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.

DP83848m debbug

Other Parts Discussed in Thread: DP83848C, DP83848M

Hi, 

I've bought an Ethernet starter kit using DP83848c with PIC32 (in RMII mode) with the provided code and it works very well. 

Then i've made my own PCB but using DP83848m with PIC32 (in RMII mode).

It's detected in the network, and i could load the html page twice. 
It is still detected on the network but i can't load the pages any more.  So i no idea of what could be happening. 
Any idea please ? 

  





Here are some DC points i've check :

Description   DC Value

AGND            0 Volts

PFBIN1         1.72 Volts

 AGND            0 Volts

 AVDD33        3.3 Volts

PFBout          1.72 Volts

RBIAS            1.19 Volts

RESET_N     3.3 Volts

IOGND          0 Volts

DGND           0 Volts

PFBIN2         1.72 Volts

IOGND          0 Volts

IOVDD33     3.3Volts


 


 


PS : 

  • Hi,

    You have certainly started off on the right track in terms of debug.  Note that loading pages is very high level functionality from a PHY perspective.  We will need to focus on lower level functionality in order to determine what is happening.

    A few comments on the schematic:

    1. L1 and L2 are not recommended.  TD+/- should connect directly to the transformer.
    2. Both center taps for the transformer (pins 2 and 5) should connect to 3.3.V.
    3. The RBIAS resistor should be 4.87 kOhms.

    Regarding the value on the RX_DV pin, you will need to measure it when the RESET_N pin is asserted in order to see the strap value pulled to the expected voltage. 

    Do you have the ability to read the registers via MDC/MDIO? 

    Patrick

  • I have moved your post to the Ethernet Forum.  To help with better supporting and tracking your future questions in Ethernet, please post any new Ethernet related questions into this forum.

    Thanks,

    Patrick

  • Hi, 
    thanks for replying ! 
    1. change done
    2. change done
    3. change done
    Now the hardware looks to be exactly the same.  What could i miss ?
    RESET_N is not used and left unconnected  (like in starter kit).
    I can't read MDC/MDIO, my analyser is to slow. 
    I'm running the code in debug mode to see difference between my card and starter kit. 
    He never go there ;( 

    // Determine if this connection is eligible for processing
    if(httpStubs[conn].sm != SM_HTTP_IDLE || TCPIsGetReady(httpStubs[conn].socket))
    {
    HTTPLoadConn(conn);
    HTTPProcess();
    }

     
  • The http aspects of the functionality is still to high level for me.  I need to find ways to evaluate this from a lower level, closer to the level of the PHY electrical characteristics.

    Could you clarify the comment on being too slow to read MDC/MDIO?  The interface itself has no minimum speed.  It could be bit-banged via GPIO if required. 

    If the starter kit worked, but the new board is not providing the same level of functionality, perhaps we should focus on the differences between the boards.  What changes were made and what level of A-B comparison between the boards is possible?  Could you compare the clocks and RMII signals, for example?

    Patrick