Because of the Thanksgiving holiday in the U.S., TI E2E™ design support forum responses may be delayed from November 25 through December 2. Thank you for your patience.

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.

DP83848C: DP83848 problem Rj45 not even LED's ON

Part Number: DP83848C

Hello there!
 
I have made a custom board with Eagle (Autodesk), I used the PIC32MZ2048EFH64 with a dp8384 Ethernet physical layer transceiver. Also a few other things, like regulated voltage supply and i/o ports (the basic stuff).
I can program the pic etc, working fine. But I do have problems with the dp83848, and it maybe looks like a hardware problem?the strange thing is that when I put a ethernet cable at my rj45 and so I connect it to my phy,

The strange thing is that the lights doesn’t blinks on, On the rj45. Without Any software it should detect a connection with the network or switch right?

I also tried a cross over cable, I thought maybe it had to do something with that.
Because my board has a tap ub rj45 connector, so it is the same as a switch TP-Link right, so I need a cross over.
 
But that didn't fixed the problem ass well for the led Activity situation.

I will put down a schematic of the situation,  I think the part of the PIC microcontroller doesn't matter a lot for the problem now. Because I cannot ping the device, but the problem already starts in my opinion by not see the activity and LINK led ON, ass wel on the rj45 connector as the switch.
 
Hopefully there is someone out there that can help me!


Best regards,
 
Marc
 

  • Hi Marc,

    Yes the PFBOUT and PFBIN capacitors need to be placed close to the pins to reduce parasitic inductance of the traces and PCB. This is expected to be a secondary effect that may impact data errors over a large time scale. A primary effect of no link up indicates a much larger issue.

    The placement of the termination resistors on the TD and RD pins are a bit more critical to link than the bypass caps on PFB circuit. The termination resistors should be placed close to the PHY pins. Please see the layout of any of the DP83848 EVMs. I suggest the 848I-MAU EVM as one of the best to reference.

    You can also post images of your TD/RD layout here.

    Best regards,
  • Hi Marc,

    I don't see any immediate issues with your schematic. Your nets appear to be connected properly for PFBOUT.

    Do you have MDIO access from the PIC? If so, can you read out the registers and post them here?

    For hardware verification, please do the following below:
    I'd recommend verifying the voltages are in spec at all the supply pins with a multimeter.
    Please verify the 25MHZ_OUT pin is outputting a 50MHz(because RMII mode) reference.
    Measure the PFBOUT pin and PFBIN1 and 2 pins. These should be 1.8V.
    Finally, without the switch connected, please probe the TD+/- pins of the PHY to see the pulse train for auto-negotiation.

    Best Regards,
  • Thank you very much for the fast reply.
    I will try what you told me to do and will let you know the results.

    By the way I read something about that the capacitors of the PFBOUT pin and PFBIN1 and 2 pins, should be placed closely as swell for the resistors and capacitors on the rj45 jack connector. Is this very important and true? I can show you my board layout, to determine if it is ok or may cause problems and why..
  • I made a screenshot with indicators, I turned the top and bottom layer off (route wires) to show it clearly.

    So what you are saying is that those 4 times 49.9ohm could be a problem?

  • PFBOUT pin and PFBIN1 and 2 pins are 1.755V. so ~1.8V, I think this is ok.
    But how to measure the 25MHz out pin? Probably can only read that one out with a scope, but I only have a old scope that goes to max 15MHz.
  • On the TD+/- and RD+ I measured a ~3.25V and on the RD- I measured almost nothing.
  • Hi Marc,

    That definitely seems like an issue with the front end of the PHY. You should be measuring ~VDD/2 on the TD and RD pins.

    Are the nets connected properly?

    The termination resistors should be local to the PHY pins as I mentioned.

    One decoupling cap should be local to the termination resistors. One decoupling cap should be close to the pull-ups for the magnetics center taps.

    Best Regards,
  • Hi,

    I got now a the link led working and activity led when I send something over the network.

    But I got a problem that I can't ping the device, but I don't know if it have to do something with my microcontroller.

    Still measuring ~3.2V on those td en rd pins
  • Hi Marc,

    Can you read the registers? If you can read register 0x10 bit0 to look at the link status, and register 0x10 bit1 to decode link speed. If you are getting an active, 100M link, then the issue may still be with your microcontroller.

    Because your MDI layout is less than ideal, I suggest creating a loopback cable and running the BIST mode of the DP83848 if you can get to the registers.

    BIST will generate arbitrary data in the PHY, loop it through the cable and decode. If you are having data errors due to the front end, BIST will fail.

    BIST is described in section 6.3.5 of the datasheet. A loopback cable can be created using these instructions: www.juniper.net/.../fe-ge-loopback-plug-rj-45.html

    Best Regards,