• Language
• English
• Deutsch
• Française
• Español
• Português
• Italia
• Русский
• 日本語
• 한국어
• 简体中文
• 繁體中文
}

# Design principles/network structure/parameter selection of EMI filters for DC power supplies.

• March 17, 2021
• 1377

Smart hardware will inevitably not avoid the EMI problem, and the power supply EMI filter is an effective method. Today, let’s talk about the design principles, network structure and parameter selection of EMI filters for DC power supplies.

1. Design principle-to meet the maximum impedance mismatch

The insertion loss should be increased as much as possible, that is, the signal reflection should be increased as much as possible. Suppose the output impedance of the power supply and the input impedance of the filter connected to it are respectively ZO and ZI. According to the signal transmission theory, when ZO≠ZI, reflection will occur at the input port of the filter, and the reflection coefficient

p=(ZO－ZI)/(ZO＋ZI)

Obviously, the greater the difference between ZO and ZI, the greater p, and the greater the reflection from the port, the more difficult it is for EMI signals to pass. Therefore, the input port of the filter should be in a mismatched state with the output port of the power supply, so that the EMI signal will be reflected. In the same way, the output port of the filter should be in a state of mismatch with the load, causing the EMI signal to reflect. That is, the design of the filter should follow the following principles:

If the internal resistance of the source is high impedance, the input impedance of the filter should be low impedance, and vice versa. If the load is high impedance, the output impedance of the filter should be low impedance, and vice versa.

For EMI signals, the inductance is high resistance, and the capacitor is low resistance. Therefore, the termination of the power supply EMI filter and the source or load should follow the following principles:

If the internal resistance or load of the source is resistive or inductive, the filter interface terminated with it should be capacitive. If the source resistance or load is capacitive, the filter interface terminated with it should be inductive.

2. Network structure of EMI filter

EMI signal includes common mode interference signal CM and differential mode interference signal DM, the distribution of CM and DM is shown in Figure 1. It can be used to guide how to determine the network structure and parameters of the EMI filter.

When selecting the power supply EMI filter structure, pay attention to the following issues:

Two-way filtering function-the power grid should have a filtering function for the power supply and the power supply for the power grid.

It can effectively suppress differential mode interference and common mode interference-the main consideration is the suppression of common mode interference in engineering design.

Satisfy the impedance mismatch principle to the greatest extent.

3. Parameter determination method of power supply EMI filter

(A) The value of the discharge resistance

When allowed, the smaller the resistance value is, the better, and the following situations need to be considered:

First, the resistance is required to be used with two-stage derating to ensure reliability. The derating factor is 0.75V, 0.6W. According to Ohm's law, n>(0.75Ve)2/(0.6Pe) can be obtained.

Second, there is residual voltage after lightning surge, and its instantaneous value is generally set at 1000V; its instantaneous power value cannot exceed 4 times the rated power value, and R>(Vcy)2/(4Pe) can also be calculated. The R value is considered comprehensively between the two. In general, the value of the resistance R is between 75-200K. The power is 2-3W. Metal film resistors.

(B) The value of Cx capacitor

If allowed, the larger the capacity requirement is, the better, and its value is difficult to accurately estimate. In general, the required value is between l-5uf (for each capacitor). The withstand voltage value of the capacitor must be taken after lightning surge. There is residual voltage, and its instantaneous value is generally not damaged at 1000V/s. It is selected according to the principle of secondary derating, and the value is 275V. The frequency characteristics and capacitance are selected. The value is related, the smaller the value, the better the frequency characteristic.

(C) The value of Cy capacitor

If allowed, the larger the capacity requirement is, the better, and its value is difficult to accurately estimate, but it cannot be too large, and the leakage current is too large. Under normal circumstances, the value is required to be between 2200-4700pf (for each Capacitors). The withstand voltage value of the capacitor must be taken after lightning surge. There is residual voltage, and its instantaneous value is generally not damaged at 1000V/S. It is selected according to the principle of secondary derating, and the value is 275V. Frequency characteristics and capacitor value Related, the smaller the value, the better the frequency characteristics. Cx capacitors and Cy capacitors are generally connected in parallel with smaller capacitors to meet the capacity requirements, so that the high frequency characteristics of the filter are good.

(D) The value of inductance

The selection principle of materials-consider from the following aspects:

First, the frequency range of the magnetic core material should be wide, and the highest frequency should be guaranteed at 1GHz, that is, there is a relatively stable magnetic permeability in a wide frequency range.

Second, the magnetic permeability is high, but it is difficult to meet this requirement in practice. Therefore, the magnetic permeability is often considered in sections. The core material is generally ferrite.

Estimation of inductance-considering impedance and frequency. The value of common mode choke is 1.5-5mH, and the value of differential mode choke is 10-50uH;

4. Installation requirements of EMI filter for DC power supply

The suppression effect of the filter on electromagnetic interference not only depends on the design of the filter itself and its actual working conditions, but also depends to a large extent on the installation of the filter.

The filter lead and installation location are also very important issues. This is to take into account that in addition to transmitting electromagnetic interference along the power line, the power line will also radiate electromagnetic interference during the transmission process, causing radiation coupling to nearby sensitive circuits (or equipment). Therefore, it must be considered that there is no coupling between the input line and the output line of the filter, otherwise the performance of the filter will be reduced. For this reason, the input line of the filter is best not to lead directly into the inside of the equipment, but to enter the inside of the equipment after filtering, and the natural shielding effect of the equipment casing is used to exclude the radiation field generated by the power supply from the outside of the equipment. The power input line should not be too long. The installation position of the filter should be close to the power supply population. Do not place the filter input line and output line too close. The filter should be well grounded, and it should not be grounded by a single wire, and should be in contact with a large area of metal casing. The pin or lead of the filter should be as short as possible. If the filter is to be connected to other circuits, it is best to use a connector.

To sum up

1. The metal shell of the filter must be in good contact with the chassis shell, and the grounding wire must be connected well.

2. The input line and output line of the filter must have a certain distance to avoid degradation of the filter performance.

3. It is better to use twisted pair for the connection of the filter, which can eliminate some high-frequency interference.

4. The installation location of the filter should be at the entrance of the power supply first to shorten the length of the input line in the chassis and reduce the radiation interference.

5. The selection of the filter mainly considers its insertion loss. The curve given by the manufacturer is generally taken from the test result of a 50Ω system, while the actual application system is generally not 50Ω. The insertion loss will have some deviations. In actual application, the power supply The filter is a band-pass filter, so pay attention to the influence of high-frequency characteristics.

6. The second point of filter selection is to pay attention to the working voltage, working current and leakage current, so as not to affect the use.

7. The low-frequency characteristic performance of the filter is related to the volume. If you want good low-frequency performance, you usually sacrifice the volume.

NXP

22+

NXP

19+/20+

NXP

21+

TI

22+

TI

22+

TI

22+

TI

22+

TI

21+