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.

DP83TD510E-EVM: Cannot get DP83TD510E-EVM boards to connect to each other.

Part Number: DP83TD510E-EVM
Other Parts Discussed in Thread: USB-2-MDIO,

Powering the DP83TD510E-EVM using the benchtop power supply helped to get the board to connect to the USB-2-MDIO software.  I believe I am using the USB-2-MDIO tool correctly and wrote the previously supplied (AutoNegotiation_Initialization and Force_1Vpp_Master_Init.txt) scripts to PHY 0 of one board and (AutoNegotiation_Initialization and Force_1Vpp_Slave_Init.txt) scripts to PHY 0 of one the other board.  I made sure that in each case "Extended register access was set to "No".  The boards each establish a link with the Raspberry Pis connected to their RJ45 ports but the boards do not connect to each other over the single pair I have connected between them.  Can you suggest some steps I might be missing such that I can get the two boards to connect to one another via the single pair? I am going to try and include photos of the board and setup if helpful.

  • I didn't see the images on the original post so I am attaching them as a Word document.DP83TD510E-EVM setup photos.docx

  • Hi Wayne,

    How have you confirmed that the script was successfully uploaded? Can you share the register dump of the DP83TD510 devices? 

  • Hi Cecilia,

    I am not sure if the script was uploaded correctly.  I was assuming that the screen shown in the attached file indicated that it was successfully uploaded.  I'm not sure how to get the register dump but I am willing to try. 

    WayneScript upload.docx

  • Hi Cecilia, Justin,

    I think I figured out how to read the register values and clearly, the script I'm trying to load is not loading.  The first register "001f" is supposed to contain "8000" but it reads as "0000".  Is there a special mode this board needs to be in to read and write to it?  I am powering it externally and loading the program to it with the USB port. 

    Wayne

    Board LEDs.docx

  • Hi Wayne,

    It looks like you have an open discussion on E2E with Justin already. Lets consolidate our threads so that we have one single E2E chain on the debug. I have added Justin to this thread as well.

    And to your question about 001f, that register actually is a write for reset and reads back 0000 once it has been written to. I suggest reading registers 0x0000 and 0x0001

    Thanks,

    Cecilia

  • Hi Cecilia,

    Thanks for consolidating threads.  Thanks also for the tip on reading the correct register.  After executing what I believe is the correct 1VppMaster and 1VppSlave scripts I still can't get the boards to reliably link with each other or with the equipment on the other end of the board (RJ45 side).  I was able to read the 0000/0001 registers as 1100/0149 for the "Master" and 0100/0149 for the "Slave".  I'll be in a position to work on this more tomorrow.

    Wayne

    (wayne.hopkinson@commscope.com)

  • Hi Wayne,

    Can you please verify which setup you are using to run the configuration scripts from below?

    There are three possible configurations to run to establish a link between the DP83TD510E boards:

    1. AutoNegotiation_Initialization.txt on both Board A and Board B. (Extended registers menu should be set to "YES")

    2. Force_1Vpp_Master_Init.txt on Board A and Force_1Vpp_Slave_Init.txt on Board B. (Extended registers menu should be set to "NO")

    3. Force_2p4Vpp_Master_Init.txt on Board A and Force_2p4Vpp_Slave_Init.txt on Board B. (Extended registers menu should be set to "NO")

    Please do not run both Auto-Negotiation and Forced mode script on the same board. 

    Regards,
    Justin 

  • Hi,

    Initially, I tried to use the AutoNeg option AND 1V Master/Slave option at the same time.  I understand this is incorrect.  I don't know if I need to do something to the boards to get them to drop this earlier attempt.

    Since then, I have been focusing on the1V Master/Slave option.  I have done so by trying to upload the 1V Master script onto one of the boards (via the USB port) and the 1V Slave script onto another board using its corresponding port.  I am NOT using the AutoNeg script any longer.  I am uploading this script to PHY 00 and with extended registers set to No.  In each case the script(s) seem to load and last post I read the contents of registers 0000 and 0001.  The board is being powered with an external power supply and the jumper selecting USB power is not shorted.

    An broader description of this relatively simple application follows:  I can get two Raspberry Pi computers to talk to one another via iPerf.  They connect to one another via a hub and 4-pair category style cables.  I want to replace one of the 4-pair cables with a 1-pair cable and establish the same contact between Pis.  To do this I am using two TI DP83TD510E-EVM modules facing one another with their RJ45 portions facing "outward".  So, really, I'm just using the TI boards to shift from 4-pair to 1-pair and back again. 

    Wayne

  • I don't know if this solved the case but I shifted to a different USB hub and after power cycling the boards and retrying the Master/Slave scripts sent out earlier, my two boards are communicating with one another.  This issue can be closed.