• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 33 subscribers
  • Views 335 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: Straps-only configuration in 100 Mbps Media Converter mode， copper cable pulled out， DP83869 enter a crash state

ych_chn
Intellectual 345 points
Part Number: DP83869HM
Other Parts Discussed in Thread: DP83869

Tool/software:

Hello, I am currently working in the 100 Mbps Media Converter mode. Can I also achieve the Media Converter function through Straps-only configuration?

In addition, I found that when the equipment is operating normally, it is fine to disconnect the optical fiber and then reinsert it. However, if the copper cable is pulled out, LED1 (Table 7-13 LED1 Default Fiber Link established: Stable ON) will flash rapidly. If the copper cable is reinserted, DP83869 will no longer function properly and needs to be powered on again to normal work.

My Straps are: OPMODE[2-0] = 101, LED_1 = 1, LED_2 = 1, RX_CTRL and RX_CLK are left floating.

As mentioned in 

 DP83869HM: Straps-only configuration possible in 1000 Mbps Media Converter mode with the DP83869 at all? 

  • 0
    TI__Mastermind 39855 points

    Hello,

    100M MC functionality can be achieved via strapping. The behavior your described may be indicative of link loss pass through, but the specifics are a bit odd. This feature, upon enabling, will knock out the other media (FX in this case) when the initial media (CU in this case) has lost link. I would expect that the FX link would sever instead of blink rapidly, and recall checking this functionality in our lab on our EVM a while back.

    Are you evaluating on our EVM, or on your own custom design?

    Sincerely,

    Gerome

  • 0 in reply to Gerome Cacho
    Intellectual 345 points

    custom design，please checking this functionality in your lab,83869 rash when led fast blink, 

  • 0 in reply to ych_chn
    TI__Mastermind 39855 points

    Hello,

    Please expect results on this EoD tomorrow.

    Sincerely,

    Gerome

  • 0 in reply to Gerome Cacho
    TI__Mastermind 39855 points

    Hello,

    I tested out this behavior and found similar phenomena present:

    - 2x EVMs plugging in via CU and FX to each other (circular topology)

    - Both EVMs configured to 100M MC mode with LLPT enabled

    - Upon breaking FX cable, CU and FX link both broke and upon reconnection of FX, both CU and FX links were up (verified via LED_0 [CU] and LED_1 [FX])

    - Upon breaking CU cable, CU and FX link both broke and upon reconnection of CU, both links were still down. However, upon Reset of device (Reg 0x1F = 0x4000), functionality was restored. One of of my boards, FX link was up, but on the other, it was down. I believe that because of this, the CU was also affected by this.

    I wanted to do a quick check as the board with FX link down was using an older revision of silicon. What does Reg 0x3 indicate on your design?

    Sincerely,
    Gerome

  • 0 in reply to Gerome Cacho
    Intellectual 345 points

    Yes, this phenomenon is consistent with the one we tested. However, unfortunately, the board we made did not have the SMI interface, so we couldn't obtain the value of the registers. This can be read on the EVM board.

    Sincerely,
    ych_chn

  • 0 in reply to ych_chn
    TI__Mastermind 39855 points

    Hello,

    Do you either have the ESD bag with device label that you had gotten the parts from? The revision should be there.

    Alternatively, you may snap a photo of the top marking of the device and I can reverse lookup the revision. However, if there are multiple reels/lots you have, this may be unsustainable.

    Sincerely,

    Gerome