JP2009260088A - Pulse transformer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a pulse transformer which has superior conversion efficiency, signal quality (signal integrity), and electromagnetic compatibility (EMC) although it is very easy to design. <P>SOLUTION: A loss pulse transformer includes a spool 15, a magnetic core 16, a winding 17, terminals 19 to 26 and 32 to 37, and a printed wiring board 38. The terminals 19 and 32 are connected to a ground plane 90 of the printed wiring board 38. A terminal 20 is connected to one end of a strip conductor 89. The terminal 33 is connected to the other end of the strip conductor 89 on the printed wiring board 38. The the printed wiring board 38 is constituted by sticking two one-side copper clad plates with a semiconductor prepreg. The strip conductor 89, the ground plane 90, insulator layers 86 and a microstrip line formed with an a semiconductor layer 87 constitute a loss line. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

The present invention relates to a pulse transformer, and in particular, is used for high-speed digital data communication equipment and high-frequency DC-DC converters that use high-speed switching elements, and has high conversion efficiency, can be reduced in size and weight, and can improve signal quality (signal integrity) and The present invention relates to a lossy pulse transformer excellent in electromagnetic environment compatibility (EMC).

In recent years, transistors have been increased in speed in order to achieve higher performance and higher functionality of information technology devices and multimedia devices. Information technology devices and multimedia devices are also strongly demanded to save energy and reduce size and weight.

However, driving a pulse transformer at a high repetition rate using high-speed transistors has the problem of increasing electromagnetic noise and heat generation. Conventional pulse transformer design technology meets the demands for energy saving and miniaturization and weight reduction. It was difficult.

It is physics that dominates the theory of circuit design technology, and more directly, electromagnetism. According to electromagnetism, circuit states can be considered as active states (exited states), steady states (stationary states), and quasi-steady states (quasi-stationary states that can be regarded as steady states in practice)
There are stationary states. The active state is a state in which the electric field and magnetic field on the circuit are changing or oscillating, and an AC circuit is one example. The oscillating electric and magnetic fields travel as electromagnetic waves in the insulator. When the insulator is a vacuum space, the electromagnetic wave travels at the speed of light.

The steady state is a state where the electric field and magnetic field on the circuit are stationary, and a DC circuit is an example. The quasi-steady state means that an electric field and a magnetic field travel on the circuit as an electromagnetic wave, but the wavelength of the electromagnetic wave is very long compared to the circuit length, and even if the behavior of the electromagnetic wave in the circuit is regarded as only strong and weak vibration, it is practical. This is a state where no inconvenience occurs. When a low-frequency analog circuit or a switching element having a rise time of about 1 [ns] or more is used in a circuit having a wiring of 10 [cm] or more, this is an example that can be regarded as a quasi-stationary state.

According to electromagnetics, the current in a circuit in an active state is defined as Ampere's law and is given by

According to electromagnetism, the electric potential V is defined as an integral value of an electric field from an infinite point where the electric field does not reach to one point of the conductor, but practically as an integral value of the electric field from the ground plane to one point of the conductor, E is obtained from the following equation as the slope of the potential V.

Maxwell published Maxwell's equation, which merged the theory of magnetic fields and the theory of electric fields, in 1873, and then transformed this equation into the form of D'Alembert's wave equation to derive the vector wave equation. Maxwell, who had been insisting since 1862, theoretically proved that both electromagnetic waves and light propagate at the speed of light using this equation, and completed linear electromagnetic wave theory (hereinafter referred to as electromagnetic wave theory). Was completed. In 1887, Hertz demonstrated the existence of electromagnetic waves by experiment and proved the correctness of Maxwell's electromagnetic wave theory.

According to electromagnetism, an electric field and a magnetic field that change with time interact with each other and propagate in a space or a dielectric as a transverse wave. The speed of the electromagnetic wave propagating in the vacuum is the speed of light. The propagating electromagnetic wave propagates power according to the pointing vector theory. An electromagnetic wave propagating in space is composed of an electric field wave and a magnetic field wave whose period and polarity coincide and whose amplitude vector is orthogonal to the traveling direction. The electromagnetic wave in this state is called a TEM (transverse electromagnetic) wave. The value obtained by dividing the amplitude of the electric field wave constituting the TEM wave by the amplitude of the magnetic field wave is the wave impedance (surge
impedance or wave impedance).

According to electromagnetism, electromagnetic waves travel not only in space but also in media. The speed of the electromagnetic wave traveling through the lossless dielectric is slowed by the square root of the relative permittivity with respect to the speed of light, and the wavelength is shortened by the square root of the relative permittivity. The latter is called wavelength compression.

According to electromagnetism, an electromagnetic wave traveling in a lossy medium follows an attenuation constant γ expressed by the following equation, and the amplitude decreases and the phase changes with progress. α which is a real term of γ is called an attenuation constant, and β which is an imaginary term of γ is called a phase constant. α is expressed in units of nep / m (neper / meter). 1 [nep / m] means that the amplitude decreases by exp- ¹ or 0.368 times after proceeding 1 meter.

According to electromagnetics, the term in parentheses in the following equation obtained by transforming γ ² in equation (3) is defined as the complex permittivity for a lossy dielectric, and the imaginary part (σ / ε ₀ ω) is defined as The value divided by the real part (ε _r ) is called the dielectric loss tangent and is represented by tan δ. However, tan δ has no deep meaning in electromagnetics.

When the electromagnetic wave travels in the conductor, there is no electric charge acting on the electromagnetic wave in the conductor, and the conductivity σ is much larger than ωε, so γ is expressed by the following equation. Δ, which is the reciprocal of the attenuation constant α in the following equation, is called the skin thickness.

According to electromagnetics, the intrinsic impedance Z _0, which is the ratio of the electric field to the magnetic field of the electromagnetic wave traveling in the conductor, is assumed to be very large compared to ωε in the intrinsic impedance in a lossy medium. Given.

In an active state or quasi-stationary state where the electric and magnetic fields on the circuit are changing or oscillating, the electromagnetic wave theory dominates the circuit, and in this case, it is difficult for the electromagnetic wave to travel through the conductor. However, in a steady state where the electric and magnetic fields on the circuit are stationary, the current can easily move through the conductor.

According to physics, there are almost inexhaustible free electrons or charges in the conductor. However, the total amount of charge in the conductor is determined by physical properties, and the value is constant. When a static load is connected to the direct current power source, a current flows due to the movement of charges in the conductor, but in general, only a small electric field can be applied to the movement axis of charges, so the average movement speed of charges is extremely slow.

For example, when a current of 10 amperes defined by a charge velocity (dq / dt) in a conductor is traveling in a copper wire having a cross section of 1 square millimeter, the current progression rate is calculated according to physics It becomes 0.368 [mm / s] at room temperature. Since the charge in the conductor can move although it is slow, if the same amount of charge is constantly supplied from one end of the conductor when the charge is constantly consumed at the other end of the conductor, Energy supply to a steady load such as a resistor connected to the other end is performed without any trouble.

Telecommunications engineering handles the progression of electrical signals on transmission lines. According to telecommunications engineering, when an electric signal is applied between two DC-insulated conductors, the electric signal becomes a current wave and a voltage wave and travels through the transmission line.

In telecommunications engineering, as in AC circuit theory, the current is the average charge velocity (dq / dt) in the conductor, that is, the conductor current. However, in Maxwell's equations that form the basis of electromagnetism, the conductor current corresponds to a current density J that is not a function of time.

The AC circuit theory and telecommunications engineering define the current as dq / dt for the following reasons. Kirchhoff's law, one of the important laws supporting AC circuit theory, was announced in 1845, and Maxwell theoretically proved the existence of the electromagnetic wave, and the existence of the electromagnetic wave was confirmed by Hertz in an experiment 42 years Previously, the telegraph equation, one of the important theories supporting telecommunications engineering, was developed in 1874, 13 years before the existence of electromagnetic waves was also confirmed. Therefore, at the time when AC circuit theory and telecommunications engineering were put into practical use, there was no idea that the action of the circuit was the action of electromagnetic waves. Furthermore, the theory was not revised after that.

In the telegraph equation that forms the basis of telecommunications engineering, the basis of the fact that the conductor current can flow at the speed of light is the D'Alembert wave equation. In D'Alembert's wave equation, the subject of the wave is expressed by a vector function with a Laplacian of scalar quantity and is not specified. Therefore, it is necessary to compare the circuit equation relating to voltage and current with the Duramber wave equation after matching the conductor current into a wave with the voltage between conductors and the electromagnetics governing the electrical circuit. However, as described above, in electromagnetism, it is assumed that the conductor current does not change with time.

If the definition of current is contrary to electromagnetism, the concept of contradiction with electromagnetism also arises in relation to the voltage, impedance, electromagnetic wave, and transmission loss of the transmission line. Since telecommunications engineering has a long history and is still in practical use for transmission line design, conventional transmission line design for continuous waves can be devised to avoid the emergence of contradictions with electromagnetism.

Transmission line design for intermittent waves such as switching waves or digital waves also seems to be efficient based on telecommunications engineering. However, since the practical application of telecommunications engineering to digital circuits is not so strong, the contradiction with electromagnetics becomes obvious unless careful design and analysis are performed in contrast to electromagnetics.

According to electromagnetics, when the insulator sandwiched between the two conductors constituting the transmission line is in a vacuum, the electromagnetic wave of the TEM wave travels in the vacuum at the speed of light. In other words, the current and voltage in this case travel through the insulator, not the conductor of the transmission line, and have values obtained from the equations (1) and (2), respectively. Since an actual current or voltage is a magnetic field or an electric field, it can travel in a quasi-light speed as a wave in the insulator. A value obtained by dividing the amplitude of the electric field wave constituting the TEM wave on the transmission line by the amplitude of the magnetic field wave is the characteristic impedance.

According to telecommunications engineering, the behavior of a signal traveling on a transmission line is determined by the characteristic impedance and propagation constant of the transmission line. The characteristic impedance Z ₀ of the parallel plate line the ideal flat conductor is parallel to opposite sides of the ideal insulator is determined from the following equation by the physical constants of the transmission line. The material characteristics of the flat conductor and the insulator do not have a large practical effect on the characteristic impedance of the transmission line.

According to the telecommunications engineering, the characteristic impedance of the Rehel line having a structure in which the centers of two conductors having a diameter a are arranged in parallel with a distance d can be obtained from the following equation.

According to telecommunications engineering, the characteristic impedance of a microstrip line can be obtained from the following equation.

According to telecommunications engineering, when electromagnetic waves are injected into a transmission line having an unknown characteristic impedance Z ₁ through a transmission line having a known characteristic impedance Z ₀ ,
The reflection coefficient S ₁₁ at the connection point of the two transmission lines is expressed by the following equation.

According to telecommunications engineering, a transmission line having a loss with a reflection coefficient S ₁₁ with respect to a known characteristic impedance Z _0, that is, the transmission coefficient S ₂₁ of the transmission line is expressed by the following equation.

According to electromagnetics, the practical transmission line attenuation constant is determined by the attenuation when the electromagnetic wave travels through a lossy dielectric body, and part of the electromagnetic wave penetrates into the conductor in the course of the electromagnetic wave traveling through the dielectric. It can be considered that this is the sum of the conductor loss and the radiation loss leaking out of the transmission line.

Wiring design of high-speed digital data communication equipment is performed according to telecommunication engineering. However, telecommunications engineering is suitable for transmission line design that handles continuous waves such as sine waves, but as mentioned above, transmission line design that handles intermittent waves such as digital signals is not suitable because of inconsistencies with electromagnetics. . For this reason,
Even in accordance with telecommunications engineering, it is difficult to design a pulse transformer that has high conversion efficiency, can be reduced in size and weight, and has excellent signal quality (electromagnetic integrity) and electromagnetic compatibility (EMC).

A direct-current power source in high-speed digital data communication equipment and a high-frequency DC-DC converter is considered to supply electric charge to a circuit.

According to electromagnetism, Maxwell modified Coulomb's law, assuming that the force acting on the unit (test) point charge is due to the electric field where the unit point charge exists. This fact is not well known.

According to the modified electromagnetism, the electrostatic energy w _E related to the electric field is expressed as:

Thus, the electrostatic energy w _E is not held by electric charges but is accumulated in the medium as the product of the electric field E and the electric flux density D or the electric field E.

The electrostatic energy w _C stored in the capacitor having the capacitance C to which the voltage V is applied is expressed by the following equation, where d is the electrode distance and S is the electrode area from the equation (13).

According to electromagnetism, the magnetostatic energy w _H relating to the magnetic field is assumed to be accumulated in the medium as the product of the magnetic field and the magnetic flux density, and is expressed by the following equation.

The magnetostatic energy w _L accumulated in the reactor of the induction L to which the current I is applied is expressed by the following equation, where l is the magnetic path length of the reactor and S is the cross-sectional area of the magnetic path.

According to the isolated electromagnetic wave concept shown in Non-Patent Document 1 and Non-Patent Document 2, the switching element excites an isolated electromagnetic wave, which is a kind of nonlinear wave or soliton, at the moment of switching. The power conversion circuit and the switching element in the control circuit constituting the switching power supply device also excite a solitary electromagnetic wave, which is a kind of nonlinear wave or soliton, at the moment of switching by the same mechanism.

The excitation mechanism of isolated electromagnetic waves during the switching operation of the switching element is caused by suddenly opening a gate for storing water in various experiments conducted by John Scott Russell in 1834 when he discovered solitons. It is very similar to the generation mechanism of solitons and the tsunami generation process that has been confirmed to be a kind of solitons.

According to the isolated electromagnetic wave concept shown in Non-Patent Document 1 and Non-Patent Document 2, at the moment when the switching element switches from OFF to ON, the potential at the point where the switching element connects the power line and the signal line is the voltage of the DC power supply. Is a value obtained by dividing the characteristic impedance of the power line and the signal line. Therefore, isolated electromagnetic waves with a polarity that lowers the voltage to the divided voltage are excited on the power line, and isolated electromagnetic waves with a polarity that raises the voltage to the divided voltage are excited simultaneously on the signal line, and the isolated amplitude vectors are orthogonal to each other according to the electromagnetic wave theory. It travels on the transmission line with electric field waves and solitary magnetic field waves.

FIG. 1 is an example of an equivalent circuit related to a push-pull circuit 1 for explaining the behavior of an isolated electromagnetic wave. In Figure 1, a push-pull circuit 1 is connected to the middle of the transmission line of the characteristic impedance Z _0, the transmission line 5 of the characteristic impedance Z ₀ is connected between the DC power source 4 and the push-pull circuit 1 supply line The transmission line 6 having the characteristic impedance Z ₀ is connected between the push-pull circuit 1 and the matching termination resistor 7 to form a signal line. Push-pull circuit 1 is a P-channel MOS
Complementary configuration with FET 2 and N-channel MOS FET 3.

In FIG. 1, the on state of the push-pull circuit 1 is a state where the P-channel MOS FET 2 is on and the N-channel MOS FET 3 is off, and the push-pull circuit 1 is off. The relationship between the magnetic field and the current and the relationship between the electric field and the potential with respect to the TEM wave traveling through the transmission line are shown as the amperage law and the definition of the potential in electromagnetics, respectively.

FIG. 2 shows a potential waveform 9 on the transmission line 6 when the push-pull circuit 1 is turned on, and an electric field waveform 8 that travels on the transmission line 6 obtained by back calculation from the definition of the potential shown in electromagnetics. FIG. 3 shows a potential waveform 11 on the transmission line 5 when the push-pull circuit 1 is on, and an electric field waveform 10 that travels on the transmission line 5 on the power source side obtained by back calculation from the definition of the potential shown in electromagnetics. Show.

As shown in FIGS. 2 and 3, the waveform of the electric field generated by the switching of the push-pull circuit 1 is the product of the rise time and the circumference obtained by regarding the tangent of the maximum slope of the rising waveform of the switching element as the rising waveform. It approximates to a half waveform of a sine wave having an effective frequency defined by the frequency obtained as the reciprocal of. To quote the concept of effective frequency, the accuracy of the approximation is expected to be 92% or more. Therefore, practically, it can be performed at an effective frequency as far as design is concerned.

1 to 3, when the push-pull circuit 1 is turned on, the potentials at points B and C in FIG. 1 are equal to E / 2 [V]. The isolated electric field wave 8 traveling on the transmission line 6 having the opposite polarity and the isolated electric field wave 10 traveling on the transmission line 5 excited by the push-pull circuit 1 travel in opposite directions with respect to the push-pull circuit 1. The isolated electric field wave 8 traveling on the transmission line 6 travels while increasing the potential of the transmission line 6 from 0 [V] to E / 2 [V], and travels toward the matching termination resistor 7. On the other hand, the isolated electric field wave 10 traveling on the transmission line 5 is an insulator that constitutes the transmission line toward the DC power source 4 while lowering the potential of the transmission line 5 from E [V] to E / 2 [V]. Proceeds with quasi-light speed.

According to Non-Patent Document 1 and Non-Patent Document 2, the wavelength of an isolated electromagnetic wave traveling on the transmission line is defined by the following equation.

Conventional printed wiring boards are described in the following patent documents and non-patent documents. The point will be described later.
JP-A-6-224052 JP 7-161533 A JP 7-45451 JP-A-2004-253500 (P2004-253500A) JP2005-19766 (P2005-19766A) Japanese Patent Application No. 2006-332245 (P2006-332245A) JP2007-67177A (P2007-67177A) Hirokazu Tohya and Noritaka Toya "A Novel Design Methodology of the On-Chip Power Distribution Network Enhancing the Performance and Suppressing EMI of the SoC", IEEE International Symposium on Circuits and Systems 2007, pp. 889-892, May 2007. Hirokazu Toya, Norio Naoya Toya "New design method of on-chip power distribution circuit that greatly improves SoC performance and EMC", IEICE Technical Report, Vol.107, No.149, EE2007-20, pp .73-78, July 2007.

The first problem to be solved relates to Patent Document 1. In Patent Document 1, in order to improve the degree of coupling between windings in a transformer winding, the coil shape is made up of a large number of individual wires (for example, copper lacquer wires) that have a small cross section and are insulated from each other. With flexible litz wire wound, individual wires are grouped individually or in parallel and closely twisted together, and primary and secondary windings can be selected arbitrarily Disclosed is a technique formed by a bundle of a number of individual wires.

According to electromagnetism, the transformer stores the magnetostatic energy by the steady energy applied to the primary winding and releases the magnetostatic energy from the secondary winding. Transformers with high magnetostatic energy storage and release efficiency are generally considered to have high performance. On the other hand, when the electromagnetic wave applied to the primary winding of the transformer leaks to the secondary winding, the electromagnetic environment compatibility (EMC) and the signal quality (signal integrity) of the device on which the transformer is mounted are reduced.

In Patent Document 1, since a plurality of primary windings and secondary windings are gathered in parallel and closely twisted with each other, a coupling of magnetostatic energy is provided between the primary winding and the secondary winding. It is considered that the degree of coupling of electromagnetic waves due to not only the degree but also the displacement current or the electric flux current due to the capacitance between the windings is very high. Therefore, it has been impossible to significantly improve both the signal quality (signal integrity) and electromagnetic compatibility (EMC) of switching equipment using a transformer by the disclosed technology.

A second problem to be solved relates to Patent Document 2. Patent Document 2 discloses a pulse transformer used for data communication using pulse code communication and modulation / demodulation, ensuring insulation between the transmission path side and the transmitter / receiver side, removing common mode noise, and primary winding and secondary winding. In order to increase the electromagnetic coupling between the windings, a shield plate is provided between the primary winding and the secondary winding, or the primary winding, the winding having a shielding function, and the secondary winding Is not integrally formed with an adhesive by interposing a rod-shaped shield winding 6 having substantially the same wire diameter between the primary conductor winding 2 and the secondary conductor winding 2 instead of the conventional method of winding the wire alternately. Disclosed is a technology in which this is a composite member, and the composite member is spirally formed at a constant pitch on the toroidal core, and the interval between the adjacent primary conductor winding 1 and secondary conductor winding 2 is evenly wound. Yes.

Since the composite member which is a wiring material used in Patent Document 2 has a ribbon shape, the size of the toroidal core increases. In order to reduce the size of the toroidal core, it is necessary to wind the composite member over multiple layers. However, when the multi-layer winding is used, the effect of the shield winding in the composite member is impaired. Therefore, it has been impossible to significantly reduce the signal quality (signal integrity) and electromagnetic environment compatibility (EMC) of digital devices and to reduce the size and weight with the disclosed technology.

A third problem to be solved relates to Patent Document 3. Patent Document 3 discloses a power supply primary side, a magnetic core, and a secondary side winding in order to remove noise of a switching power supply circuit by a switching transformer having an electrostatic shield function that is easy to assemble automatically and inexpensive. Compared to the conventional shield using copper foils connected to the primary and secondary stable potential terminals between the wire and the magnetic core, respectively, the main winding of the primary winding constituting the live part An electrostatic shield layer formed by independent one-layer windings connected to the active part and the non-active part between the constituent part and the main constituent part of the secondary winding constituting the non-active part The technique of the switching transformer provided with this is disclosed.

In the above conventional technique, since the voltage excited on the copper foil is small, the copper foil functions as an electrostatic shield. However, the electrostatic shield winding disclosed in Patent Document 3 includes a primary winding or a secondary winding. A large voltage equivalent to the line voltage is excited. Since the primary winding and the primary-side electrostatic shield winding, and the secondary winding and the secondary-side electrostatic shield winding are densely wound, the technique disclosed in Patent Document 3 is eventually used. It is impossible to obtain a shield characteristic equivalent to that of an electrostatic shield performed with a conventional copper foil. Therefore, it has been impossible to remove the noise of the switching power supply circuit by the disclosed technique.

A fourth problem to be solved relates to Patent Document 4. Patent Document 4 discloses that a coil, a bobbin, and a magnetic core are provided in place of a conventional method using a short ring for shielding in order to suppress noise generation of a switching transformer from an inverter power supply, and the bobbin is wound around which the coil is wound. And a side plate portion provided so as to sandwich the winding portion, and a transformer body provided with a concave portion for arranging a magnetic core so as to go around the winding portion on the surface of each side plate portion A low-profile switching transformer comprising: a storage case that covers the transformer body; a shield case that is provided on the inner surface of the storage case and is made of a conductor that covers the transformer body; and a resin that is injected into the storage case provided with the transformer body. The technology which comprises is disclosed.

In Patent Document 4, the noise problem of the switching transformer is regarded as only the radiated electromagnetic wave from the switching transformer, but the largest noise problem of the switching transformer is the high frequency electromagnetic coupling between the primary side and the secondary side. Therefore, it has been impossible to solve the noise problem of the inverter power supply by the disclosed technique.

The fifth problem to be solved relates to Patent Document 5. Patent Document 5 increases the transmission efficiency in a low frequency range of a pulse transformer used for an inverter or the like, and secures an attenuation amount of a coupled state in a high frequency range of a predetermined level or higher, thereby preventing unnecessary signals and noise (noise). ), A closed magnetic circuit core that mechanically forms a closed magnetic circuit without a gap, and a primary winding and a secondary winding wound around the closed magnetic circuit core around the closed magnetic circuit core as winding axes, respectively. The coupling control core unit is arranged between the wire and the primary winding and the secondary winding, and controls the capacitive coupling so as to reduce the capacitive coupling between the primary winding and the secondary winding in a high frequency band above a predetermined level. The technique provided with these is disclosed.

According to electromagnetism, the transformer stores the magnetostatic energy by the steady energy applied to the primary winding and releases the magnetostatic energy from the secondary winding. Transformers with high magnetostatic energy storage and release efficiency are generally considered to have high performance. On the other hand, when the electromagnetic wave applied to the primary winding of the transformer leaks to the secondary winding, the electromagnetic environment compatibility (EMC) and the signal quality (signal integrity) of the device on which the transformer is mounted are reduced. In the technique disclosed in Patent Document 5, since the primary winding and the secondary winding are arranged separately at the two leg portions of the E-type core, the primary winding and the secondary winding The high frequency electromagnetic coupling between is considered low. However, the magnetic coupling to magnetostatic energy is low. Therefore, it has been impossible to achieve both the resolution of the noise problem of the inverter power supply and the reduction in size and weight and the improvement of power conversion efficiency by the disclosed technology.

The sixth problem to be solved relates to Patent Document 6 and Patent Document 7. Patent Document 6 and Patent Document 7 are intended to realize a stable DC superposition characteristic with a simple structure, reduce leakage magnetic flux, and provide a low-cost wire ring component, that is, a choke coil. A technology that enables sufficient DC superimposition characteristics by molding with a mixture of soft magnetic metal powder consisting of paste-like fluid and thermosetting resin in order to create a closed magnetic circuit around the road and ring parts. Disclosure.

Patent Document 6 and Patent Document 7 are limited to choke coils, and are limited to soft magnetic powder as a mixture for molding. Therefore, high-speed digital data communication equipment and high-frequency DC-DC converters that use high-speed switching elements. It has been impossible to apply to pulse transformers that have high conversion efficiency, can be reduced in size and weight, and are excellent in signal quality (signal integrity) and electromagnetic environment compatibility (EMC).

In analog circuits, changes in the circuit state are relatively gradual and the beginning and end are often unclear. Analog circuits have a long history, and especially in engineering, there are almost no problems in practical use in circuit design according to conventional AC circuit theory or telecommunications engineering without having to return to electromagnetism by applying rules of thumb, etc. It was.

On the other hand, unlike the case of the analog circuit, the beginning and end of the state change in the switching circuit are clear. Changes in the state of the switching circuit are very rapid, and sudden changes in electric or magnetic fields naturally excite large levels of electromagnetic waves. The change in the electric or magnetic field in the switching circuit is intermittent. Furthermore, in a data processing circuit that occupies about 90% of a semiconductor integrated circuit, the switching cycle is generally indefinite.

As described above, the analog circuit and the switching circuit are greatly different from the viewpoint of electromagnetics. However, in conventional telecommunications engineering and AC circuit theory, the design of a circuit that assumes intermittent circuit operation, that is, a pulse circuit, is performed by the above-mentioned method that has nothing to do with electromagnetics, and the analysis is performed using a switching wave. The Fourier transform method has been applied, which is considered as a kind of distorted wave.

According to the Fourier transform method, the distorted wave is composed of a number of harmonics that are sine waves. These harmonics are numerous sine waves with no beginning and no end. If the signal on the circuit is analyzed for each harmonic and the results are added, the switching circuit can be analyzed. However, the Fourier transform method is a mathematical method, and it is not used in the design and analysis of electrical and electronic circuits after confirming the consistency with the higher theory of electromagnetism. The analysis of instantaneous phenomena is impossible because of the large deviation from reality.

For example, when a switching wave having a duty of 1/10 and a repetition frequency of 1 [GHz] is Fourier transformed, it can be decomposed into a DC component having a value of 1/10 of the amplitude and a harmonic having 1 [GHz] as a fundamental wave. If a certain length of wiring or transmission line in a semiconductor integrated circuit using a CMOS circuit that hardly passes direct current has a loss that reduces the amplitude of 1 [GHz] to 1/2, The amplitude of the switching wave at the end of the transmission line is reduced to almost ½ or less in the analysis result.

However, according to electromagnetics, the amplitude of the switching wave is maintained by electrostatic energy supplied from a DC power source. Since electrostatic energy is not a wave, it is not affected by the loss of wiring or transmission lines. Therefore, the amplitude of the switching wave observed at the end of the transmission line should not be attenuated.

This fact indicates that it is an error to treat the switching wave as a distorted wave. This fact also shows that the conventional theory concerning signal quality (signal integrity) in digital signal wiring, which follows the concept of group velocity generated based on the Fourier transform method, needs to be modified. In other words, this fact means that a unified theory for design and analysis that can explain instantaneous changes in switching circuits and digital circuits and behavior over a relatively long period of time without contradiction must be newly established. Suggests.

An object of the present invention is to provide means for fundamentally solving the above problems.

In order to solve the above-mentioned problem, an invention according to claim 1 relates to a pulse transformer, and includes a pulse transformer having at least one primary winding and at least one secondary winding driven by a switching element. The pulse transformer is defined in the primary winding of the pulse transformer with a frequency obtained as the reciprocal of the product of the rise time and the circumference calculated by regarding the tangent of the maximum slope of the rising waveform of the switching element as the rising waveform. It is characterized by being designed as a lossy pulse transformer having an ability to attenuate the amplitude of the electromagnetic wave to 1/10 or less inside the pulse transformer when an electromagnetic wave having an effective frequency is applied.

The invention described in claim 2 relates to a pulse transformer. In the loss pulse transformer according to claim 1, the wiring connecting the winding structure constituting the loss pulse transformer and the external connection terminal includes a conductor, Loss line composed of an insulator formed on a conductor surface and a semiconductor formed on the surface of the insulator, or a conductor, the semiconductor formed on the conductor surface, and formed on the surface of the semiconductor It is characterized by a lossy line composed of an insulator.

The invention described in claim 3 relates to a pulse transformer, wherein in the loss pulse transformer according to claim 1 or 2, the length of the loss line constituting the loss pulse transformer is 10 [mm] or more. It is characterized by being.

According to a fourth aspect of the present invention relates to a pulse transformer, and in the lossy pulse transformer according to any one of the first to third aspects, the characteristic impedance of the lossy line constituting the lossy pulse transformer is mounted with the lossy pulse transformer. It is characterized by being equal to the characteristic impedance of the printed wiring board to be performed with an accuracy within ± 20%.

The invention according to claim 5 relates to a pulse transformer, and in the loss pulse transformer according to claims 1 to 4, the loss line constituting the loss pulse transformer is covered with two semiconductor films. It is characterized by being formed by embedding a conductor in a flexible insulating resin.

The invention described in claim 6 relates to a pulse transformer, wherein in the loss pulse transformer according to claims 1 to 5, the loss line constituting the loss pulse transformer is insulated between two conductor layers. It is formed on a printed wiring board having a layer and a semiconductor layer.

The invention according to claim 7 relates to a pulse transformer, wherein the semiconductor is an inorganic semiconductor or an organic semiconductor and has a conductivity of 10 [S / m] or more. It is designed to have a rate.

The invention described in claim 8 relates to a pulse transformer, wherein in the loss pulse transformer according to claims 1 to 7, the semiconductor is poly (p-phenylene) (22.7 °) or poly (p -Phenylene sulfide), or carbon graphite, or manganese dioxide, or polyacetylene, or polythiophene, or polypyrrole, or polyphenylene vinylene, or tetrathiafulvalene-tetraquinodimethane (TTF-TCNQ).

The invention according to claim 9 relates to a pulse transformer, wherein in the loss pulse transformer according to claims 1 to 8, the semiconductor is n-type silicon doped with phosphorus of 10 ¹⁵ [cm ⁻³ ] or more, P-type silicon doped with boron of 10 ¹⁶ [cm ⁻³ ] or more, or amorphous silicon, alumina, zirconia, carbide, nitride, silicide, silicon carbide, or silicon nitride mixed with impurities Or magnesium nitride or zinc oxide.

The invention described in claim 10 relates to a pulse transformer. In the loss pulse transformer according to claims 1 to 9, the semiconductor is the semiconductor according to claims 4 to 6, or the maximum length is 100 nanometers. It is a resin in which the following semiconductor particles are mixed.

When the present invention according to the isolated electromagnetic wave concept shown in Non-Patent Document 1 and Non-Patent Document 2 is applied to a pulse transformer mainly used in high-speed digital data communication equipment and high-frequency DC-DC converters, Design and analysis can be performed easily.

In addition, when the present invention based on the isolated electromagnetic wave concept is applied to a pulse transformer mainly used in a high-speed digital data communication device or a high-frequency DC-DC converter, a snubber or a matching termination circuit that is indispensable for a pulse transformer that applies a high-speed switching pulse. Is completely unnecessary.

In addition, when the present invention based on the isolated electromagnetic wave concept is applied to a pulse transformer, the conversion efficiency, signal integrity, and electromagnetic environment compatibility (EMC) of high-speed digital data communication equipment and high-frequency DC-DC converters are improved. In addition, it is possible to reduce the size and weight and improve the performance.

DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, the best embodiment according to the invention will be described in detail with reference to the drawings.

(Embodiment 1)
FIG. 4 is an example of a lossy line.

In FIG. 4, the loss winding is composed of a conductor 12, an insulator 13, and a semiconductor 14.

FIG. 5 is an example of a cross-sectional structure of the loss pulse transformer.

In FIG. 5, the loss pulse transformer includes a winding frame 15, a magnetic core 16, a winding 17, a resin 18, primary winding external terminals 19 and 20, primary winding internal terminals 21 and 22, fixing terminals 23 and 24, two It consists of secondary winding internal terminals 25 and 26, secondary winding external terminals 27 and 28, a resin substrate 29, and a resin box 30.

FIG. 6 is an example of the internal wiring structure of the loss pulse transformer.

In FIG. 6, the internal structure of the loss pulse transformer is such that the loss line 31 having the cross section of FIG. 4 for connecting the primary winding external terminals 19 and 20 and the primary winding internal terminals 21 and 22 is made of resin. Built in the box 30. The lossy line 31 has the cross-sectional structure of FIG. 4, the radius of the conductor 12 is 0.5 [mm], the thickness of the semiconductor 14 is 0.1 [mm], and the center distance between the two conductors 12 is 2
If it is [mm], the characteristic impedance of the lossy line 31 is 50.8 [Ω] from the equation (8).

When the conductivity of the insulator constituting the loss line having the characteristic impedance of Z ₀ is infinite and the conductivity of the semiconductor is σ _P , a part of the electromagnetic wave having the impedance Z ₀ traveling through the insulator is a specific impedance. penetrating into the semiconductor having a Z _P. The electromagnetic waves in progress in the semiconductor are electromagnetic waves that are not useful for communication other than TEM waves, and all of them are lost. If we define the effective conductivity of the semiconductor as the actual conductivity of the semiconductor, corrected by the ratio related to the loss, the effective conductivity σ
_P1 can be obtained from the following equation.

The attenuation constant α _P1 when the effective conductivity is σ _P1 can be obtained from the following equation.

In FIG. 5, the lossy line 31 having the cross-sectional structure of FIG. 4 includes an insulator with infinite conductivity and a semiconductor with a conductivity of 100 [S / m], a length of 1 [cm], when the impedance is connected to the transmission line of the printed wiring board of 50.8 [Omega] through the terminals 19 and 20, the transmission coefficient _{S 21} of the loss line 31, formula (10), (11) , (17) and (18), −43 dB at 1 [GHz] and −162 dB at 10 [GHz]. The validity of these values will be verified by the application example to the DC-DC converter shown in the third embodiment.

(Embodiment 2)

FIG. 7 shows another example of the cross-sectional structure of the lossy pulse transformer.

FIG. 8 is an example of the wiring structure of the loss pulse transformer mounting surface of the printed wiring board 38. FIG. 9 is a cross-sectional view of the printed wiring board 38 as viewed from X-X ′ in FIG. 8.

7 and 8, the loss pulse transformer includes a winding frame 15, a magnetic core 16, a winding 17, primary winding external terminals 19 and 20, primary winding internal terminals 21 and 22, and a primary wiring printed wiring board 38. The connection terminals 32 and 33, the secondary winding internal terminals 25 and 26, the secondary winding external terminals 34 and 35, and the winding terminals 15 are arranged and fixed in isolation and connected to the vias 83 and 84 of the printed wiring board 38. Terminal 36 and 37, and printed wiring board 38. The terminals 34 and 35 pass through the central portions of the hole 81 and 82 of the printed wiring board 38 in a non-contact manner, and are directly connected to the printed wiring board on which the loss pulse transformer is mounted.

7 and 8, the terminals 19 and 32 are connected to vias 77 and 79 connected to the ground plane 90 of the printed wiring board 38. The terminal 20 is connected to a via 78 connected to one end of a strip conductor 89 on the printed wiring board 38. The terminal 33 is connected to a via 80 connected to the other end of the strip conductor 89 on the printed wiring board 38.

The printed wiring board 38 is configured by bonding two single-sided copper-clad plates with a semiconductor prepreg.

FIG. 9 is a cross-sectional view of the printed wiring board 38 as viewed from X-X ′ in FIG. 8.

In FIG. 9, the microstrip line composed of the strip conductor 89, the ground plane 90, the insulator layer 86, and the semiconductor layer 87 constitutes a loss line.

8 and 9, the strip conductor 89 and the ground plane 90 have a thickness of 35 [μm], the insulator layer 86 has a thickness of 500 [μm], and the semiconductor layer 87 has a thickness of 50 [μm]. When the width of the 89 strip conductor is 1 [mm] and the relative dielectric constant of the insulator layer is 4, the characteristic impedance of the microstrip line is 49.6 [Ω] from the equation (9).

8 and 9, when the conductivity of the insulator layer is infinite, the conductivity of the semiconductor layer is 100 [S / m], and the effective length of the strip conductor is 1 [cm], the terminals 78 and 79 are connected. The transmission coefficient S ₂₁ of the loss line on the printed wiring board 38 when connected to the transmission line on the printed wiring board of 50 [Ω] via the equations (10), (11), (17) From (18), -43 dB at 1 [GHz] and -162 dB at 10 [GHz]. The validity of these values will be verified in the application example of the lossy pulse transformer shown in the third embodiment to the DC-DC converter.

(Embodiment 3)
This embodiment relates to the behavior of an isolated electromagnetic wave on a transmission line when the lossy pulse transformer of Embodiments 1 and 2 is used for a forward type DC-DC converter.

In FIG. 10, the forward type DC-DC converter includes a DC power source 51, a switching N-channel MOS FET 55, a transmission line 54 connected between the N-channel MOS FET 55 and the DC power source 51 and made up of conductors 52 and 53, Loss pulse transformer 60, transmission line 58 composed of conductors 56 and 57 connected between primary winding terminals 62 and 63 of loss pulse transformer 60 and MOS FET 55, and primary winding terminals 61 and 63 of loss pulse transformer 60. Are connected between the choke coil 71, the choke coil 71, the choke coil 71, and the diode 67. A transmission line 70 composed of conductors 68 and 69 and a capacitor 72 are included. The characteristic impedance of the transmission lines 54 and 58 is 50 [Ω].

In FIG. 10, when the N-channel MOS FET 55 is turned on, an isolated electromagnetic wave having a negative polarity shown in FIG. 11 is excited at the point C on the transmission line 58 during the rising period. The isolated electric field wave 41 constituting the isolated electromagnetic wave excited at the point C has an exponential amplitude along the leading envelope curve 43 shown in FIG. 11 while maintaining the wavelength λ _s defined by the equation (16). It proceeds in the transmission line 58 while decreasing the point D and reaches the point D.

The exponentially decaying electric field generates a potential that compensates for the attenuation of the solitary electric field wave according to equation (2). This electric field is an electrostatic field, and is generated when an isolated electric field wave draws electrostatic energy from the DC power source 51 in FIG. 10 onto the transmission line.

The isolated electromagnetic wave travels while keeping the potential of the transmission line 58 constant from the C point to the D point while distributing the electrostatic field on the transmission line. In this embodiment, since the transmission line 52 and the transmission line 58 have the same characteristic impedance, the potential of the transmission line 50 at which the isolated electric field wave 7 reaches the point D becomes a steady value of −E / 2 [V].

The loss pulse transformer of the first or second embodiment is connected to the point D. Although the lossy pulse transformers of the first and second embodiments do not have a primary winding connected to the terminal 61 and the terminal 62 in FIG. 10, they are omitted because they are unnecessary for the description in the present embodiment.

The isolated electromagnetic wave 7 reaching the point D is attenuated by −43 dB at 1 [GHz] by the loss line in the loss pulse transformer 60 in the process of traveling toward the primary winding of the loss pulse transformer 60. The greatly attenuated isolated electromagnetic wave 7 proceeds to the secondary winding of the loss pulse transformer 60 in accordance with the concept of displacement current or electric flux current. However, since the electrostatic energy extracted by the isolated electromagnetic wave cannot pass through the insulator, The rise in potential stops until the primary winding. Note that the effective frequency of an N-channel MOS FET having a rise time of 0.3 [ns] is 1 [GHz].

FIG. 11 is an example of an isolated electric field wave at points C and D on the transmission line 58 and an envelope curve at the tip of the isolated electric field wave.

The envelope curve at the tip of the isolated electromagnetic wave when the isolated electromagnetic wave travels on the loss line in the loss pulse transformer 60 is an attenuation curve obtained from the exponential term of Equation (11). The change in potential in the length direction of the transmission line due to the isolated electric field wave when the point of the isolated electromagnetic wave attenuates by the exponential term of Equation (11) is considered to depend on the attenuation characteristic of the electric field, and is obtained from the following equation. .

Equation (19) shows that the electrostatic energy compensates for the decrease in the ability to raise the potential of the loss line while the isolated electric field wave is traveling on the loss line in the loss pulse transformer 60. The potential of −E / 2 [V] of the loss line in the ongoing loss pulse transformer 60 and the potential of −E / 2 [V] of the primary winding of the loss pulse transformer 60 are not attenuated.

In FIG. 10, when the N-channel MOS FET 55 is turned on, the isolated electromagnetic wave having the positive polarity shown in FIG. 12 is excited at the point B on the transmission line 54 at the same time as the point C on the transmission line 58. Is done. The isolated electric field wave 42 constituting the isolated electromagnetic wave excited at the point B maintains the wavelength λ _s defined by the equation (16), and the potential of the transmission line 54 is changed from −E [V] to −E / 2.
Proceed toward point A while increasing to [V].

It is assumed that the DC power source 51 connected to the point A is an ideal power source and has a terminal impedance of zero. When the isolated electric field wave 42 reaches A, the reflected electric field wave 42 is reflected to reverse the polarity, and the potential of the transmission line 54 is returned to -E [V] to advance to the point B. When the isolated electric field wave 42 reaches the point B, if the N-channel MOS FET 55 is kept on, the isolated electric field wave 42 goes to the point D through the point C on the transmission line 58.

The isolated electromagnetic wave 42 that reaches the point D is attenuated by −43 dB at 1 [GHz] by the loss line in the loss pulse transformer 60 in the process of traveling toward the primary winding of the loss pulse transformer 60, as described above. The potential of the primary winding of the loss pulse transformer 60 is set to a steady value of −E [V]. Therefore, the loss line in the loss pulse transformer 60 does not reduce the electrostatic energy supply capability of the N-channel MOS FET 55 for driving the loss pulse transformer 60.

The isolated electromagnetic wave 7 greatly attenuated on the loss line in the loss pulse transformer 60 proceeds to the secondary winding of the loss pulse transformer 60 according to the concept of displacement current or electric flux current, but the electrostatic energy extracted by the isolated electromagnetic wave is insulated. Since it cannot penetrate the body, the potential drop to -E [V] stops at the primary winding.

Until the isolated electromagnetic wave 7 reaches the primary winding of the loss pulse transformer 60, electrostatic energy for charging the line 58 is supplied from the DC power supply 51. After the isolated electromagnetic wave 7 reaches the primary winding of the loss pulse transformer 60, electrostatic energy is continuously supplied from the DC power source 51 in order to accumulate magnetostatic energy in the loss pulse transformer 60 until the N-channel MOS FET 55 is turned off. Is done. During this period, supply of direct current from the direct current power supply 51 is observed, but this current is a steady current. The current during the period for charging the transmission line 58 is a value obtained by dividing the potential on the transmission line 58 by the characteristic impedance of the transmission line 58.

When an isolated electromagnetic wave passes on the transmission line 58 and through the lossy pulse transformer 60, a displacement current or an electric flux current is observed, which is an instantaneous current due to the electromagnetic energy held by the isolated electromagnetic wave, that is, a pointing vector. .

FIG. 13 is an example of a potential curve 45 on the transmission line represented on the time axis of point A on the transmission line 54. FIG. 14 is an example of a potential curve 46 on the transmission line represented on the time axis of point B on the transmission line 54. FIG. 15 is an example of a potential curve 47 on the transmission line represented on the time axis of point C on the transmission line 58. FIG. 16 is an example of a potential curve 48 on the transmission line represented on the time axis of point D on the transmission line 58. The current is as described above, and the illustration is omitted.

In FIG. 13 to FIG. 16, t ₁ is the time until the isolated electromagnetic wave excited at the point C on the transmission line 58 reaches the point D. t ₂ is the time from when the isolated electromagnetic wave excited at the point B on the transmission line 54 is reflected at the point A and returns to the point B. t ₃ is the time from when the isolated electromagnetic wave excited at the point B on the transmission line 54 is reflected at the point A and reaches the point D on the transmission line 58 via the points B and C.

From FIG. 16, the rise time of the signal waveform at point D on the transmission line 58 is independent of the loss or attenuation constant of the transmission line 54 and the transmission line 58, and the rise time of the N-channel MOS FET 55 that is the wave source. It can be seen that this depends on the length of the transmission line 54 connected between the N-channel MOS FET 55 and the DC power supply 51.

The electric field intensity E at a distance r [m] when a linear electromagnetic wave having radiated power P is radiated from the antenna can be obtained from the following equation shown in IEC CISPR 16-2-3.

For example, the permissible value of the interference wave electric field strength at a distance of 10 [m] from a class B information technology device intended for home use is determined by VCCI (CISPR22), and is from 30 [MHz] to 230 [MHz]. 30 [dBμV / m] and 230 [MHz] to 1 [GHz] and 37 [dBμV / m]. From the equation (20), for example, the allowable radiated power value at 230 [MHz] is 2 [nW].

FIG. 17 is an example of a cross-sectional structure of a conventional pulse transformer.

In FIG. 17, the conventional pulse transformer includes a winding frame 92, a magnetic core 93, a winding 94, a resin 91, and a terminal 95.

When the input power of the forward type DC-DC converter of FIG. 10 is 100 [W], the rise time of the N-channel MOS FET 55 is 0.3 [ns], and the on-time is 30 [ns], the energy of the isolated electromagnetic wave is 1 [W]. Assuming that 0.1% of this is radiated from the pulse transformer into the atmosphere, and that 10 ⁻⁴ of that is included in the linear electromagnetic wave of 230 [MHz], the radiation disturbance wave of 230 [MHz] Is 100 [nW]. This value significantly exceeds the allowable radiated power value of the class B information technology device.

In the lossy pulse transformers of the first and second embodiments, the transmission coefficient of the loss line in the lossy pulse transformer is −43 dB at 1 [GHz], which is the effective frequency when the rise time is 0.3 [ns]. Therefore, the power of the 230 [MHz] radiated disturbance wave under the above condition is 0.7 [nW], which sufficiently satisfies the allowable radiated power value of the class B information technology device.

In the loss pulse transformers of the first and second embodiments, the length of the loss line in the loss pulse transformer is set to 1 [cm]. Since the transmission coefficient decreases exponentially with respect to the length of the loss line as shown in the equation (11), a large attenuation characteristic can be easily obtained by selecting the length. Therefore, according to this patent, it is considered that the EMC problem caused by the pulse transformer is almost solved without using any EMC countermeasure parts or electromagnetic shielding materials.

(Embodiment 4)
FIG. 18 is an example of the structure of a prototyped loss line having a low impedance.

The loss line having a low impedance is composed of a conductor 96, a valve metal 97, an insulator 98, a semiconductor 99, and a conductive adhesive 100. The valve metal 97 is drawn out in the line length direction as shown in FIG. Yes. Both ends of the drawn valve metal 97 in the line length direction are anode terminals, and both ends of the conductor 96 in the line length direction are negative terminals.

An experimentally produced lossy line having a low impedance uses an aluminum thin film subjected to an etching process having a line part width of 1 [mm] and a length of 16 [mm] as the valve metal 97. An aluminum oxide film having a thickness of 10 nm formed by chemical conversion treatment on the etched portion of the aluminum thin film corresponds to the insulator 98. Polypyrrole deposited on the etched portion of the aluminum thin film by chemical polymerization corresponds to the semiconductor 99 and has a thickness of about 2.5 μm.

The conductive graphite 100 corresponds to carbon graphite coated on polypyrrole to a thickness of about 30 [μm] and thermosetting silver paste coated on the carbon graphite. A copper plate having a width of 1 [mm] and a length of 16 [mm] is bonded by the conductive adhesive 100, and this corresponds to the conductor 99. The electrical conductivity of polypyrrole used as a semiconductor is 1500 [S / m], and the relative dielectric constant of aluminum oxide used as an insulator is 10.

FIG. 19 is an example of the transmission coefficient (S ₂₁ ) characteristic of a prototyped loss line with low impedance.

FIG. 19 shows characteristics when the length of the loss line portion having low impedance is 4 [mm], 8 [mm], 16 [mm], and 24 [mm], and two conventional chip ceramics. The characteristic of the capacitor is shown.

When the capacitance of the parallel plate constituting the loss line having a low impedance is C, the expansion ratio k of the facing area by etching can be obtained from the following equation.

The impedance Z _{C of} the capacitor assumed to be used in parallel with the circuit can be obtained from the scattering matrix transmission coefficient (S ₂₁ ). When the characteristic impedance Z ₀ of the measurement system cable is 50 [Ω] and the transmission coefficient (S ₂₁ ) is considerably smaller than 1, the relationship between Z _C and the transmission coefficient (S ₂₁ ) is simplified to become that way.

In FIG. 19, the transmission coefficient (S ₂₁ ) of a lossy line having a low impedance of a length of 16 [mm] and a width of 1 [mm] at 100 k [kHz] considered to be acting as a capacitor is about −46 dB. It is. The impedance Z _C as a capacitor at this time can be obtained from the equation (22), and the capacitance C is 12 [μF]. When this value and other predetermined parameters are substituted into equation (21), the enlargement ratio k of the facing area by etching is 85, and the enlargement ratio k ′ of the width is 9.2.

In this case, the characteristic impedance when acting as a parallel plate line can be obtained by setting w in Equation (7) as 9.2w, and is 0.13 [mΩ].

The transmission coefficient (S ₂₁ ) of the low-impedance loss line of the present embodiment with respect to a line having a characteristic impedance of 50 [Ω] is obtained using equations (10), (11), (17) and (18). I can do it. When the line length is 8 [mm], −63 dB at 10 [MHz], −82 dB at 100 [MHz], and −129 dB at 1 [GHz]. S ₂₁ when the line length is 16 [mm] is similarly 10
[MHz] is −77 dB, 100 [MHz] is −116 dB, and 1 [GHz] is −208 dB.

These characteristics substantially coincide with the characteristics shown in FIG. The difference between the actual measurement and the calculation result is that the structure of the etched portion of the aluminum thin film is very complicated. Therefore, the characteristic impedance and transmission coefficient (S ₂₁ ) of the loss line in the first and second embodiments can be obtained almost accurately by calculation when the etching process is not performed, and when the etching process is performed. However, it can be obtained by calculation with accuracy that does not hinder practical use.

The present invention makes it possible to design and analyze a pulse transformer used in a high-frequency DC-DC converter very easily by following the isolated electromagnetic wave concept.

The present invention can easily improve conversion efficiency, signal quality (signal integrity), and electromagnetic environment compatibility (EMC) of a pulse transformer used for a frequency DC-DC converter even by an unskilled engineer.

In addition, the present invention can reduce the use of electromagnetic shielding parts of a pulse transformer, and EMC countermeasure parts and materials of equipment using the pulse transformer, and a snubber and a matching termination circuit that are indispensable for a pulse transformer that applies a high-speed switching pulse. This makes it possible to reduce the size and weight of the equipment, reduce manufacturing costs, and shorten the design period.

FIG. 1 is an example of an equivalent circuit related to a push-pull circuit. FIG. 2 is an example of an electric field waveform and a potential waveform on the load-side transmission line. FIG. 3 is an example of an electric field waveform and a potential waveform on the transmission line on the power supply side. FIG. 4 is an example of a transmission line structure. FIG. 5 is an example of a cross-sectional structure of the loss pulse transformer. FIG. 6 is an example of the internal wiring structure of the loss pulse transformer. FIG. 7 shows another example of the cross-sectional structure of the lossy pulse transformer. FIG. 8 is an example of the wiring structure of the loss pulse transformer mounting surface of the printed wiring board 38. FIG. 9 is a cross-sectional view of the printed wiring board 38 as viewed from X-X ′ in FIG. 8. FIG. 10 is an example of a forward type switching power supply circuit. FIG. 11 is an example of an isolated electric field wave traveling on the transmission line 58 of FIG. 10 and an envelope curve of the cusp of the isolated electric field wave. FIG. 12 is an example of an isolated electric field wave traveling on the transmission line 54 of FIG. 10 and an envelope curve of the cusp of the isolated electric field wave. FIG. 13 is an example of a time-axis potential waveform at point A on the transmission line 54 in FIG. FIG. 14 is an example of a time-axis potential waveform at point B on the transmission line 54 in FIG. FIG. 15 is an example of a time-axis potential waveform at point C on the transmission line 58 in FIG. FIG. 16 is an example of a time-axis potential waveform at point D on the transmission line 58 in FIG. FIG. 17 is an example of a cross-sectional structure of a conventional pulse transformer. FIG. 18 is an example of the structure of a prototyped loss line having a low impedance. FIG. 19 is an example of transmission (S ₂₁ ) characteristics of a prototyped loss line having low impedance.

Explanation of symbols

1
Push-pull circuit
2
P-channel MOS FET
Three
55 N-channel MOS FET
Four
51 DC power supply
Five
, 6,54,58,70 Transmission line
7
Matching termination resistor
8
, 10, 41, 42 Isolated electric wave
9
, 11, 45, 46, 47, 48 Potential waveform on the transmission line
12
, 52, 53, 56, 57, 68, 69 conductors
13
, 98 insulator
14
, 99 semiconductors
15
, 92 reel
16
, 93 magnetic core
17
94 windings
18
, 91 resin
19
, 20, 21, 22, 23, 24, 25, 26, 27, 28, 31, 32, 33, 34, 35, 61, 62, 63, 64, 65, 73, 74, 95 terminals
29
Resin substrate
30
Resin box
31
Lossy track
38
Printed wiring board
43
Envelope curve of the cusp of an isolated electric field wave
59
, 66, 67 Diode
60
Loss pulse transformer
77
78, 79, 80, 83, 84 Via
81
, 82 Stupid hole
85
88 conductor layers
86
Insulator layer
87
Semiconductor layer
89
Strip conductor
90
Ground plane
96
conductor
97
Valve metal
100 conductive adhesive

Claims (10)

In a pulse transformer having at least one primary winding and at least one secondary winding driven by a switching element, the pulse transformer has a rising waveform of the switching element on a primary winding of the pulse transformer. When an electromagnetic wave having an effective frequency defined by a frequency obtained as a reciprocal of the product of the rising time and the circumference is obtained by regarding the tangent line of the maximum slope as a rising waveform, the electromagnetic wave is generated inside the pulse transformer. A pulse transformer characterized by being designed as a lossy pulse transformer having an ability to attenuate the amplitude to 1/10 or less. The loss pulse transformer according to claim 1, wherein the wiring connecting the winding structure constituting the loss pulse transformer and the external connection terminal includes a conductor, an insulator formed on the conductor surface, and a surface of the insulator Or a loss line composed of a conductor, the semiconductor formed on the surface of the conductor, and an insulator formed on the surface of the semiconductor. Characteristic pulse transformer 3. The lossy pulse transformer according to claim 1, wherein a length of the loss line constituting the loss pulse transformer is 10 [mm] or more. 4. The loss pulse transformer according to claim 1, wherein the characteristic impedance of the loss line constituting the loss pulse transformer is within ± 20% of the characteristic impedance of a printed wiring board on which the loss pulse transformer is mounted. Pulse transformer characterized by equality 5. The loss pulse transformer according to claim 1, wherein the loss line constituting the loss pulse transformer is formed by embedding two conductors covered with a semiconductor film in a flexible insulating resin. A pulse transformer characterized by 6. The loss pulse transformer according to claim 1, wherein the loss line constituting the loss pulse transformer is formed on a printed wiring board having an insulating layer and a semiconductor layer between two conductor layers. Pulse transformer featuring 7. The lossy pulse transformer according to claim 1, wherein the semiconductor is an inorganic semiconductor or an organic semiconductor and is designed to have a conductivity of 10 [S / m] or more. Pulse transformer 8. The lossy pulse transformer according to claim 1, wherein the semiconductor is poly (p-phenylene) (22.7 °), poly (p-phenylene sulfide), carbon graphite, manganese dioxide, or polyacetylene. , Or polythiophene, or polypyrrole, or polyphenylene vinylene, or tetrathiafulvalene-tetraquinodimethane (TTF-TCNQ) 9. The loss pulse transformer according to claim 1, wherein the semiconductor is n-type silicon doped with phosphorus of 10 ¹⁵ [cm ⁻³ ] or more, p-type doped with boron of 10 ¹⁶ [cm ⁻³ ] or more. It is characterized by being silicon, or amorphous silicon mixed with impurities, or alumina, or zirconia, or carbide, or nitride, or silicide, or silicon carbide, or silicon nitride, or magnesium nitride, or zinc oxide. Pulse transformer 10. The lossy pulse transformer according to claim 1, wherein the semiconductor is a resin in which the semiconductor according to claim 4 or the semiconductor particles having a maximum length of 100 nanometers or less are mixed. Characteristic pulse transformer