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.

[FAQ] DP83TG720R-Q1: Why is my SGMII eye measurement distorted?

Part Number: DP83TG720R-Q1

SGMII signal is high frequency low-voltage differential signal. Both layout design and measurements need to be accurate for good eye diagram capture.

Common board design guidelines:

  • 100ohms+/-10% differential traces,
  • ac-coupling caps close to transmitter +
  • other high-speed signal layout guidelines as per app note:

Common measurement guidelines:

  • high impedance differential probes
  • probing signals close to the receiver pins (at the receiver pins if possible)
  • ~4GHz or higher bandwidth oscilloscope


If above guidelines are met and SGMII transmitter meets the specification, then the eye diagram at the input of the PHY should be good for PHY to function without packet errors.


But sometimes even if there are no packet errors + board/measurement guidelines are met, SGMII eye diagram looks distorted. Following may explain the distortion in eye measurement but still no data errors.


Impact of Package: Eye Diagram @ Pin vs Silicon Die

In between the pin and the die, there will be some parasitic capacitance (C), parasitic inductance (L), and some resistance (R). You can read more about these parasitics in this app note:

Fig 1. Package parasitics block diagram.


Signal probed at pin will have reflections because of bond-wire and related parasitics, while die will still see the clean eye and hence no packet errors.

                   Fig 2. SGMII Eye Diagram at Package                           Fig 3. SGMII Eye Diagram at DIE


For this example, we used the DP83TG720S as the DUT, with R =  0.1 Ohms, C = 3.2 pF, and L = 3.2 nH.

The simulation setup is shown in the diagram below.[1] To learn how to set up a simulation similar to this, refer to this app note:

Fig 4. DP82TG720S SGMII IBIS Eye Simulation Setup in ADS.

Customers can fit the below eye in the measured eye diagram and if it passes the mask the signal can be considered passing the requirement (as actual eye opening seen by silicon die will be even better than what is measured on pin)


Fig 5. This is the DP83TG720S SGMII mask requirement found in the datasheet.


[1] Note that these are modeled with ideal transmissions lines and the simulated eye would be more accurate if they were replaced with the s parameter files extracted from the PCB.