• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 33 subscribers
  • Views 660 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.

DP83848K: Communication error when cable connect/disconnect test

Toshihiro Watanabe
Genius 10235 points
Part Number: DP83848K

Hello,

We are testing cable plug/unplug test and error occurs around once in 5 to 10 times. 1:1 connection and both phys are DP83848K.

When this error occurs, Bit9 of 0x0010 PHY Status Register (PHYSTS) becomes 0. bit0 Link Status is 1 so we think link is established.

This error is gone if power cycle does.

What is occurring this and would you advice us how to solve this issue?

Best regards,

Toshihiro Watanabe

  • 0
    TI__Genius 10235 points
    Hello,
    Auto MDI/MDIX is invalid and datarate is 100Mbps fixed. This info is added just in case.
    Best regards,
    Toshihiro Watanabe
  • 0 in reply to Toshihiro Watanabe
    TI__Mastermind 23675 points
    Hi Toshihiro-san,

    Can you provide a register dump from 0x0 to 0x1F please?
    What is the cable type and length used?
    Are both PHYs configured the same way?
    Can you share a schematic of the PHY portion?
  • 0 in reply to Ross Pimentel
    TI__Genius 10235 points

    Hello Ross-san,

    As I shared the answer of your question offline but let me add it online too.

    Cable type is category 5e and its length is 1 to 2m. Both PHYs are the same configuration. The schematic is to let us share offline. the register dump is attached.

    Fullscreen DP83848RegisterDump.txt Download 
         OK     NG
0x00 0x2100 0x2100
0x01 0x784D 0x784D
0x02 0x2000 0x2000
0x03 0x5C90 0x5C90
0x04 0x0181 0x0181
0x05 0x0000 0x0000
0x06 0x0004 0x0004
0x07 0x2001 0x2001
0x08 0x0000 0x0000
0x09 0x0000 0x0000
0x0A 0x0000 0x0000
0x0B 0x0000 0x0000
0x0C 0x0000 0x0000
0x0D 0x0000 0x0000
0x0E 0x0000 0x0000
0x0F 0x0000 0x0000
0x10 0x0605 0x2C05
0x11 0x0000 0x0000
0x12 0x0000 0x0000
0x13 0x0000 0x0000
0x14 0x0000 0x00** bit0-6 keep changing
0x15 0x0000 0x000* bit0 keep changing
0x16 0x0100 0x0100
0x17 0x0001 0x0001
0x18 0x0000 0x0000
0x19 0x0020 0x0020
0x1A 0x0804 0x0804
0x1B 0x0000 0x0000
0x1C 0x0000 0x0000
0x1D 0x6011 0x6011
0x1E 0x083E 0x083E
0x1F 0x0000 0x0000

    Best regards,

    Toshihiro Watanabe

  • 0 in reply to Ross Pimentel
    TI__Genius 10235 points
    Hello Ross-san,

    We tried the same test by 7m cable and the issue gone in 100 times plug/unplug.

    Why longer cable worked? Could you find out the issue is relating with cable length from register value? We want to know the reason.

    Best regards,

    Toshihiro Watanabe
  • 0 in reply to Toshihiro Watanabe
    TI__Mastermind 23675 points
    Hi Toshihiro-san,

    Could you try the same experiment but with a longer cable?
    Please try this test with a 10m cable or longer.
  • 0 in reply to Ross Pimentel
    TI__Genius 10235 points

    Hello,

    The longer cable works for the issue. Why longer cable works?

    Best regards,

    Toshihiro Watanabe

  • 0 in reply to Toshihiro Watanabe
    TI__Genius 10235 points
    Hello Ross-san,
    I could understand the signal is saturated in case of shorter cable. We will try 10pF adding measurement.
    We would like to confirm some.
    1) Where and how to add 10pF?
    -- add 10pF at TD+(pin15) to TD-(pin14), RD+(pin12) to RD-(pin11) ?
    -- add 10pF at TD+(pin15) to GND, TD-(pin14) to GND, RD+(pin12) to GND and RD-(pin11) to GND?
    -- other portion?

    2) Can we use adding 10pF solution if it worked? or Adding 10pF is just debug purpose?

    Please let us confirm how you are thinking.
    Best regards,
    Toshihiro Watanabe
  • 0 in reply to Toshihiro Watanabe
    TI__Mastermind 23675 points
    Hi Toshihiro-san,

    Please place the 10pF on each of the MDI pins to ground (TD+, TD-, RD+, RD-).
    The caps can be placed anywhere between the PHY and transformer.

    There is no issue using added 10pF cap on the line for normal use.