• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 100 subscribers
  • Views 111 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

AM67: Failing to meet RGMII timing requirements with DP83867IRRGZ PHY

Douglas Smith
Prodigy 20 points
Part Number: AM67

Tool/software:

I have been testing the signal quality of the RGMII1 signals in our design and have been unable to get all the RGMII signals to meet a 20%/80% rise/fall time of 0.75ns. In our design the RGMII signals are run to a 1x3 pin jumper that can connect the RGMII signals to either the PHY or a connector (in this case they are connected to the PHY). We have set the drive strength of the RGMII signals to the maximum value and while the RX lines meet the listed timing, several TX lines fail too. I have included a few examples.

TD0 rise

TD2 rise

We also tried changing the RGMII signal voltage level to 1.8V, but that seemed to either not help or make things slightly worse.

TD3 rise (the yellow and blue were both TD3, just with two different probes)

We also tried testing the signal quality of our AM67x EVM, but found the signal quality to also be failing on that.

TD3 on the EVM

We also checked our voltage rails to see if the power was dipping during operation, but could not find any signs of that. From our perspective it seems that the AM67x just isn't putting out enough drive strength on these signals to meet the 0.75ns rise/fall time requirement. Is there something we are missing that could be adjusted to fix this problem?

  • 0
    TI__Guru 60300 points

    Hi Doug,

    Is this failing to meet the RGMII functionality? Are you seeing TX/ RX packets dropped or errors due to this?

    What were the adjustments made to increase the drive strength to maximum? Is there other pullup or pulldown resistors on the board, or in the SoC during the pin mux selection?

  • 0 in reply to ShreyasRao
    Prodigy 20 points

    RGMII appears to be functional as I have not seen any issues with dropped TX/RX packets. For software we are just setting output impedance to the minimum. None of the pullups or pulldowns in the soc are enabled right now. The only pullups/pulldown resistors present on the board are on RD0, RX_CTL, and LED1 for strapping purposes.

  • 0 in reply to Douglas Smith
    TI__Guru 60300 points

    Douglas,

    Yes, we have used the RGMII on several of the devices and for several years without any issues. With the minimum impedance setting, its usually 50ohm which gives a decent Tr/Tf to not cause issues. On the other hand, with faster rise and fall times, there are concerns on the EMC emissions being higher. With that in mind, the termination is tuned such that the RGMII performance is maintained as long as the setup and hold times are maintained.