With the popularity of automotive Ethernet, as a device supporting ADAS, a variety of sensors and cameras have been gradually configured in the car. LVDS and other interfaces are usually used for camera data transmission, while the case of vehicle Ethernet is gradually increasing when the data of lidar and other sensors are transmitted. Office Ethernet adopts 100base TX or 1000BASE-T standard, while automobile Ethernet adopts 100base-t1 or 1000base-t1 standard.
02 transmit signal through vehicle Ethernet
The vehicle interface of Ethernet adopts the differential transmission signal with less external noise or less radiation. The external noise usually appears on two differential transmission signal lines at the same time, which has no influence on the difference, so the differential transmission signal line has a strong ability to resist external noise. In addition, the paired signal lines are adjacent to each other, so the magnetic fields generated by the signal current cancel each other, so it is not easy to radiate noise to the outside.
Noise problem of 03 vehicle Ethernet
Although differential transmission line is not easy to produce noise, it will produce common mode current due to various factors, which will lead to noise problems.
Factors of common mode noise
The characteristic of differential transmission line is that it does not produce common mode noise, but if the signal of two lines has skew (time deviation) or amplitude deviation, then the signal equalization of two lines will be broken, resulting in common mode noise.
Noise reduction measures of 05 vehicle Ethernet
Problems of Ethernet
The cable used for Ethernet is different from HDMI, USB, etc. Different from paired signal wires, HDMI or USB cables are equipped with GND wires, so even if common mode current is generated, it will return through GND wires, so that the magnetic fields generated by common mode current cancel each other and are not easy to radiate. On the contrary, Ethernet is not equipped with GND line, so the common mode current returns through the stray capacitance to ground, which tends to radiate easily.
HDMI, USB, etc
Common mode choke (CMCC) is more effective in differential transmission noise reduction measures such as vehicular Ethernet. The common mode choke coils wind two wires on the same coil core in opposite directions. For the differential mode current, the magnetic flux generated by the two wires cancels each other, so it does not affect the differential current. For the common mode current, the magnetic flux generated by the two wires strengthens each other, so it can be used as an inductor. Therefore, the common mode choke can effectively reduce the common mode noise without affecting the differential signal.
06 precautions of CMCC for vehicle Ethernet
In the aspect of vehicle Ethernet, the balance of CMCC is also very important. If the length and winding mode of the two lines of CMCC deviate, the current balance will be broken and the mode will change, which may cause common mode noise. Therefore, CMCC with balanced design should be selected.
As the most suitable CMCC for 1000base-t1, dlw32mh101xt2 has been developed as a product. The product has the impedance value suitable for 1000base-t1, and adopts the balanced design which is not easy to occur mode conversion.
Transmission characteristics of dlw32mh101xt2
Countermeasures of 07 conducted emission
Measurement conditions of conducted emission
Murata used 1000base-t1 EMC evaluation board to measure conducted emission (150 Ω Law).
Comparison of noise reduction effect of different CMCC
Common mode noise is extracted from the signal line of 1000base-t1 EMC evaluation board and measured by EMI receiver.
The noise was compared by replacing CMCC.
CMCC used to reduce conducted emission
The CMCC used the product dlw32mh101xt2 for 1000base-t1, the comparison object used the product dlw33mh201xk2 for 100base-t1, and the can used the product dlw32sh101xk2.
Conducted emission measurement results
The conducted emission measurement results show that the noise suppression effect of dlw32mh101xt2 for 1000base-t1 is the most obvious and meets the limit requirements. However, dlw43mh201xk2 and dlw32sh101xk2 can not suppress the noise before meeting the limit requirements.
Noise generation mechanism
The reason why the noise suppression effect varies with CMCC may be due to the influence of the mode conversion characteristic ssd12 of CMCC. When the value of ssd12 is higher, the proportion of the input differential mode signal converted into common mode noise will increase, resulting in higher noise level.
Key points of noise reduction
The measurement results of conducted emission may be affected by the degree of common mode noise suppression of scc21 in low frequency and the degree of common mode conversion reduction of sd12 in high frequency.
Considerations of evaluation board design
Through the evaluation of CMCC, Murata understands the matters needing attention in the design of evaluation board.
After mounting the same CMCC sample for 1000base-t1 on the same evaluation board, the results show that the noise levels of the two are different, and one of the evaluation boards becomes ng.
Even for the same CMCC sample, ng may occur due to the status of the evaluation board.
Through the analysis of the transmission line characteristics of the evaluation board, it is found that the mode conversion characteristics of CMCC output are different, and the value of # 2 of the evaluation board is higher.
The reason why the conducted noise level varies with the evaluation board is that the differential mode signal after CMCC is converted into common mode noise on the evaluation board.
The factors that lead to mode switching may include the resistance of CMCC output, capacitor and circuit board lead, etc., which lead to the unbalanced phenomenon due to its characteristic deviation.
Therefore, for the parts other than CMCC, we should also pay attention to keep the characteristics of each line balanced.
Conducted emission of 100base-t1
After the same conducted emission measurement for 100base-t1, it is found that dlw32sh101xk2 for can exceeds the limit value, while dlw43mh201xk2 for 100base-t1 can effectively suppress the noise and meet the limit value requirements.
The difference between 100base-t1 and 1000base-t1
The frequency components of the differential mode signals of 100base-t1 and 1000base-t1 are different, so the required mode conversion characteristics are also different. Therefore, CMCC designed according to respective standards should be selected.
08 DPI test Countermeasures
Murata used the same 1000base-t1 EMC evaluation board as conducted emission to conduct DPI (direct power injection) test.
The common mode noise is injected into the signal line of 1000base-t1 EMC evaluation board from the outside, and the communication error is confirmed in the control PC.
Similar to conducted emission, CMCC for 1000base-t1, namely dlw32mh101xt2, dlw33mh201xk2 for 100base-t1, and dlw32sh101xk2 for can were used for evaluation.
DPI test results of 1000base-t1
In the low frequency band below 2MHz, although the level will be different due to different CMCC, there is no difference in other performances, meeting all the limits.
DPI test results of 1000base-t1
Below 2MHz, the difference due to CMCC may depend on scc21. The differences due to the mode switching characteristics did not affect the DPI test.
09 main points of DPI test Countermeasures
DPI test results of 100base-t1
After 1000base-t1, the DPI test results under 100base-t1 show that CMCC for 100base-t1 meets the limit requirements, while CMCC for can not only performs worse than CMCC for 100base-t1 in the low frequency band below 1MHz, but also is lower than the limit in the range of 8-60mhz, and the result is ng.
The difference in the frequency band below 2MHz may depend on the scc21 factor. In addition, the difference in 8-60mhz frequency band may depend on the mode conversion characteristics.
Noise intrusion mechanism
Under 100base-t1, the mode conversion characteristics of CMCC have an impact on the test results. The reason may be that the common mode noise injected from the outside is converted into differential mode noise, which distorts the signal waveform and leads to communication errors.
Considerations of evaluation board design
The same as conducted emission, except CMCC, the mode conversion caused by the unbalanced design on the evaluation board will also have an impact, so attention should be paid to the design of the evaluation board.
The vehicular Ethernet standard 1000base-t1 requires high performance of CMCC, especially mode conversion.
When evaluating conducted emission, it is necessary to prepare CMCC which meets the requirements of 1000base-t1 and has mode conversion characteristics to achieve the purpose of noise suppression. CMCC for can or 100base-t1 cannot meet the limit requirements.
Even if the CMCC used for 1000base-t1 is used, the performance of the mode conversion characteristics may be reduced and the noise may be increased due to the design of the circuit board or the deviation of the components to be mounted.
The performance of CMCC in dpi test is lower than that in conducted emission test, but the noise immunity varies with PHY, so it is recommended to choose CMCC with lower mode conversion characteristics.
CMCC for vehicular Ethernet is introduced in this paper
Editor in charge: PJ