FILTER TOPOLOGY AND GROUNDING CONSIDERATION

Get Complete Project Material File(s) Now! »

Chapter 2 EMI Analysis

EMI Definition and Noise Propagation

Common-Mode and Differential-Mode Noise Characterization

The conducted EMI noise is usually decoupled and characterized as common-mode (CM) and differential-mode (DM) noise. CM noise is defined as the noise flowing between the power circuit and the ground, while DM noise is defined as the current following the same path as the power delivery as shown in Fig. 2-1. The same analysis applies to the power converter in Fig. 1-2. As explained in [27], it is common to separate the CM and DM noise from the measured noise with a noise separator. Many benefits are gained from this, including simplicity of the filter design, as each mode could be designed independently; and troubleshooting with the possibility to know directly which mode is not meeting the standard and therefore reduces the number of tries and errors during the design stage.
A simple model is used in Fig. 2-2 to understand the relationship between the CM and DM current. The noise source of the motor drive can be represented by voltage source V_N, while Z_g represents the impedance from the motor drive to the ground, and Z₁ and Z₂ represent the line impedance. As mentioned earlier, a LISN must be used to measure the EMI noise, and it is represented by its equivalent output impedance of 50 Ω per line [28].
A noise separator [29] is used in the research experiments to extract both CM and DM noise components, from the measure total voltage v₁ and v₂. The integrity of the noise separator is verified in a later section. It is essential to note that the EMI standard is dealing with the voltage v₁ and v₂ which imply that a margin of 6 dB below the limit is needed when looking at v_CM and v_DM to make sure that the total noise meets the standard.

Noise Propagation

For a typical motor drive system like the one shown in Fig. 1-2, the DM and CM noise are produced by the switching operations of the inverter, which is controlled by a PWM modulation scheme. While they have the same noise origin, the propagation path for these two modes is different. The DM noise transmitted through the power transmission line to the power source and motor. Conversely, the CM noise flows through the power lines to the ground via stray capacitances. In Fig. 1-2 C_H represents the stray capacitances between the IGBTs and the heat sink, while C_G is the equivalent capacitor between the motor frame and the ground. It is also important to take into account the stray capacitances of cables (not shown in the figure) between the inner conductors and the shield and ground. Experimentation shows that the stray capacitance of the motor is dominant, and is measured to be around 4-5 nF. On the other hand, the total stray capacitance of the IGBT module and heat sink are about 50 pF and the stray capacitance of the wire is 45 pF/m. This capacitance could be quite significant and become predominant for long cable applications such as those in aerospace where dozens of meters of cable are used [35]-[36]. From this analysis, it is clear that the EMI filter needs to be placed in a position where it will eliminate the three CM paths, even though the path from the IGBT to the heat sink is small. Therefore, placing the filter at the input on the DC link side is the best solution and should lead to a smaller design.

EMI Standards

Many standards exist to accommodate the wide variety of applications where EMI is an issue. Most of the standards differ either by their frequency range of application or the amplitude of the noise and whether the type of noise measured is voltage or current. They also have their own experimental and noise measurement setup as well as their own LISN circuit. It is common to see EMI standards beginning at 150 kHz and ending in the mega hertz range around 30 MHz, like the DO160 standard [30]. However this research is based on the Military Standard 461E described in [31] which starts at 10 kHz and end at 10 MHz.

Setup

Many experimental setups are defined by the Military Standard 461 E. However, for this research, the general setup shown in Fig. 2-3 is used. It is composed of a table covered with a ground plane where the LISN and the equipment under test (EUD) are placed. These two pieces of equipments need to be connected through a two-meter power line wire positioned on a non-conductive standoff of 5 cm height to avoid any disturbances from the ground.

Frequency Limit and Bandwidth

In contrast with other standards, the Military 461 E has a frequency range, limited from 10 kHz to 30 MHz. Fig. 2-5 defines the maximum noise limit for the conducted EMI noise. It is important to mention that this norm is given for a voltage of 28 V, and as the voltage increases some relaxation of this limit is permitted. For this research, the input voltage is set to 300 V, so another 10 dB of noise is allowed. On the other hand, this limit relaxation will be served as margin to compensate for the CM and DM voltage measurements via the noise separator. The amplitude boundary drops from 94 dBµV at 10 kHz to 60 dBµV at 500 kHz with a negative slope of 20dB/dec and then stays at 60 dBµV until 10 MHz.
Moreover, the spectrum analyzer connected to the LISN needs to be set with a certain bandwidth that is dependent on the frequency. As shown in Table 2-1, for a frequency range of 10 kHz to 150 kHz, the bandwidth set on the spectrum analyzer is 1 kHz while from 150 kHz to 30 MHz the bandwidth is 10 kHz. This helps to explain the discontinuity that will be observed in the future large signal results in the frequency domain.

Restrictions

Other restrictions and standards exist, such as the power characteristic given by the Military Standard 704 F [32], but this research focuses on the subject of EMI and does not consider other power standards. On the other hand, some constraints apply to the maximum common mode capacitance allowed in the EMI filter due to the grounding current safety standard. According to the SAE AS 1831 standards [33], the maximum capacitance value is set to 100 nF per line to ground “6.1.6 User Equipment AC Ground Isolation: The leakage capacitance to ground at the user interface shall not exceed the lower of 0.005 µF/kW of connected load or 0.1 µF measured at 1 kHz for each user equipment DC power and return line.” The filter design follows this standard to limit the CM current going through the ground as well as setting a baseline for comparison between the different versions of version.

CHAPTER 1 INTRODUCTION
I. Background
II. Literature Review and Motivations
III. Objectives & Scope of Work
IV. Thesis Outline and Summary of Contribution
CHAPTER 2 EMI ANALYSIS 
I. EMI Definition and Noise Propagation
II. EMI Standards
III. Experiment Setup
CHAPTER 3 BASELINE EMI FILTER DESIGN 
I. Introduction
II. Design Procedure
III. Material Consideration for Baseline Filter
V. EMI Filter Modeling
VI. Baseline Filter Large Signal Measurement
VII. Baseline Filter Size
VIII. Summary
CHAPTER 4 FILTER TOPOLOGY AND GROUNDING CONSIDERATION
I. Introduction
II. Filter Topology Consideration
III. Grounding Effect
IV. Summary
CHAPTER 5 APPROACHES TO MINIMIZING SIZE AND IMPROVING PERFORMANCE OF EMI FILTER
I. Introduction
II. CM Choke Core Material
III. CM Choke Parameters
IV. Integrated CM and DM Inductor
V. Capacitor Impact on Size and Performance
VI. Final Design Approach and Latest Filter Version
CHAPTER 6 SUMMARY AND CONCLUSIONS
REFERENCES .
GET THE COMPLETE PROJECT