SNLA385 September 2021 DP83561-SP , DP83620 , DP83630 , DP83640 , DP83822H , DP83822HF , DP83822I , DP83822IF , DP83825I , DP83826E , DP83826I , DP83843 , DP83847 , DP83848-EP , DP83848-HT , DP83848C , DP83848H , DP83848I , DP83848J , DP83848K , DP83848M , DP83848Q-Q1 , DP83848T , DP83848VYB , DP83848YB , DP83849C , DP83849I , DP83849ID , DP83849IF , DP83867CR , DP83867CS , DP83867E , DP83867IR , DP83867IS , DP83869HM
E2E is a trademark of Texas Instruments.
Microsoft are registered trademarks of Microsoft Corporation.
Tektronix is a registered trademark of Tektronix, Inc..
All trademarks are the property of their respective owners.
The IEEE 802.3 standard defines a set of compliance tests for the 1000BASE-T Ethernet physical layer. These tests are intended to ensure reliable communications over an Ethernet network. To perform this testing, the device under test (DUT) is placed in each of four IEEE 802.3 specified modes. The standard defines the test modes as listed in Table 1-1.
Test Mode | Test | IEEE 802.3 Reference |
---|---|---|
Test Mode-1 | Peak Droop Template |
40.6.1.2.1 40.6.1.2.2 40.6.1.2.3 |
Test Mode-2 Test Mode-3 |
Transmitter Jitter Receiver Jitter |
40.6.1.2.5 |
Test Mode-4 | Distortion MDI Return Loss MDI Common Mode Voltage |
40.6.1.2.4 40.8.3.1 40.8.3.3 |
Ethernet compliance is executed with an oscilloscope that operates a specific Ethernet compliance software test suite. Such Ethernet compliance test equipment is made available for purchase from many test equipment manufacturers like Tektronix, Agilent, Lecroy and others.
Ethernet compliance test equipment implements a software test suite. The software suite executes the test cases step by step in a semi-automated way. The IEEE specification in some cases does not provide a fixed formula but describes the type of math in text form. Each test equipment vendor has the freedom to implement their own measurement ways and formulas. As result, different interpretations of the math formula have been implemented by manufacturers in their Ethernet compliance testing software suite.
Specifically, we are looking at test case “Peak differential output voltage and level accuracy” as described in IEEE 802.3-2018 Subclause 40.6.1.2.1. This test case requires Test Mode-1 to be activated in the Ethernet PHY.
What we have observed in the market is that two different kinds of formulas are implemented by test tool manufacturers. One formula calculates percentage error, while the other calculates percentage difference.
Some manufacturers implemented the percentage error formula in a different way to other manufacturers. In some cases, this resulted in a doubling of test values, and subsequently erroneous failures. Figure 1-1 illustrates an example.
Figure 1-1 uses the wrong formula and therefore fails to pass Peak A versus Peak B test case. In some cases, the Ethernet compliance test vendors have already updated their Ethernet compliance test suite firmware with the corrected formula. The engineer performing Ethernet compliance test cases may need to update the test equipment software with the latest software version from the manufacturer. For the update the vendor of the test equipment has to be contacted.
Contact the test equipment vendor for updates. Use this application note to determine whether or not the correct formula for the Peak A versus Peak B test is used by the test equipment.
The IEEE 802.3-2018, Subclause 40.6.1.2.1 is written as follows:
See Figure 40–20—Example of transmitter test mode 1 waveform (1 cycle) in the IEEE specification for further explanation of the points A and B:
The second paragraph of the specification description in section 40.6.1.2.1 contains the description of the math formula required to calculate “Peak A vs. Peak B”: The absolute value of the peak of the waveforms at point A and B shall differ by less than 1% from the average of the absolute value of the peaks of the waveform at point A and B.
The first part of the sentence “the absolute value of the peak of waveforms at point A and B” means that each of them individual waveforms need to comply to the next part of the sentence. It also means that signs of the measurement values have to be removed and the absolute value must be used.
The second part of the sentence “… average of the peaks of the waveforms at point A and B” refers to the following percentage error formula.
The percentage error formula is defined as follows:
E is the measured value, either Peak A or Peak B.
The theoretical quantity T is defined as the average of Va and Vb, which is calculated as:
When Peak A absolute value is Va and Peak B absolute value is Vb, the following formula of the IEEE specification and the percent error formula can be derived: