Isolated active EMI (Electro-Magnetic Interference) filter without additional element on power line

Abstract

The present disclosure relates to an isolated active EMI filter without an additional element on a power supply line to prevent noise emitted through a power supply line cable, the isolated active EMI filter comprising: a common mode choke disposed on a power supply side; a Y capacitor disposed on the EMI source side; a sensing winding sensing a current; an amplifying unit that amplifies the noise current; and the signal of the secondary coil is injected into the transformer of the Y capacitor as a compensation signal.

Description

[0001] This application is a divisional application of an invention patent with a priority date of May 2, 2018, application number 201980029888.2, and invention name “An isolated active EMI filter without additional components on the power line”. Technical Field

[0002] The present invention relates to an EMI filter, and more particularly to an isolated active EMI filter without additional components on a power line. Background Art

[0003] In most home / industrial electrical systems, EMI filters are installed to prevent conducted EMI noise emitted through the power line cables.

[0004] In order to prevent common-mode conducted noise, a filter consisting of a common-mode choke and a Y capacitor is generally used. In high-power / high-current electrical systems, the self-saturation of the common-mode choke will reduce the noise reduction performance. To prevent this, a multi-stage filter or an expensive high-performance choke must be used to obtain sufficient attenuation performance, which results in a significant increase in the size and price of the filter. Therefore, attempts have been made to use active EMI filters, which can overcome the limitations of passive EMI filters and effectively improve performance, and preferably without the choke added to the power line due to the active EMI filter.

[0005] The active EMI filter has a feedback circuit structure that can sense the noise voltage or current through a capacitor or transformer and apply a compensation voltage or current back to the transformer or capacitor to eliminate it. However, if the active EMI filter senses and compensates the noise by adding a transformer to the power line, the performance of the high-power / high-current electrical system will be greatly reduced due to the self-saturation of the transformer. That is, in the related art, the active EMI filter without adding a transformer to the power line performs noise sensing and noise compensation through a capacitor.

[0006] However, when the active EMI filter connects a capacitor to the power line to sense and compensate for noise, the active circuit element is not isolated from the power line, so the reliability and stability against electrical overstress (EOS) are greatly reduced. In other words, a structure is needed that does not add a choke to the power line due to the active EMI filter and can isolate the active circuit element from the power line. Summary of the invention

[0007] Technical issues

[0008] The problem to be solved by the present invention is to provide an isolated active EMI filter without additional components on a power line, wherein the active circuit components of the EMI filter are isolated from the power line even if there are no additional components on the power line.

[0009] Another problem to be solved by the present invention is to provide a method for reducing EMI noise by using an isolated active EMI filter without additional components on a power line, wherein active circuit components are isolated from the power line even if there are no additional components on the power line.

[0010] Technical solutions

[0011] In order to achieve the above object, an isolated active EMI filter without additional components on the power line according to the first embodiment of the present invention comprises: a common mode choke, wherein two power lines connected to the EMI source of the common mode choke are respectively wound into windings; a sensing winding, wherein the sensing winding is rewound on the common mode choke through a coil and senses the noise current of the common mode choke; an amplifying unit, wherein the amplifying unit amplifies the noise current sensed by the sensing winding; a transformer, wherein the transformer receives the amplified signal from the amplifying unit and generates a compensation signal; and a Y capacitor, wherein the Y capacitor is connected between the transformer and the two power lines; wherein when the capacitance of the parasitic circuit of the common mode choke is C cm , the capacitance of the parasitic circuit of the sensing winding is C sen When the number of turns of the sensing winding is N sen Less than 2C cm / Cs en The square root of .

[0012] In order to achieve the above object, an isolated active EMI filter without additional components on the power line according to the second embodiment of the present invention includes: a common mode (CM) choke, which is arranged on the power supply side of the power supply, and the live line and the neutral line connected to the EMI source are respectively wound with windings; a Y capacitor, which is arranged on one side of the EMI source generating the EMI, and is composed of two capacitors connected in series, and the two capacitors are connected in parallel between the live line and the neutral line, and are connected to the ground; a transformer unit, which is installed at the front end of the Y capacitor, the primary coil senses the noise voltage of the Y capacitor, and the secondary coil transforms it and is isolated from the power line; an amplifier unit, which amplifies the noise voltage sensed by the transformer unit and transformed; a compensation winding, which is rewound on the common mode choke through a coil, and injects the noise signal amplified by the amplifier unit into the common mode choke.

[0013] In order to achieve the above object, an isolated active EMI filter without additional components on a power line according to a third embodiment of the present invention comprises: a common mode (CM) choke, which is arranged on the side of an EMI source generating EMI, and the live line and the neutral line connected to the EMI source are respectively wound with windings; a Y capacitor, which is arranged on the power supply side of the power supply and consists of two capacitors connected in series, the two capacitors are connected in parallel between the live line and the neutral line, and are connected to the ground; a sensing winding, which is rewound on the common mode choke through a coil and senses the noise current of the common mode choke; an amplifying unit, which amplifies the noise current sensed by the sensing winding; and a transformer, which is arranged at the front end of the Y capacitor, the primary coil receives the amplified signal from the amplifying unit, the secondary coil is connected to the ground connected to the Y capacitor and is isolated from the power line, and the signal transformed by the secondary coil is injected into the Y capacitor as a compensation signal.

[0014] In order to achieve the above object, an isolated active EMI filter without additional components on a power line according to a fourth embodiment of the present invention includes: a common mode (CM) choke, which is arranged on one side of an EMI source generating EMI, and a live line and a neutral line connected to the EMI source are respectively wound with windings; a Y capacitor, which is arranged on the power supply side of the power supply and consists of two capacitors connected in series, which are connected in parallel between the live line and the neutral line and are connected to the ground; a transformer, which is arranged at the front end of the Y capacitor, and a primary coil senses a noise voltage from the Y capacitor and transforms it through a secondary coil, and is isolated from the power line; an amplifying unit, which amplifies the noise voltage transformed by the transformer; and a compensation winding, which is rewound on the common mode choke through a coil, and injects the noise signal amplified by the amplifying unit into the common mode choke as a compensation signal.

[0015] In order to achieve the other purposes, a method for reducing EMI noise by using an isolated active EMI filter without additional components on a power line according to the first embodiment of the present invention is provided. The method is a method for reducing EMI noise by adding active components to a passive EMI filter. The passive EMI filter includes a common mode choke, wherein two power lines connected to an EMI source of the common mode choke are respectively wound into windings; a sensing winding, wherein the sensing winding is rewound on the common mode choke through a coil and senses the noise current of the common mode choke; an amplifying unit, wherein the amplifying unit amplifies the noise current sensed by the sensing winding; a transformer, wherein the transformer receives the amplified signal from the amplifying unit and generates a compensation signal; and a Y capacitor, wherein the Y capacitor is connected between the transformer and the two power lines; wherein when the capacitance of the parasitic circuit of the common mode choke is C cm , the capacitance of the parasitic circuit of the sensing winding is C sen When the number of turns of the sensing winding is N sen Less than 2C cm / Cs en The method comprises: a step of sensing the noise current of the common mode choke by forming a sensing winding by rewinding a coil on the common mode choke; a step of amplifying the noise current induced by the sensing winding; and a step of transforming the amplified signal received by the primary coil of the transformer installed at the front end of the Y capacitor through the secondary coil and injecting it into the Y capacitor, wherein the secondary coil of the transformer is connected to the ground connected to the Y capacitor, thereby being isolated from the power line.

[0016] In order to achieve the other objects, a method for reducing EMI noise by using an isolated active EMI filter without additional components on a power line according to a second embodiment of the present invention is provided, as a method for reducing EMI noise by adding active components to a passive EMI filter, wherein the passive EMI filter includes a common mode (CM) choke, which is arranged on the power supply side of the power supply and is connected to the live line and the neutral line of the EMI source. line) are respectively wound with windings; and a Y capacitor, the Y capacitor is arranged on one side of an EMI source generating EMI, and is composed of two capacitors connected in series, the two capacitors are connected in parallel between the live line and the neutral line, and are connected to the ground in common, and is characterized in that: the method comprises: a primary coil of a transformer installed at the front end of the Y capacitor uses the Y capacitor as a sensing capacitor to sense the noise voltage, and transforms the voltage through the secondary coil of the transformer; a step of amplifying the noise voltage transformed by the secondary coil; and a step of injecting the amplified noise signal into the common mode choke through a compensation winding, the compensation winding being formed by rewinding the common mode choke with a coil, wherein the secondary coil of the transformer is connected to the ground connected to the Y capacitor, thereby being isolated from the power line.

[0017] In order to achieve the other objects, a method for reducing EMI noise by using an isolated active EMI filter without additional components on a power line according to a third embodiment of the present invention is provided, as a method for reducing EMI noise by adding active components to a passive EMI filter, wherein the passive EMI filter includes a common mode (CM) choke coil, which is arranged on one side of an EMI source generating EMI, and is connected to a live line and a neutral line of the EMI source. line) are respectively wound with windings; and a Y capacitor, the Y capacitor is configured on the power supply side of the power supply, and is composed of two capacitors connected in series, the two capacitors are connected in parallel between the live wire and the neutral line, and are commonly connected to the ground, and is characterized in that: the method comprises: a step of sensing the noise current of the common mode choke by forming a sensing winding by rewinding the common mode choke with a coil; a step of amplifying the noise current sensed by the sensing winding; and a step of transforming the amplified signal input through the primary coil of the transformer installed at the front end of the Y capacitor through the secondary coil, and injecting it into the Y capacitor as a compensation signal, wherein the secondary coil of the transformer is connected to the ground connected to the Y capacitor, thereby being isolated from the power line.

[0018] In order to achieve the other objects, a method for reducing EMI noise by using an isolated active EMI filter without additional components on a power line according to a fourth embodiment of the present invention is provided, as a method for reducing EMI noise by adding active components to a passive EMI filter, wherein the passive EMI filter includes a common mode (CM) choke coil, which is arranged on one side of an EMI source generating EMI, and is connected to a live line and a neutral line of the EMI source. line) are respectively wound with windings; and a Y capacitor, the Y capacitor is arranged on the power supply side of the power supply, and is composed of two capacitors connected in series, the two capacitors are connected in parallel between the live wire and the neutral line, and are connected to the ground together, and is characterized in that: the method comprises: a step of sensing the noise voltage of the Y capacitor through the primary coil of a transformer installed at the front end of the Y capacitor, and transforming the voltage through the secondary coil; a step of amplifying the transformed noise voltage; and a step of injecting the amplified noise signal as a compensation signal into the common mode choke through a compensation winding, the compensation winding is formed by rewinding the common mode choke with a coil, wherein the secondary coil of the transformer is connected to the ground connected to the Y capacitor, thereby being isolated from the power line.

[0019] Effects of the Invention

[0020] In most household / industrial electrical and electronic devices, an EMI filter must be installed to prevent conducted EMI noise emitted through a power line cable, but according to the isolated active EMI filter without additional components on the power line and the method of reducing EMI noise using the filter of the present invention, the same noise reduction performance can be obtained with a smaller size and lower cost than the case of using only a passive filter.

[0021] In addition, according to the present invention, when conventional multi-stage passive EMI filters are used to sufficiently reduce noise, the number of filter stages can be reduced and the size and cost of most electrical and electronic equipment can be reduced by adding an isolated active EMI filter without additional components on the power line according to the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A circuit diagram showing the configuration of a first embodiment of an isolation type active EMI filter having no additional components on a power supply line according to the present invention.

[0023] Figure 2 An example of a configuration of an AEF according to the present invention shows a configuration of the proposed transformer-isolated AEF additionally (add-on) mounted to a CM LC EMI filter.

[0024] Figure 3A circuit model of an AEF according to an embodiment of the present invention is shown.

[0025] Figure 4 An equivalent circuit including a half portion of the parasitic component is shown.

[0026] Figure 5 An equivalent circuit model of a CM choke in the power line including the effect of the sensing winding is shown.

[0027] Figure 6a , 6b 6c shows the current path of the active EMI filter in each frequency domain and the capacitance effect of the Y capacitor (C Y , eff (s)).

[0028] Figure 7 shows the power line impedance (Z) when viewed from the Y capacitor position along the power supply direction. line )curve. Fig. 7A Shows N sen When Formula 19 is violated Figure 7B Shows N sen The situation when formula 19 is satisfied.

[0029] Figure 8 is a comparison of loop gain, where Fig. 8A It shows that without damping component R d1 , C d , R d2 And phase compensator R c , C c The loop gain in the unstable case is Figure 8B The loop gain for the stable case with these components present is shown.

[0030] Fig. 9 FIG. 1 is a circuit diagram of a second embodiment of an isolated active EMI filter having no additional components on a power line according to the present invention.

[0031] Fig.10 FIG. 1 is a circuit diagram of a third embodiment of an isolated active EMI filter having no additional components on a power line according to the present invention.

[0032] Fig.11 FIG. 4 is a circuit diagram of a fourth embodiment of an isolated active EMI filter having no additional components on a power line according to the present invention.

[0033] Fig.12 A flow chart showing a method for reducing EMI noise by adding active elements to a passive EMI filter, which corresponds to a first embodiment of an isolated active EMI filter without additional elements on a power line according to the present invention.

[0034] Fig.13 A flow chart showing a method for reducing EMI noise by adding active elements to a passive EMI filter, which corresponds to a second embodiment of an isolated active EMI filter without additional elements on a power line according to the present invention.

[0035] Fig.14 A flow chart showing a method of reducing EMI noise by adding active elements to a passive EMI filter, which corresponds to a third embodiment of an isolated active EMI filter without additional elements on a power line according to the present invention.

[0036] Fig.15 A flow chart showing a method for reducing EMI noise by adding active elements to a passive EMI filter, which corresponds to a fourth embodiment of an isolated active EMI filter without additional elements on a power line according to the present invention.

[0037] Preferred specific implementation

[0038] The present invention relates to an isolated active EMI filter without additional components on a power line and a method for reducing EMI noise using the isolated active EMI filter, wherein the isolated active EMI filter includes a common mode (CM) choke configured on the power supply side of the power supply; a Y capacitor; a sensing winding rewound on the common mode choke through a coil and sensing the noise current of the common mode choke; an amplifying unit amplifying the noise current sensed by the sensing winding; and a transformer, wherein the transformer is arranged at the front end of the Y capacitor, the primary coil receives the amplified signal from the amplifying unit, the secondary coil is connected to the ground connected to the Y capacitor and isolated from the power line, and the signal of the secondary coil is injected into the Y capacitor as a compensation signal. DETAILED DESCRIPTION

[0039] Hereinafter, the preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Since the embodiments described in this specification and the configurations shown in the accompanying drawings are only preferred embodiments of the present invention and do not represent the entire technical spirit of the present invention, it should be understood that various equivalent forms and variations of the embodiments may be substituted for them for this application.

[0040] Figure 1 The configuration circuit diagram of the first embodiment of an isolated active EMI filter without additional components on the power line according to the present invention is shown. The first embodiment of the present invention includes a passive EMI filter composed of a common mode (CM) choke 110 and a Y capacitor 120, and an EMI filter 100 composed of a sensing winding 130, an amplifying unit 140 and a transformer 150.

[0041] The common mode (CM) choke 110 is provided at the power supply side of the power supply, and a live line and a neutral line connected to an EMI source are respectively wound with windings.

[0042] The Y capacitor 120 is disposed at one side of the EMI source generating the EMI, and is composed of two capacitors connected in series. The two capacitors are connected in parallel between the live line (L) and the neutral line (N), and are connected to the ground.

[0043] The sensing winding 130 is rewound on the common mode choke 110 through a coil, and senses the noise current flowing in the common mode choke 110. When the capacitance of the parasitic circuit of the common mode choke 110 is C cm, the capacitance of the parasitic circuit of the sensing winding 130 is C sen, and the number of turns (N sen) of the sensing winding 130 is preferably less than the square root of 2C cm / C sen.

[0044] The amplification unit 140 amplifies the noise current sensed by the sensing winding 130 .

[0045] The transformer 150 is arranged at the front end of the Y capacitor 120, the primary coil receives the amplified signal from the amplifying unit, the secondary coil is connected to the ground connected to the Y capacitor 120 and is isolated from the power line, and the signal of the secondary coil is injected into the Y capacitor 120 as a compensation signal.

[0046] Figure 2 An example of a configuration of an AEF according to the present invention shows a configuration of the proposed transformer-isolated AEF additionally (add-on) mounted to a CM LC EMI filter. Figure 3 A circuit model of an AEF according to an embodiment of the present invention is shown.

[0047] According to the present invention, the isolated active EMI filter without additional components on the power line may further include a low pass filter to prevent stability problems caused by winding resonance in the high frequency range (refer to Figure 3 The low-pass filter consists of a resistor (R f ) and a capacitor (C f ).

[0048] Resistance (R f ) is connected to the sensing winding at one end, and the other end is connected to the + input terminal of the amplification unit. The capacitor (C f ) is connected to the resistor (R f ) and the +input terminal of the amplifying unit, and the other end is connected to the ground and is located at the input end of the amplifying unit.

[0049] In addition, preferably, within the frequency range of interest, the impedance (Z in，AEF ) is set to be higher than the parasitic RC component impedance (Z sen，para ).

[0050] Preferably, the cut-off frequency (1 / 2ΠR f C f ) is greater than the maximum operating frequency (f op,max ), and is less than the frequency 1 / 2π√(1-k sen 2 )N sen 2 L cm C sen (k sen is the coupling coefficient of the sensing winding 130; N sen is the number of turns of the sensing winding 130; L cm is the inductance of the common mode choke 110; C sen is the capacitance of the parasitic circuit of the sensing winding 130).

[0051] In addition, the isolated active EMI filter without additional components in the power line according to the present invention may also include a bypass branch, which has the stability of being a bypass and damping circuit to avoid resonance in the transformer and alleviate the performance degradation caused by the resonance between the EMI source impedance and the Y capacitor. (Refer to Figure 3 The bypass branch may include a first resistor (R d1 ), capacitor (C d ) and the second resistor (R d2 ).

[0052] First resistor R d1 One end of the capacitor is connected to the Y capacitor, and the other end is connected to the secondary coil of the transformer. d One end of the second resistor R is connected to one end of the resistor. d2 One end is connected in series with the other end of the capacitor, and the other end is connected to the ground.

[0053] In addition, the isolated active EMI filter without additional components in the power line according to the present invention may further include a phase compensator for maintaining stability in the low frequency range (refer to Figure 3 ).

[0054] The phase compensator includes a resistor (Rc) and a capacitor (Cc) connected in parallel, and one end of the resistor (Rc) and the capacitor (Cc) connected in parallel is connected to the (-) input terminal of the amplification unit, and the other end of the resistor (Rc) and the capacitor (Cc) connected in parallel is connected to the output terminal of the amplification unit.

[0055] The present invention proposes a new structure of a fully transformer-isolated AEF. Figure 2 , an AEF according to one embodiment of the present invention is additionally installed (add-on) to an existing CM LC EMI filter composed of a CM choke and a Y capacitor. The structure of the AEF according to one embodiment of the present invention is similar to the traditional CSCC AEF topology, but an injection transformer is added between the output of the amplification unit and the compensation Y capacitor. Since the injection transformer is not installed on the main power line, only a small amount of compensation signal current flows through the transformer. The injection transformer can be implemented in a small size because its current is small and has nothing to do with the operating current of the application, thereby reducing the risk of magnetic saturation and thermal problems. In addition, the sensing part of the AEF does not require an additional transformer, but a thin noise sensing line is additionally wound on the existing commercial CM choke. Attempts have been made to add the sensing winding directly to the commercial CM choke, but the adverse effects and the maximum allowable number of turns of the sensing winding have not yet been studied. In summary, the main new feature of the AEF according to one embodiment of the present invention is that the main power line is fully transformer-isolated without the use of separate components, and a specific compact sized design can be adopted. Due to these characteristics, the AEF according to the present invention has a smaller size and better performance compared to other transformer-isolated CSVC AEFs.

[0056] The present invention provides many useful and clear design guidelines for the complete design of an AEF according to an embodiment of the present invention. As described below, the transformer-isolated AEF is analyzed to evaluate the noise attenuation performance, on which basis appropriate design guidelines are provided for the performance and stability of the AEF, and the filter insertion loss and loop gain of the AEF are measured and verified by a vector network analyzer (VNA). The reduction of CM CE noise by the AEF can also be seen on the actual product SMPS board. In addition, the leakage current to ground is measured to confirm the safety of the use of the AEF.

[0057] The AEF according to one embodiment of the present invention will be analyzed below. Figure 3In the figure, CY represents the capacitance of the Y capacitor. The CM choke is modeled with Lcm and Mcm, which represent the self-inductance and mutual inductance of the windings on the power line.

[0058] The AEF is mainly composed of a sensing winding wound on a CM choke, an amplification unit, and an injection transformer. The winding ratio of the power line winding to the sensing winding is set to 1:N. sen , the self-inductance of the sensing winding is about N sen 2 L cm . M sen represents the mutual inductance between the winding of the power line and the sense winding of the AEF input. Similarly, M inj represents the mutual inductance injected into the transformer, and the ratio of the primary and secondary windings is set to 1:N inj The self-inductance of each side is L inj and N inj 2 L inj . M cm 、M sen and M inj k cm L cm , k sen N sen L cm and k inj N inj L inj Calculate. Here, k cm , k sen and k inj represents each coupling coefficient. In practical design, k cm , k sen and k inj The value of is usually in the range of 0.99 to 1. The amplifier unit is configured with a resistor R 1 and R 2 A non-inverting operational amplifier.

[0059] Considering the feedback stability of AEF, several additional components are required, such as low-pass filter, bypass branch and phase compensator, such as Figure 3 As shown. f and C f A low-pass filter is formed in the operational amplifier to prevent stability problems due to resonance of the sense winding in the high frequency range. d1 , C d and R d2 For the stability of the bypass branch and damping circuit, it is necessary to avoid resonance in the transformer, thereby further alleviating the performance degradation caused by the resonance between the noise source impedance and the Y capacitor. c and Cc It is a phase compensator that ensures stability in the low frequency range.

[0060] Even if the ground reference voltage of AEF is set to something different than ground, AEF will be symmetrical with respect to the AC zero potential and can be analyzed by voltage division of the circuit. Figure 4 An equivalent circuit including half the parasitic components is shown. Figure 4 , for a more accurate display, the parasitic circuit parameters of the CM choke, sense winding, and injection transformer are also modeled, including R cm , C cm , R sen , C sen , R inj1 , C inj1 , R inj2 and C inj2 The CM noise sources of the equipment under test (EUT) are modeled as the Thevenin equivalent circuit V n and Z n , to represent the voltage and impedance of the CM noise source. LISN Represents the impedance of the Line Impedance Stabilization Network (LISN). line , Z line , AEF and Z Y , eff Respectively represent the impedance relative to ac zero potential in each direction.

[0061] According to the expressions of the effective inductance of the CM choke and the effective capacitance of the Y capacitor branch, the working principle of AEF is analyzed and the effective inductance of the CM choke is explained.

[0062] Using Kirchhoff's law, the impedance from the front end of the CM choke toward the power line is as follows.

[0063] [Formula 1]

[0064]

[0065] in,

[0066] [Formula 2]

[0067] L cm，eff (s) = (1 + k cm -X(s))L cm

[0068]

[0069] [Formula 4]

[0070]

[0071] [Formula 5]

[0072]

[0073] L cm，eff represents the effective inductance of the CM choke, in which the inductance cancellation term X(s) appears. X(s) is defined as (2M sen I sen ) / (L cm I cm ), where I cm and I sen Respectively Figure 4 Through the inductor branch (L cm +M cm ) and 2N sen 2 L cm The current. Z sen，para Z represents the parasitic RC component impedance of the sensing winding. in，AEF It represents the impedance from the input of the amplifier unit toward the low-pass filter. Assuming that the input impedance of the operational amplifier is larger than the target frequency range, it is ignored in Formula 5.

[0074] Figure 5 An equivalent circuit model of a CM choke in the power line including the effect of the sensing winding is shown. Figure 5 The effect of the sense winding on the inductance of the CM choke is summarized. Figure 5 , the right box shows the consideration of 2M sen Equivalent circuit model of CM choke in the power line of the induced voltage. Induced voltage 2sM sen I sen The polarity and s(L cm +M cm )I cm The polarity of the voltage drop is opposite. If X(s) is defined as (2M sen I sen ) / (L cm I cm ), the total voltage of the choke inductor can be simplified to s(1+k cm -X(s))L cm I cm Therefore, as shown in Equation 2, the effective inductance of the CM choke, L cm，eff It is expressed as (1+k cm -X(s))L cm .

[0075] If there is no sense winding, then k sen=X(s)=0, so L cm，eff It can be simply expressed as (1+k cm )L cm However, when sN in Formula 3 sen 2 L cm Much higher than (Z in,AEF ||Z sen,para ), k cm and k sen Very close to 1, so X(s)≈2k sen 2 , L cm,eff ≈L cm (1+k cm -2k sen 2 ). This means that the current flowing through the sense winding can have a decisive influence on the choke inductance. Therefore, in order to maintain the choke inductance, the number of turns N of the sense winding must be limited. sen .

[0076] Next, the effective capacitance of the Y capacitor will be described. Y，eff The impedance can be expressed by Equations 6 to 11.

[0077] [Formula 6]

[0078]

[0079] in,

[0080] [Formula 7]

[0081]

[0082] [Formula 8]

[0083]

[0084] [Formula 9]

[0085]

[0086] [Formula 10]

[0087]

[0088] [Formula 11]

[0089]

[0090] Here, α(s) and β(s) can be understood physically as a boosting factor and a bypass factor, respectively, as described later. 1(S) is from V in To V in，amp The voltage gain, G amp (s) is from V in，amp To V out，amp The gain of the amplifier unit. Assume that the frequency bandwidth of the operational amplifier is sufficiently higher than the target frequency range. Z in Formula 6 Y，eff The expression can be understood as the effective capacitance C Y，eff The impedance of (S) is defined as shown in Formula 12.

[0091] [Formula 12]

[0092]

[0093] Figure 6a , 6b and 6c show the C changes with frequency Y , eff (s) changes, Figure 6a shows the AEF operation over the frequency range, Figure 6b A plot showing α(s) and β(s) is shown. Figure 6c Shows Z Y，eff The impedance curve.

[0094] exist Figure 6a , 6b and 6c summarize the frequency-dependent C Y , eff (s) changes. Figure 6a The influence of AEF on the Y capacitor branch and the change of current path are shown in Figure 2. The dashed box shows the influence of AEF on C Y , eff The impact of op，min and f op，max are the minimum and maximum target operating frequencies of the AEF that can be designed through the circuit parameters of the AEF, respectively.

[0095] For example, according to Figure 6b The magnitude of α(s) and β(s) for a properly designed AEF is plotted against the frequency in . Figure 6c The impedance Z of the Y capacitor branch is also plotted. Y，eff . At sufficiently lower f op，min At frequencies of , α(s) and β(s) are both much less than 1, and Z in Formula 6 Y，eff This means that the bypass circuit and injection transformer are negligible compared to the impedance of CY, and the noise voltage compensated by AEF is very small. op，min to f op，max In the AEF operating frequency range, α(s) is greater than 1, but β(s) is still much less than 1. That is, AEF provides the compensation voltage to the Y capacitor branch, such as Figure 6a -αV in As shown, the bypass circuit can still be ignored.

[0096] The size of α(s) is mainly maintained at the product of the voltage gain of the sensing winding, the amplification unit and the injection transformer within the operating frequency range, that is, N sen N inj (1+R 2 / R 1 ) range. Therefore, the CM current flowing through the Y capacitor branch is calculated according to (1+N sen N inj (1+R 2 / R 1 ))Enlarge, such as Figure 6c As shown, the effective capacitance increases to (1+N sen N inj (1+R 2 / R 1 ))C Y As the frequency increases to near f op，max , α(s) starts to decrease, which indicates that the compensation voltage from AEF decreases. At the same time, β(s) is similar to 1, which means that the impedance of the bypass branch (R d2 +1 / sC d ) is lower than the impedance of the injection transformer path. Therefore, the CM noise current mainly flows through the bypass branch, Z Y，eff The impedance is approximately (1 / sC Y +2(R d2 +1 / sC d )).

[0097] Figure 6c Displayed in f op，max After the frequency range, because C d and R d2 is added to the current path, so Z Y，eff The size is relative to 1 / sC Y However, the damping resistor R d2 It plays an important role in alleviating the resonance between the Y capacitor and the CM noise source impedance. Because the performance of the entire CM filter is usually greatly reduced by resonance, resonance should be avoided.

[0098] Next, the insertion loss of the entire filter will be described. The noise attenuation performance of a filter is usually quantified as insertion loss (IL), which is defined as the ratio of the noise voltage received by the LISN without the filter to the noise voltage of the LISN with the filter installed. Figure 4 , the IL of the entire EMI filter is derived as shown in Equation 13.

[0099] [Formula 13]

[0100]

[0101] As the frequency increases, the insertion loss IL in formula 13 mainly occurs at Z Y，eff becomes smaller than Z line The low-frequency boundary of the filter operation can be approximated as like Figure 6c As shown, the proposed AEF significantly reduces the Z Y，eff , which increases the IL of the entire filter. In addition, AEF allows the entire filter to start operating at a lower frequency.

[0102] On the other hand, the design guidelines of the AEF provided by the present invention will be described. Considering the performance and stability, practical design guidelines are formulated for the AEF. First, the design of the sensing winding and the input low-pass filter will be described.

[0103] The sense winding is wound directly on the CM choke, so there is no need to add an additional sense transformer. In terms of size and cost, it is preferred not to use a separate sense transformer. However, as mentioned above, the CM choke inductance L can be reduced by an additional sense winding. cm，eff and the power line impedance Z line Even though the AEF according to the present invention improves the Y capacitance C Y，eff , reducing the power line impedance Z line It will also deteriorate the noise attenuation performance of the entire CMEMI filter. line To reduce this, appropriate design guidelines for the sense winding are required.

[0104] Z line The first self-resonant frequency f of the CM choke r，cm (Given as ) is mainly determined by parasitic capacitance. Therefore, the frequency f r，cm The inductance cancellation term X(s) then affects Z line In addition, only when sN sen 2 L cm Higher than (Z in,AEF ||Z sen,para ), X(s) in Formula 3 is meaningful. Therefore, if sN sen 2 L cm Begins to become greater than (Z in,AEF ||Z sen,para ) is higher than f r，cm , then Z lineThese conditions can be met by the following design procedure. First, Z in，AEF Designed to be higher than Z sen，para That is, it can be expressed as the condition of formula 14.

[0105] [Formula 14]

[0106]

[0107] In simpler terms, R f and C f The selection of is shown in Equations 15 and 16.

[0108] [Formula 15]

[0109]

[0110] [Formula 16]

[0111] C f <C ser

[0112] If the inequality condition in formula 14 is fully satisfied, then (Z in,AEF ||Z sen,para )≈Z sen,para Then, sN sen 2 L cm Begins to become greater than (Z in,AEF ||Z sen,para ) is approximately equal to N sen 2 L cm and C sen The resonant frequency between is shown in Equation 17.

[0113] [Formula 17]

[0114]

[0115] Then, due to f r，sen Must be higher than f r，cm , as shown in formula 18,

[0116]

[0117] A design guideline for the number of turns of the sensing winding is derived as shown from Equation 19.

[0118] [Formula 19]

[0119]

[0120] Here, L cm,eff Approximately (1+k cm )L cm ≈2L cm . It is guaranteed to maintain Z line The design guide for Equation 19 is derived by sensing the maximum allowable winding of the winding. The CM choke and C cm ,C sen The exact value of the parasitic capacitance is not actually known prior to design, but Equation 19 can still provide a useful guide for the number of turns of the sense winding.

[0121] Figure 7 shows the power line impedance (Z line ) curve. Fig. 7A Indicates that when N sen When Formula 19 is violated, (b) means N sen The situation when formula 19 is satisfied. The numerical example is shown in Figure 7. By setting C cm and C sen Set the same fixed value to design AEF, which is used to sen The values ​​of each Z were compared between the case without AEF and the case with AEF. line The size of Z linew / o,AEF和 Z linew / ,AEF .exist Fig. 7A In, N sen =2 violates conditional formula 19, and f r，sen Below f r，cm .on the contrary, Figure 7B In the formula, Nsen=0.5 satisfies formula 19, and f r，sen Higher than f r，cm As a result, Fig. 7A In, with Z line,w / oAEF In comparison, Z linew / ,AEF Significantly reduced, while Figure 7B In, Z linew / ,AEF Almost unchanged.

[0122] In addition, when using AEF, Z line Another resonance occurs. Fig. 7A and B, through the sensing winding in Since this resonance will adversely affect the stability of the system feedback in the high frequency range, it is necessary to connect R f and C f For a low-pass filter that does not affect the AEF performance within the operating frequency range, the filter cutoff frequency should be greater than the maximum operating frequency f op，max, but should be less than the resonant frequency of Formula 20.

[0123] [Formula 20]

[0124]

[0125] Equation 15, Equation 16, and Equation 20 can be a design guide for the low-pass filter.

[0126] Next, the design of the injection transformer and the amplification unit will be described.

[0127] The design of the injection transformer and the amplifier unit components mainly determines the main performance parameters f of the AEF in Figure 6 op，min , f op，max and C y，eff The capacitor C at the output of the amplifier unit o Used to block harmful signals with frequencies below the target operating frequency range. inj Series C o A high-frequency filter is constructed, and its cut-off frequency is derived as shown in Formula 21 to determine the minimum operating frequency of AEF.

[0128]

[0129] In f op,min frequency, with L inj Series connected C o The impedance of R decreases rapidly, thus increasing the output current of the op amp. o is added to the op amp output to limit the impedance at the resonant frequency, but it must be sufficiently smaller than sL over all operating frequency ranges. inj .

[0130] At the same time, if Figure 6a As mentioned in the above, the maximum operating frequency of AEF is f op，max The frequency limit at which the impedance of the bypass branch is lower than the impedance of the injection transformer path is determined. Similar to the resonance caused by the inductive winding shown in Figure 7, the resonance in the secondary winding of the injection transformer will cause feedback instability, so the bypass branch must start operating at a frequency below the resonant frequency. As shown in Equation 22, the resonance of the secondary winding occurs above f op，max的 frequency superior.

[0131] [Formula 22]

[0132]

[0133] According to the inductance part of the injection transformer (1-k inj 2 )N inj2 L inj With bypass branch capacitor C d The resonance between the two determines f op,max , as shown in Formula 23.

[0134] [Formula 23]

[0135]

[0136] Substituting Equation 23 into Equation 22, we get C d and C inj 2 The relationship between them is as follows.

[0137] [Formula 24]

[0138] C d >C inj2

[0139] In order to maintain stability at high frequencies, some damping resistor R is required d1 and R d2 It is recommended to use R of tens of ohms. d2 To alleviate the resonance between the Y capacitor and the CM noise source impedance in the high frequency range, which will be shown experimentally in Chapter IV.

[0140] Assume that the resistance R in AEF operation can be ignored except for the resonance point. d1 and R d2 and R inj 2 The influence of the formula 24, the formula 7 to formula 11 can be changed at f op,max The frequency range is estimated according to Formula 25.

[0141] [Formula 25]

[0142] α(s)≈N sen N inj G amp (s), β(s)≈0

[0143] In formula 12, the Y capacitor and effective capacitor C Y,eff (s) is simplified by formula 26.

[0144] [Formula 26]

[0145] C Y，eff (s)≈(1+N sen N inj G amp (s))C Y

[0146] G amp (s) The phase compensation element Rc and C c should have little impact on the AEF operation and Equation 26 further simplifies to a frequency independent value as shown in Figure 6.

[0147] [Formula 27]

[0148] C Y，eff ≈(1+N sen N inj (1+R 2 / R 1 ))C Y

[0149] Finally, several useful design guidelines for AEF can be derived as follows. sen Limited by Formula 19, but in Formula 27, it can be increased by increasing N inj and the gain of the amplifier unit (1+R 2 / R 1 ), C Y，eff Designed for C Y However, as N inj The increase in the maximum operating frequency f op，max will be reduced according to Equation 23. In addition, high amplifier gain requires the op amp to have a large output voltage swing and a large gain bandwidth. Therefore, the cost of the op amp and the f of AEF should be considered. op，max , choose a suitable N inj and (1+R 2 / R 1 )value.

[0150] In addition, the condition of Equation 22 means that N inj , L inj and C inj 2 To adjust the AEF f op，max Due to the parasitic capacitance C inj 2 is not an independent design parameter, so N inj and L inj Designed to be smaller to achieve higher f op，max However, for smaller N inj Make C y，eff Reduce, L inj decreases, then f in Formula 21 op，min will increase. Therefore, we propose the following design process to optimize the performance of AEF. First, in a physical package of a given size, C o Designed to be as large as possible, and for the purpose of op，min L injReduce to the limit of Formula 21. Next, for the target f op，ma'x N inj Increase to the limit of Equation 22 to obtain the maximum C Y，eff .

[0151] Next, the stability check will be described.

[0152] The AEF is essentially a feedback system with analog input and analog output that must be carefully designed and guaranteed to be stable. If the system is unstable, it will oscillate even if no EUT noise source is applied. Feedback stability can be verified by the phase and gain margin of the loop gain. Figure 4 The loop gain is derived from the circuit model of , separating the feedback loop from the output of the op amp and connecting the test voltage source V from the separation node t applied to the injection transformer, and without applying the noise source voltage V n In the state, the voltage V in With test voltage V t The ratio can be calculated according to formula 28.

[0153] [Formula 28]

[0154]

[0155] in,

[0156] [Formula 29]

[0157]

[0158] (V in,amp / V in ) and (V out,amp / V in,amp ) is derived as G in equations 9 and 10 respectively. 1 (s) and G amp (s). Therefore, the loop gain of the system can be expressed as Equation 30.

[0159] [Formula 30]

[0160]

[0161] Using R c and C c The purpose is that the effective inductance L of the choke cm,eff and the effective capacitance C of the Y capacitor branch Y,eff The resonance between them will bring the risk of instability, so for the stability of the low frequency range, increase G loop (s) Phase margin. Resonant frequency It determines the low frequency boundary of the filter operation and should be lower than the low frequency limit of the CE standard in proper EMI filter design. As shown in Equation 31, the calculation of the c and C c The maximum amount of phase compensation caused.

[0162] [Formula 31]

[0163]

[0164] Formula 31 at frequency

[0165]

[0166] The following happens.

[0167] ∠G loop，wlocomp (S) represents G with phase compensator loop (s), and ∠G loop，wlocomp (S) represents G without phase compensator loop By setting the maximum phase compensation frequency to The resonant frequency of R c and C c Other expressions of are as shown in Formula 32.

[0168] [Formula 32]

[0169]

[0170] Formulas 31 and 32 are expressed as R c and C c Design guidelines are provided.

[0171] Since G in Formula 28 2 (s) According to the EUT noise source impedance Z n , so it should be noted that the loop gain in Equation 30 is also the same. In Equation 30, it can be seen that as Z n As the value of increases, the loop gain increases, so the gain margin tends to decrease. n Designing for stability at infinity usually provides stability under worst-case conditions. Therefore, in this specification, the value of Z at infinity is n The loop gain of the designed AEF is calculated or measured under the specified conditions to ensure stability in any EUT application.

[0172] Figure 8 is a comparison of loop gain, where Fig. 8A It shows that without damping component R d1 , C d and R d2 And phase compensator Rc , C c The loop gain in the unstable case is Figure 8B The loop gain for the stable case with these components is shown.

[0173] For example, the G of a filter with AEF can be expressed using Equation 30 loop (s), as shown in Figure 8. Bypass branch and phase compensator R d1 , C d , R d2、 R c , C c Not installed on Fig. 8A but already installed in Figure 8B Their influence on stability has been clearly shown. Fig. 8A Instability is caused by a sudden phase shift at about 10 MHz due to injection into the secondary winding of the transformer and in Figure 8B In the case of, the bypass branch solves the problem. Fig. 8A In the case of low frequencies less than 100kHz, L cm,eff and C Y,eff Resonance between the two can cause excessive phase shift and risk of instability. Figure 8B It can be seen that using the phase compensator R c and C c Greatly increases the gain margin.

[0174] Next, the selection and overall design steps of the operational amplifier will be described. The high frequency limit of the operational amplifier in the non-inverting amplifier is f OPamp Must be higher than the high frequency limit f of the CE standard CE,max .

[0175] [Formula 33]

[0176] f OPmp >f CE，ma x

[0177] In addition, the voltage and current capabilities of the op amp must be sufficient to compensate for the noise. To calculate the required op amp capabilities, use Equation 34 and Equation 35, respectively, based on Figure 4 The circuit model calculates the voltage V at the output of the operational amplifier out,amp (s), current I out,amp (s).

[0178] [Formula 34]

[0179] V out,amp (s) = V in (s)G 1 (s)G amp (s)

[0180] [Formula 35]

[0181]

[0182] in,

[0183] [Formula 36]

[0184]

[0185] Since V in (s) is determined not only by the filter impedance including AEF, but also by Z n and V n (s) is determined, so noise source model information is needed to estimate V out,amp (s) and I out,amp (s). Various measurement methods have been developed to derive the operating SMPS, Z n and V n (s) noise source model. Derive Z n and V n (s), the operational amplifier output voltage v out,amp (t) and output current i out,amp The time domain waveform of V (t) can be obtained by using the V out,amp (s) and I out,amp. Therefore, the voltage capacity v of the operational amplifier output OPamp,max And current capacity i OPamp,max , should be sufficient to provide the calculated v out,amp (t) and i out,amp (t).

[0186] [Formula 37]

[0187] v OPamp,max >max(|v out,amp (t)|)

[0188] [Formula 38]

[0189] i OPamp,max >max(|2i out,amp (t)|)

[0190] Since I is defined in the half-circuit model out,amp (s), so the actual current flowing through the op amp is twice the calculated current, as shown in Equation 38.

[0191] As shown in formula 35, in the influence of I out,amp (s) In the various design factors, increase N injThis will greatly increase the I out,amp (s). inj The voltage gain induced by the injection transformer can reduce the output voltage of the op amp, as opposed to increasing the output current as described above. The injection transformer not only isolates the AEF ground from the SMPS ground, but also provides additional design flexibility for the gain and op amp circuits.

[0192] Other embodiments of an isolated active EMI filter without additional components on the power line according to the present invention will be described. Fig. 9 1 is a circuit diagram of a second embodiment of an isolated active EMI filter without additional components on a power line according to the present invention. The second embodiment of an isolated active EMI filter without additional components on a power line according to the present invention includes a common mode (CM) choke 1710, a Y capacitor 1720, a transformer unit 1750, an amplifying unit 1740, and a compensation winding 1730.

[0193] Reference Fig. 9 A common mode (CM) choke 1710 is provided on the power supply side, and a live line (Liveline) and a neutral line (Neutral line) connected to an EMI source are respectively wound with windings.

[0194] The Y capacitor 1720 is arranged at one side of the EMI source generating the EMI, and is composed of two capacitors connected in series. The two capacitors are connected in parallel between the live line and the neutral line, and are connected to the ground in common.

[0195] The transformer unit 1750 is installed at the front end of the Y capacitor 1720. The primary coil senses the noise voltage of the Y capacitor, and the secondary coil transforms it and is isolated from the power line.

[0196] The amplifying unit 1740 amplifies the noise voltage sensed and transformed by the transforming unit 1750 .

[0197] The compensation winding 1730 is rewound on the common mode choke 1710 through a coil, and injects the noise signal amplified by the amplification unit into the common mode choke 1710 .

[0198] Fig.10 1 is a circuit diagram of a third embodiment of an isolated active EMI filter without additional components on a power line according to the present invention, including a common mode (CM) choke 1810, a Y capacitor 1820, a sensing winding 1830, an amplifying unit 1840 and a transformer unit 1850.

[0199] Reference Fig.10The common mode choke 1810 is arranged on one side of the EMI source generating the EMI, and the live line (Live line, L) and the neutral line (Neutral line, N) connected to the EMI source are respectively wound with windings.

[0200] The Y capacitor 1820 is arranged on the power supply side of the power supply, and is composed of two capacitors connected in series, wherein the two capacitors are connected in parallel between the live wire (L) and the neutral wire (N), and are connected to the ground together.

[0201] The sensing winding 1830 is rewound on the common mode choke 1810 through a coil, and senses the noise current of the common mode choke 1810 .

[0202] The amplification unit 1840 amplifies the noise current sensed by the sensing winding 1830 .

[0203] The transformer unit 1850 is arranged at the front end of the Y capacitor 1820, the primary coil receives the amplified signal from the amplification unit 1840, the secondary coil is connected to the ground connected to the Y capacitor 1820 and is isolated from the power line, and the signal transformed by the secondary coil is injected into the Y capacitor 1820 as a compensation signal.

[0204] Fig.11 It is a circuit diagram of the fourth embodiment of an isolated active EMI filter without additional components on a power line according to the present invention, comprising a common mode (CM) choke 1910, a Y capacitor 1920, a transformer unit 1930, an amplifying unit 1940 and a compensation winding 1950.

[0205] Reference Fig.11 The common mode choke 1910 is arranged at one side of the EMI source generating the EMI, and the live line and the neutral line connected to the EMI source are respectively wound with windings.

[0206] The Y capacitor 1920 is arranged on the power supply side of the power supply, and is composed of two capacitors connected in series. The two capacitors are connected in parallel between the live wire and the neutral wire, and are connected to the ground.

[0207] The transformer 1930 is disposed at the front end of the Y capacitor 1920 . The primary coil senses the noise voltage from the Y capacitor 1920 and transforms it through the secondary coil and is isolated from the power line.

[0208] The amplifying unit 1940 amplifies the noise voltage transformed by the transformer 1930 .

[0209] The compensation winding 1950 is rewound on the common mode choke 1910 through a coil, and injects the noise signal amplified by the amplification unit 1940 into the common mode choke 1910 as a compensation signal.

[0210] Fig.12 A flow chart showing a method for reducing EMI noise by adding active elements to a passive EMI filter, which corresponds to a first embodiment of an isolated active EMI filter without additional elements on a power line according to the present invention.

[0211] Reference Figure 1 and Fig.12 First, a passive EMI filter is prepared, in which a common mode choke 110 is arranged on the power supply side, and a Y capacitor 120 is arranged on the EMI source side (step S2010), that is, a common mode (CM) choke 110 is arranged on the power supply side of the power supply, and a live line (Live line) and a neutral line (Neutral line) connected to the EMI source are respectively wound with windings. The Y capacitor 120 is arranged on one side of the EMI source generating the EMI, and is composed of two capacitors connected in series, and the two capacitors are connected in parallel between the live line (L) and the neutral line (N), and are connected to the ground in common.

[0212] The EMI noise current of the common mode choke 110 is sensed by the sensing winding 130 formed by rewinding the common mode choke 110 with a coil (step S2020 ), and the amplifying unit 140 amplifies the noise current sensed by the sensing winding 130 (step S2030 ).

[0213] The signal amplified by the amplifying unit 140 is received by the primary coil of the transformer 150 installed at the front end of the Y capacitor 120 (step S2040), and then transformed by the secondary coil of the transformer 150 and injected into the Y capacitor 120 (step S2050). Here, the secondary coil of the transformer 150 is connected to the ground connected to the Y capacitor 120 and is isolated from the power line.

[0214] Fig.13 A flow chart showing a method for reducing EMI noise by adding active elements to a passive EMI filter, which corresponds to the second embodiment of an isolated active EMI filter without additional elements on a power line according to the present invention. Fig. 9 and Fig.13First, a passive EMI filter is prepared, in which a common mode choke 1710 is arranged on the power supply side and a Y capacitor 1720 is arranged on the EMI source side (step S2110). More specifically, the common mode choke 1710 is arranged on the power supply side of the power supply, and the live line and the neutral line connected to the EMI source are respectively wound with windings. The Y capacitor 1720 is arranged on the side of the EMI source generating the EMI, and is composed of two capacitors connected in series, and the two capacitors are connected in parallel between the live line (L) and the neutral line (N), and are connected to the ground.

[0215] The primary coil of the transformer 1750 installed at the front end of the Y capacitor 1720 uses the Y capacitor 1720 as a sensing capacitor to sense the noise voltage (step S2120). The sensed noise voltage is transformed using the secondary coil of the transformer 1750 (step S2130). Here, the secondary coil of the transformer 1750 is connected to the ground connected to the Y capacitor 1720, thereby being isolated from the power line.

[0216] The amplifying unit 1740 amplifies the voltage transformed from the secondary coil of the transformer 1750 (step S2140). The amplified noise signal is injected into the common mode choke coil through the compensation winding 1730 formed by rewinding the common mode choke coil with a coil (step S2150).

[0217] Fig.14 A flow chart showing a method of reducing EMI noise by adding active elements to a passive EMI filter, which corresponds to a third embodiment of an isolated active EMI filter without additional elements on a power line according to the present invention.

[0218] Reference Fig.10 and Fig.14 First, a passive EMI filter is prepared, in which a common mode choke 1810 is arranged on the EMI source side and a Y capacitor 1820 is arranged on the power supply side (step S2210). Specifically, a common mode (CM) choke 1810 is arranged on one side of the EMI source generating the EMI, and a live line (Live line, L) and a neutral line (Neutral line, N) connected to the EMI source are respectively wound with windings. The Y capacitor 1820 is arranged on the power supply side of the power supply, and is composed of two capacitors connected in series, and the two capacitors are connected in parallel between the live line (L) and the neutral line (N), and are connected to the ground.

[0219] The noise current of the common mode choke 1810 is sensed by the sensing winding 1830 formed by rewinding the common mode choke 1810 with a coil (step S2220), and the amplifying unit 1840 amplifies the noise current sensed from the sensing winding 1830 (step S2230).

[0220] The signal amplified by the amplifying unit 1840 is input to the primary coil of the transformer 1850 installed at the front end of the Y capacitor 1820 (step S2240), and then the signal input to the primary coil is transformed by the secondary coil of the transformer 1850 and injected into the Y capacitor 1820 as a compensation signal (step S2250). Here, the secondary coil of the transformer 1850 is connected to the ground connected to the Y capacitor 1820, and is isolated from the power line.

[0221] Fig.15 The flowchart of the method for reducing EMI noise by adding active elements to the passive EMI filter is shown, and the method corresponds to the fourth embodiment of the isolated active EMI filter without additional elements on the power line according to the present invention. Fig.11 and Fig.15 , prepare a passive EMI filter, in which a common mode choke 1910 is arranged on the EMI source side, and a Y capacitor 1920 is arranged on the power supply side (step S2310)

[0222] More specifically, the common mode choke 1910 is arranged on one side of the EMI source generating the EMI, and the live line (L) and the neutral line (N) connected to the EMI source are respectively wound with windings. The Y capacitor (1920) is arranged on the power supply side of the power supply, and is composed of two capacitors connected in series, and the two capacitors are connected in parallel between the live line (L) and the neutral line (N), and are connected to the ground in common.

[0223] The primary coil of the transformer 1930 installed at the front end of the Y capacitor 1920 senses the noise voltage of the Y capacitor 1920 (step S2320). The noise voltage sensed by the primary coil is transformed by the secondary coil of the transformer 1930 (step 2330), where the secondary coil of the transformer 1930 is connected to the ground connected to the Y capacitor 1920, thereby being isolated from the power line.

[0224] The amplifier unit 1940 amplifies the noise voltage transformed from the secondary coil (step 2340). The compensation winding 1950 injects the amplified noise signal as a compensation signal into the common mode choke 1910, and the compensation winding is formed by rewinding the common mode choke 1910 with a coil (step 2350).

[0225] In the above, the isolated active EMI filter without additional components on the power line according to the present invention is an active filter in the form of an additional installation on the existing passive EMI filter composed of a common mode choke and a Y capacitor. The present invention proposes a choke element, which is added to the power line by rewinding a noise sensing or compensation line on the common mode choke existing in the passive EMI filter. The Y capacitor existing in the passive EMI filter is used as a compensation or sensing capacitor, and a small transformer is installed in front of the compensation or sensing capacitor to isolate the active circuit from the power line, which has the advantages of not adding components to the power line and isolating the active circuit from the power line.

[0226] The isolated active EMI filter according to the present invention allows the sensing and compensation lines to be wound back with the optimal number of turns, so as not to reduce the noise attenuation performance of the passive EMI filter itself. The transformer turns ratio is adjusted to optimize the noise sensing and compensation performance through the Y capacitor and the small transformer at the front end, and to optimize the gain of the active filter amplification unit. Various stability compensation circuits can be added to ensure the feedback stability of the entire feedback circuit structure for noise sensing and compensation. The active EMI filter of the present invention is a feedback circuit structure for sensing noise and injecting a compensation signal.

[0227] According to the embodiment of the present invention, when only a passive filter is used, the low-frequency band conducted noise is reduced by 11 dB, but when the active EMI filter (AEF) of the present invention is additionally installed, it is reduced by 26 dB. In the case of using only a passive filter, an expensive common-mode choke must be used, or the total number of filter stages must be increased to sufficiently attenuate the noise in the low-frequency band.

[0228] The present invention has been described with reference to the embodiments shown in the accompanying drawings, but these are merely exemplary, and those skilled in the art will appreciate that various modifications and other equivalent embodiments may be made therefrom. Therefore, the true technical protection scope of the present invention should be determined by the technical ideas of the attached registered claims.

Claims

1. An isolated active EMI filter without additional components on the power line, characterized in that: include: A common mode choke, wherein two power lines connected to the EMI source are respectively wound into windings; A sensing winding, wherein the sensing winding is rewound on the common mode choke coil through a coil and senses a noise current of the common mode choke coil; an amplifying unit, the amplifying unit amplifying the noise current sensed by the sensing winding; a transformer, the transformer receiving the amplified signal from the amplifying unit and generating a compensation signal; and A Y capacitor connected between the transformer and the two power lines; in, When the capacitance of the parasitic circuit of the common mode choke is C cm , the capacitance of the parasitic circuit of the sensing winding is C sen When the number of turns of the sensing winding is N sen Less than 2C cm / Cs en The square root of .

2. The isolated active EMI filter without additional components on the power line according to claim 1, characterized in that: The amplification unit comprises: Non-inverting operational amplifier; a resistor R1, wherein the resistor R1 is connected to an input terminal of the non-inverting operational amplifier; and The resistor R2 is disposed between an input terminal and an output terminal of the non-inverting operational amplifier.

3. The isolated active EMI filter without additional components on the power line according to claim 2, characterized in that: A phase compensator is also included which is connected in parallel with the amplification unit.

4. The isolated active EMI filter without additional components on the power line according to claim 3, characterized in that: The phase compensator includes a resistor R connected between the input terminal and the output terminal of the non-inverting operational amplifier. c and capacitor C c .

5. The isolated active EMI filter without additional components on the power line according to claim 1, characterized in that: It also includes: a low-pass filter, one end of which is connected to the sensing winding, and the other end of which is connected to the amplification unit.

6. The isolated active EMI filter without additional components on the power line according to claim 5, characterized in that: The low-pass filter includes a resistor R connected between the sensing winding and the amplifying unit. f and a capacitor C connected between the amplifier unit and ground f .

7. The isolated active EMI filter without additional components on the power line according to claim 5, characterized in that: The impedance Z from the input terminal of the amplifier unit toward the low-pass filter in,AEF is set to be higher than the parasitic RC component impedance Z of the sensing winding sen,para .

8. The isolated active EMI filter without additional components on the power line according to claim 1, characterized in that: It also includes: a bypass branch, wherein the bypass branch is configured between the Y capacitor and the transformer.

9. The isolated active EMI filter without additional components on the power line according to claim 8, characterized in that: The bypass branch includes: a resistor R connected between the Y capacitor and the transformer d1 , connected in series with the resistor R d1 Capacitor C between the d and resistor R d2 .

10. The isolated active EMI filter without additional components on the power line according to claim 8, characterized in that: It also includes: a low-pass filter, one end of which is connected to the sensing winding, and the other end of which is connected to the amplification unit; The cut-off frequency of the low-pass filter is 1 / 2πR f C f is greater than the maximum operating frequency f determined by the inductance of the transformer and the capacitance of the bypass branch op,max , and is less than the frequency Among them, k sen is the coupling coefficient of the sensing winding, N sen is the number of turns of the sensing winding, L cm is the inductance of the common mode choke, C sen is the capacitance of the parasitic circuit of the sensing winding.

11. The isolated active EMI filter without additional components on the power line according to claim 1, characterized in that: The transformer includes: a primary winding connected to the output terminal of the amplifying unit, and a secondary winding connected to the other end of the Y capacitor.

12. The isolated active EMI filter without additional components on the power line according to claim 10, characterized in that: The effective capacitance of the Y capacitor is at least determined by the number of turns N of the sensing winding. sen , the primary winding of the transformer and the winding ratio N of the primary winding inj , and the gain of the amplification unit is determined.

Images