Noise suppression technologies/case study introduction (Automotive)

Automotive Ethernet for ADAS Noise suppression measures for 1000Base-T1 (4)

06/02/2020

INDEX

8. Testing of Immunity (DPI) Measures

A DPI (Direct Power Injection) test was conducted using the same 1000Base-T1 EMC test board as conducted emission.

Testing of Immunity (DPI) Measures_image

Frequency	1 - 1000MHz
Power(MAX)	39dBm
Power Step	0.5dB
EUT	1000Base-T1 EMC Test Board
DC Power Supply	GP035-5(Takasago)
Signal Generator	SML02(Rohde&Schwarz)
Power Amplifier	BSA1040-100 (1-400MHz) BLWA4010-100 (400-1000MHz)

Common mode noise is conducted out from an external source to the signal line on the 1000Base-T1 EMC test board, and a control PC is used to confirm whether any communication errors occurred.

Testing of Immunity (DPI) Measures_image2

In the same way as the conducted emission, the DLW32MH101XT2 (CMCC for 1000Base-T1), DLW43MH201XK2 (CMCC for 100Base-T1), and DLW32SH101XK2 (CMCC for CAN) were used as test CMCCs.

Differential Mode Transmission Characteristics_image — Sdd21（Differential Mode Transmission Characteristics）

Common Mode Transmission Characteristics_image — Scc21（Common Mode Transmission Characteristics）

Differential Mode Reflection Characteristics_image — Sdd11（Differential Mode Reflection Characteristics）

Mode Conversion Characteristics_image — Ssd12（Mode Conversion Characteristics）

1000Base-T1 DPI Test Results

At low frequencies of 2 MHz and less, the CMCCs exhibited differences in performance levels, but for other frequencies, there was no difference in performance, and all CMCCs satisfied the limit values.

1000Base-T1 DPI Test Results

The differences in the CMCCs at 2 MHz and less are thought to be due to common mode attenuation (Scc21). Differences in the mode conversion characteristics did not affect the results in the DPI test.

9. Key Points for Testing of Immunity (DPI) Measures

100Base-T1 DPI Test Results

After 1000Base-T1, the DPI test was also conducted for 100Base-T1. The CMCC for 100Base-T1 satisfied the limit values. However, the CMCC for CAN performed worse than the CMCC for 100Base-T1 at low frequencies of 1 MHz and less, and it also did not meet the limit values at 8 to 60 MHz, and failed the test.

The differences at 2 MHz and less are thought to be due to common mode attenuation (Scc21). Also, the differences at 8 to 60 MHz are thought to be due to the mode conversion characteristics.

Noise Entry Mechanism

One factor why the mode conversion characteristics of the CMCC affected the test results for 100Base-T1 is that the common mode noise that entered from an external source was converted to differential mode noise, which distorted the signal waveform, resulting in a communication error.

Notes on Board Design

In the same way as conducted emissions, in addition to the CMCC, this may also be caused by mode conversion due to imbalances on the board, and so careful attention is needed in board design.

10. Conclusion

For the automotive Ethernet standard 1000Base-T1, high performance is required in the CMCCs used for noise suppression, and the mode conversion characteristics are particularly important.
In evaluation of conducted emissions, a CMCC having mode conversion characteristics that meets the required values for 1000Base-T1 is needed for suppressing noise. The limit values are not satisfied when using CMCCs for CAN or for 100Base-T1.
Even in the CMCC for 1000Base-T1, the mode conversion characteristics can worsen due to variations in the board design and installed components, resulting in increased noise. For this reason, careful attention on this point is necessary in the design process.
In the DPI test, which is an immunity test, the required performance for the CMCC is lower than that for conducted emissions, but the noise resistance varies by the PHY, and so the one with lower mode conversion characteristics is preferred.