn't matter. 
    For 53h, feel free to use either 2054h or 2053h. 2053h is the minimum IPG the device can tolerate. We would not be able to go lower than 2053h since that is known to not work. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    Hi,

    With the script, DP83869HM doesn't link for the fiber side.

    Dump register without the script (there are errors):

    Dump registers 0x00 to 0xFF:
        |  .0  |  .1  |  .2  |  .3  |  .4  |  .5  |  .6  |  .7  |  .8  |  .9  |  .A  |  .B  |  .C  |  .D  |  .E  |  .F
        +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     0. | 1140 | 796D | 2000 | A0F3 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 7800 | 0000 | 0000 | 401F | 0000 | F000
     1. | 5048 | AF32 | 0000 | 0000 | 6FFF | 0001 | 0000 | 0040 | 6150 | 4000 | 0002 | 0000 | 0000 | 0000 | 0012 | 0000
     2. | 5668 | 5814 | 0718 | 8D1C | F01E | 0480 | 008D | F810 | FF80 | 0000 | 0000 | 0100 | 141F | 0000 | 0221 | 0000
     3. | 0000 | 10B0 | 0050 | 0000 | 0000 | 0000 | 0000 | 0001 | 4020 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     4. | 0073 | 0000 | 0000 | 07A0 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0055 | 0000 | 0196
     5. | 0453 | 9ADA | 01F4 | 2055 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     6. | 0008 | 0000 | 0001 | 0000 | 8000 | 43AA | 2334 | 03E8 | 0000 | 0020 | 0000 | 0000 | 0289 | 8912 | 083E | 0000
     7. | 0000 | 0000 | 0000 | 0600 | 0000 | 00C0 | 0000 | 0000 | 05ED | 003F | 0000 | 05DC | 007D | 0003 | 0000 | 0000
     8. | 0000 | 0000 | 5AD6 | 7FF6 | 0000 | 007E | 0077 | 0000 | 03FF | 8000 | 0000 | 0BBB | 9999 | 0000 | 0010 | 0000
     9. | 0000 | 0210 | 0210 | BBBB | 0000 | 0000 | 0000 | 0000 | 0000 | 4000 | 0285 | 0285 | 0000 | 0000 | 0000 | BBBB
     A. | 0608 | 0606 | 1010 | 1010 | 0000 | 8888 | 0783 | FFFF | 0000 | 0000 | 0000 | 4442 | 5133 | 0145 | 016A | 0000
     B. | 0000 | 0000 | 0000 | 0088 | 0000 | 1020 | 1020 | 0000 | 0000 | 0000 | 7777 | 0000 | 0000 | 0000 | 0000 | 0000
     C. | 0000 | 0000 | 0000 | 0000 | 1F1F | 0003 | 0000 | 0000 | 0003 | 7000 | 0000 | 0003 | 2600 | 0000 | 0000 | 0000
     D. | 0000 | 0000 | 0002 | 0000 | 0F00 | 43D0 | 0000 | 0300 | 0000 | 0000 | 0000 | 0010 | 3800 | 0200 | 0000 | 0000
     E. | 3220 | 30C0 | 7A56 | 204B | 0000 | 0231 | 0231 | 0C83 | 9696 | 9F22 | 2040 | 0501 | 190D | 6432 | FA96 | 1840
     F. | 3E25 | 6004 | 0000 | 00F0 | 6791 | 199A | 1027 | 051C | 1027 | AC76 | 3033 | 3104 | 0067 | 199A | E721 | 1027

    Dump registers 0x130 to 0x1EF:
         | ..0  | ..1  | ..2  | ..3  | ..4  | ..5  | ..6  | ..7  | ..8  | ..9  | ..A  | ..B  | ..C  | ..D  | ..E  | ..F
         +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     13. | 0200 | 4A23 | 6520 | 4820 | 1000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     14. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     15. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     16. | 0000 | 000C | 080C | 0110 | 4489 | 00C3 | E115 | 0D4A | 0A96 | 0000 | 2A4A | 818A | 0800 | 0800 | 318C | 0095
     17. | 0C0F | 0000 | 0000 | 0000 | 000F | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     18. | 0752 | C850 | 5326 | A01E | E976 | 19CF | 31D7 | 0D9B | 0CA9 | 0014 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     19. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1A. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | F020 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1B. | 604E | 0D01 | 6D2B | 002E | 050B | 0217 | 0000 | 0000 | 0010 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1C. | 5C29 | A829 | 7E9E | F3C6 | 01C2 | 1C70 | 530B | F1C1 | 01C1 | B1C1 | 1600 | 1600 | 1000 | 1241 | 0555 | 1708
     1D. | 1313 | 1313 | 0904 | 0000 | 0000 | F500 | FFFF | FFFF | 0800 | 30D3 | F423 | 0044 | F3DF | 0000 | 0000 | 0044
     1E. | 417A | 0002 | 0025 | 0018 | 88F7 | 0000 | 0080 | 0012 | 0920 | 036D | 0000 | 4830 | 1FFD | 0000 | 0035 | 0000

    Dump registers 0xC00 to 0xC3F:
         | ..0  | ..1  | ..2  | ..3  | ..4  | ..5  | ..6  | ..7  | ..8  | ..9  | ..A  | ..B  | ..C  | ..D  | ..E  | ..F
         +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     C0. | 1140 | 616D | 2000 | A0F3 | 0020 | 4020 | 0005 | 2001 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     C1. | 3148 | 0000 | 10B4 | 0400 | 0100 | 652A | 0029 | 0002 | 01FF | 0020 | 0000 | 0000 | 1000 | 0000 | 00DB | 0000
     C2. | 250A | 0000 | F423 | 0000 | 0044 | 0000 | F3DF | 0000 | 70D3 | 0000 | 0812 | 0000 | 4844 | 0000 | F423 | 0000
     C3. | 3056 | 0000 | 0001 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    Interesting. Is the copper link alive? Or, is fiber link never established but the copper does?
    If the fiber link is never established, but the copper is, I think it may be the issue with the script.
    Could you then follow the script in the datasheet without disabling any advertising for 100Base-T?
    This is the script:

    And then apply the other scripts and see if the fiber link can still not happen. 

    If the copper link is alive but the fiber link is dropped, this can be due to link loss pass through feature. 
    In addition, this can be due to link loss pass through feature. You can disable link loss pass through feature by writing 1FF5h to register 1ECh. 

    Best,
    J

  • 0 in reply to J
    TI__Genius 14026 points

    In addition, if you are doing a back to back media converter with two PHYs, have you considered using copper to SGMII, and connect SGMII with the other PHY? This will lower the number of wires you need unless you have to use fiber. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    When I said "without script", I meant without the scripts that modify the link margin. When the equipment starts up, I do execute script 7.4.8.5.

    I'm not doing a back to bakc media converter with two PHYs :
    Wi-Fi Router <--- 1000BASE-T ---> DP83869HM <--- 1000BASE-FX -----> 1000BASE-FX Switch

    When I use scripts which modify link margin, the fiber optic interface does not link even if I don't disable advertising 100BASE-T. But If I disable link loss pass through feature, this interface links.

    Dump with disabling advertising 100BASE-T and link loss pass through feature (link is good but there are errors):
    Dump registers 0x00 to 0xFF:
        |  .0  |  .1  |  .2  |  .3  |  .4  |  .5  |  .6  |  .7  |  .8  |  .9  |  .A  |  .B  |  .C  |  .D  |  .E  |  .F
        +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     0. | 1140 | 796D | 2000 | A0F3 | 0001 | C101 | 006D | 2001 | 6801 | 0300 | 3800 | 0000 | 0000 | 401F | 0000 | F000
     1. | 5048 | AF32 | 0000 | 0000 | 29C7 | 0025 | 0000 | 0040 | 6150 | 4000 | 0002 | 0000 | 0000 | 0000 | 0012 | 0000
     2. | 5668 | 5814 | 0718 | 8D1C | F01E | 0480 | 008D | F810 | FF80 | 0000 | 0000 | 0100 | 141F | 0000 | 0221 | 0000
     3. | 0000 | 10B0 | 0050 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     4. | 0073 | 0000 | 0000 | 07A0 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0055 | 0000 | 0116
     5. | 0453 | 9ADA | 01F4 | 2055 | 0000 | 0011 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     6. | 0008 | 0000 | 0001 | 0000 | 8000 | 43AA | 2334 | 03E8 | 0000 | 0020 | 0000 | 0000 | 0289 | 8912 | 083E | 0000
     7. | 0000 | 0000 | 0000 | 0600 | 0000 | 00C0 | 0000 | 0000 | 05ED | 003F | 0000 | 05DC | 007D | 0003 | 0000 | 0000
     8. | 0000 | 0000 | 5AD6 | 7FF6 | 0000 | 007E | 0077 | 0000 | 03FF | 8000 | 0000 | 0BBB | 9999 | 0000 | 0010 | 0000
     9. | 0000 | 0210 | 0210 | BBBB | 0000 | 0000 | 0000 | 0000 | 0000 | 4000 | 0285 | 0285 | 0000 | 0000 | 0000 | BBBB
     A. | 0608 | 0606 | 1010 | 1010 | 0000 | 8888 | 0783 | FFFF | 0000 | 0000 | 0000 | 3435 | 3535 | 00EC | 0167 | 0000
     B. | 0000 | 0000 | 0000 | 0088 | 0000 | 1020 | 1020 | 0000 | 0000 | 0000 | 7777 | 0000 | 0000 | 0000 | 0000 | 0000
     C. | 0000 | 0000 | 0000 | 0000 | 1F1F | 0003 | 0000 | 0000 | 0003 | 7000 | 0000 | 0003 | 2600 | 0000 | 0000 | 0000
     D. | 0000 | 0000 | 0002 | 0000 | 0F00 | 43D0 | 0000 | 0300 | 0000 | 0000 | 0000 | 0010 | 3800 | 0200 | 0000 | 0000
     E. | 3220 | 30C0 | 7A56 | 204B | 0080 | 0231 | 0231 | 0C83 | 9696 | 9F22 | 2040 | 0501 | 190D | 6432 | FA96 | 1840
     F. | 3E25 | 6004 | 0000 | 0000 | 6791 | 199A | 1027 | 051C | 1027 | AC76 | 3033 | 3104 | 0067 | 199A | E721 | 1027

    Dump registers 0x130 to 0x1EF:
         | ..0  | ..1  | ..2  | ..3  | ..4  | ..5  | ..6  | ..7  | ..8  | ..9  | ..A  | ..B  | ..C  | ..D  | ..E  | ..F
         +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     13. | 0200 | 4A23 | 6520 | 4820 | 1000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     14. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     15. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     16. | 0000 | 000C | 080C | 0110 | 4489 | 00C3 | E115 | 0D4A | 0A96 | 0000 | 2A4A | 818A | 0800 | 0800 | 318C | 0095
     17. | 0C0F | 0000 | 0000 | 0000 | 000F | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     18. | 0752 | C850 | 5326 | A01E | E976 | 19CF | 31D7 | 0D9B | 0CA9 | 0014 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     19. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1A. | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | F020 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1B. | 604E | 0D01 | 6D2B | 002E | 050B | 0217 | 0000 | 0000 | 0010 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     1C. | 5C29 | A829 | 7E9E | F3C6 | 01C2 | 1C70 | 530B | F1C1 | 01C1 | B1C1 | 1600 | 1600 | 1000 | 1241 | 0555 | 1708
     1D. | 1515 | 1515 | 0904 | 0000 | 0000 | F500 | FFFF | FFFF | 0800 | 30D3 | F423 | 0044 | F3DF | 0005 | 0000 | 0044
     1E. | 417A | 0002 | 0025 | 0018 | 88F7 | 0000 | 0080 | 0012 | 0920 | 036D | 0000 | 4830 | 1FFD | 0000 | 0035 | 0000

    Dump registers 0xC00 to 0xC3F:
         | ..0  | ..1  | ..2  | ..3  | ..4  | ..5  | ..6  | ..7  | ..8  | ..9  | ..A  | ..B  | ..C  | ..D  | ..E  | ..F
         +------+------+------+------+------+------+------+------+------+------+------+------+------+------+------+------
     C0. | 0140 | 614D | 2000 | A0F3 | 0020 | 0000 | 0004 | 2001 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000
     C1. | 3148 | 0000 | 10B4 | 0400 | 0100 | 652A | 0029 | 0002 | 01FF | 0000 | 0000 | 0000 | 3400 | 0000 | 00DB | 0000
     C2. | 250A | 0000 | F423 | 0000 | 0044 | 0000 | F3DF | 0000 | 70D3 | 0000 | 0812 | 0000 | 4844 | 0000 | F423 | 0000
     C3. | 3056 | 0000 | 0001 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000 | 0000




    Initialization script :
    // Disable link loss pass through feature. 
    01EC   1FF5

    // ***********
    // Set Operation Mode to 1000Base-T to 1000Base-X
    01DF   0044

    // Reset BMCR
    0000   1140

    // ***********
    // "Advertise 1000Base-T ability" and "Disable Advertise "100Base-T ability"
    0009   0300
    0014   FFFF
    // ***********
    // Disable Advertise "100Base-T ability"
    0004   0001
    // ***********

    // Reset PHY_CONTROL
    0010   5048

     // Reset FX_CTRL
    0C00   1140

    // Software Reset
    001F   4000

    // ***********

    // Hard reset
    001F   8000

    // Threshold for consecutive amount of Idle symbols for viterbi Idle detector to assert Idle Mode set to 5
    0053   2054

    // CAGC DC Compensation Disable
    00EF   3840

    // Master Training Timers - increasing time in different training states
    0102   7477

    // Master Training Timers - increasing time in different training states
    0103   7777

    // Master Training Timers - increasing time in different training states
    0104   4577

    // Timing Loop Bandwidth
    010C   7777

    // Timing Loop Bandwidth
    01C2   7FDE

    // Slave Timers - increasing time in different training states
    0115   5555

    // Slave Timers - increasing time in different training states
    0118   0771

    // Timing Loop Bandwidth
    011D   6DB2

    // Timing Loop Bandwidth
    011E   3FFB

    // Timing Loop Bandwidth
    01C3   FFC6

    // Timing Loop Bandwidth
    01C4   0FC2

    // Timing Loop Bandwidth
    01C5   0FF0

    // FEE Fix
    012C   0E81

    // Soft reset
    001F   4000

    // ****************
    // Hard reset
    001F   8000

    // Increase time for AGC
    0102   7477

    // No-AGC Re-Train
    00E4   0080

    // Soft reset
    001F   4000



  • 0 in reply to MARMONIER Cedric
    Prodigy 20 points

    With a 50-meter cable, the equipment works in the same way if I don't disabling advertising 100BASE-T.
    If I do disabling advertising 100BASE-T, fiber interface doesn't link with this cable.

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi, 

    This looks like an issue with the link loss pass through feature. I believe this is happening because the copper link between the router and the PHY is poor, and the PHY is detecting the poor link as a link drop sometimes and drops the fiber link. Please feel free to keep link loss pass through disabled to ensure the link is up. 
    Could you try
    // Threshold for consecutive amount of Idle symbols for viterbi Idle detector to assert Idle Mode set to 5
    0053   2053

    Instead of 0053 2054? This is supposed to lower the minimum IPG required on the PHY and may help with the link margin, but this is the lowest setting we can do. 

    Best,
    J

  • 0 in reply to J
    Prodigy 20 points

    If the copper link between the router and the PHY is of poor quality, then why does it work when we replace the media converter with a switch, and why does the media converter work with 100m of cables (Ethernet tester in place of the Wi-Fi router: BERT test performed)?

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Hi,

    I agree that this is a very peculiar case. What is the wifi router model again? I would like to see if we can test this in house with our EVM.

    Best,

    J

  • 0 in reply to J
    Prodigy 20 points

    Wifi router is CISCO CW9166I-E

  • 0 in reply to MARMONIER Cedric
    TI__Genius 14026 points

    Okay, I will confirm internally to see if we can get the router for in-house testing. While I am confirming, could you test 0053 2053 to see if that improves anything?