DP83TG720S-Q1: PHY layer MDI mode conversion loss test failed

hello，

I,am the one who have the problem, thansks

Hi Amy,

The 720 has been tested for mode conversion loss at UNH and it passes, however this test is very sensitive and set up dependent. Could you share how you're running this test and the set up?

Thanks,

Alvaro

Hi Alvaro,

Here's customer feedback:

(1) What does UNH mean?

(2) What should be the correct test method and setup?

(3) Our tests are commissioned to an external laboratory. 

There are two test methods, each tested by a different external laboratory, one of which is as follows and the other to be confirmed.

YF_1000Base_T1_PMA_测试项目：MDI Mode Conversion Loss Adaptor Verification test

Test Tool:

Hardware: Network Analyzer
Software: R&S ScopeSuite (80.1)
Test Process:

(1) Connect calibrator and begin calibration of both ports.

(2) Remove calibration and connect RF harness to test clamp.

(3) Set the DUT to SLAVE mode and connect the clamp to the DUT

(4) Click Next to start the test.

Best Reagrds,

Amy

hello，Mr.Alvaro (Al-vuh-roe) Reyes ，is there any result of analysis？

Hi Amy,

Mode-Conversion is sensitive. Layout and components (between PHY and connector) also make a difference. 

What is the mismatch between TRD_P & TRD_M impedance/length and could you share the CMC being used? 

Hi Alvaro,

Thank you for your response. Below is the information provided by the customer.

The differential impedance between TRD_P and TRD_M is controlled at 100 ohms, with a wire spacing of 8 mil and a wire length of 2008.19 mil. The CMC part number being used is the ACT1210G-800-2P-TL05 of TDK.

The layout is highlighted in the below figure

Hi Amy,

To identify the contribution of layout: Could you remove the PHY and place a 100 ohm resistor connecting the TRD_P & TRD_M pins and then redo the measurement.

If results are similar, then it is a layout/connector/Non-PHY related issue.

It looks like there are some spot frequencies which are failing the spec. This would NOT cause problems in the functionality or EMC/EMI performance of the PHY.

Regards,

Alvaro

Hi Alvaro,

Thank you for your suggestion!

The customer feedback that he took the DP83TG720S-Q1 EVM board to the lab for this test, and the test results were also failed, as follows:

Can you help analyze the problem and see where it is? Can we improve anything else? Thank you.

Best Regards,

Amy

Hi Alvaro,

No other team member contacted me via email so far. Or are you referring to the customer with this issue, not me?

Best,

Amy

Hi Alvaro,

I thought the team contacted this customer by email, but it didn't seem like it was.

The customer wants to know about the problem handling process and whether you can synchronize this process to his email ？His e-mail address is zouyongchang@yftech.com