High Frequency Amplifier
The high frequency amplifier addresses impedance and phase variations by using asymmetrical matching circuit patterns with adjusted transmission line lengths and chamfered corners, improving power efficiency and output.
Patent Information
- Application Number
- JP2022517059
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2020-04-23
- Filing Date
- 2021-04-20
- Publication Date
- 2024-11-13
- Estimated Expiration
- 2041-04-20
AI Technical Summary
Existing high frequency amplifiers face issues with variations in effective impedance and phase of RF signals due to mutual inductance components between bonding wires, leading to reduced power efficiency.
The high frequency amplifier design includes transistors connected via wires with asymmetrical matching circuit patterns, where the electrical length of transmission lines is adjusted to cancel out variations in effective impedance and phase, using chamfered corners and varying wire thickness to match electrical lengths.
This configuration reduces phase variations between transistors, enhancing power efficiency and output in high-output high-frequency amplifiers.
Smart Images

Figure 0007582693000001 
Figure 0007582693000002 
Figure 0007582693000003
Abstract
Description
[Technical field]
[0001] The present disclosure relates to high frequency amplifiers. This application claims priority to Japanese Application No. 2020-076598, filed on April 23, 2020, and incorporates by reference all of the contents of said Japanese application. [Background technology]
[0002] For example, Patent Document 1 discloses a field effect transistor (FET) as a high-frequency amplifier. E affect T This field effect transistor includes a plurality of amplifying elements for amplifying radio frequency (RF) signals, and a matching circuit that is connected to the input ends of the amplifying elements and the input terminal of the package via bonding wires and performs impedance conversion. [Prior art documents] [Patent documents]
[0003] [Patent Document 1] Japanese Patent Application Publication No. 63-86904 Summary of the Invention
[0004] The present disclosure provides a high frequency amplifier. This high-frequency amplifier includes a first transistor, a second transistor arranged in a first direction with respect to the first transistor, and a third transistor arranged in the first direction opposite to the first transistor with respect to the second transistor, a first drain pad electrically connected to a drain electrode of the first transistor, a second drain pad electrically connected to a drain electrode of the second transistor, and a third drain pad electrically connected to a drain electrode of the third transistor, a first transmission line electrically connected to the first drain pad, a second transmission line electrically connected to the second drain pad, and a third transmission line electrically connected to the third drain pad, and includes a matching circuit pattern for performing impedance matching of a high-frequency signal for each of the first transistor, the second transistor, and the third transistor, a first wire electrically connecting the first transmission line and the first drain pad, a second wire electrically connecting the second transmission line and the second drain pad, and a third wire electrically connecting the third transmission line and the third drain pad, and a wiring pattern electrically connected to the first drain pad via the first transmission line and the first wire, and electrically connected to the second drain pad via the second transmission line and the second wire. The effective impedance of the second wire is greater than the effective impedance of the first wire. The matching circuit pattern has an asymmetrical outer shape with respect to a second virtual line that is perpendicular to a first virtual line connecting a first connection point to which the first wire of the first transmission line is connected and a second connection point to which the second wire of the second transmission line is connected and that passes through a midpoint between the first connection point and the second connection point. The electrical length of the second transmission line is shorter than the electrical length of the first transmission line. [Brief description of the drawings]
[0005] [Figure 1] FIG. 1 is a plan view showing the internal configuration of a high-frequency amplifier 1A according to a first embodiment and a high-frequency amplifier 1B according to a second embodiment. [Diagram 2] FIG. 2 is a plan view showing the amplifying element 11 and the matching circuit 50. As shown in FIG. [Diagram 3]FIG. 3 is an enlarged plan view of one matching circuit pattern 52 in FIG. [Figure 4] FIG. 4 is a perspective view for explaining the electrical length of each of the bonding wires W41, W42, W43, W44, W45, W46, W47, and W48. [Diagram 5] FIG. 5 is a Smith chart showing the impedance of each of the bonding wires W41, W42, W43, W44, W45, W46, W47, and W48. [Figure 6] FIG. 6 is a graph showing the imaginary part of the impedance of each of the bonding wires W41, W42, W43, W44, W45, W46, W47, and W48. [Figure 7] FIG. 7 is a diagram for explaining the relationship between the length of a physical path along which an electromagnetic wave travels and the phase. [Figure 8] FIG. 8 is a diagram for explaining the relationship between the length of a physical path along which an electromagnetic wave travels and the phase. [Figure 9] FIG. 9 is a diagram for explaining a distribution state of current in microwaves. [Figure 10] FIG. 10 is a simulation diagram showing a state of current concentration in microwaves. [Figure 11] FIG. 11 is a simulation diagram showing the distribution state of the electric field intensity of microwaves. [Figure 12] FIG. 12 is a plan view showing a reference example of a pattern 80A through which electromagnetic waves pass. [Figure 13] FIG. 13 is a plan view showing another reference example of a pattern 80B through which electromagnetic waves pass. [Figure 14] FIG. 14 is a graph showing the phase difference occurring between two electromagnetic waves that have passed through paths DA and DB in patterns 80A and 80B, respectively. [Figure 15] FIG. 15 is a plan view showing the matching circuit 60. As shown in FIG. [Figure 16] FIG. 16 is an enlarged plan view of one matching circuit pattern 62 in FIG. [Figure 17]FIG. 17 is a Smith chart showing the load impedances of the transistors 13A, 13B, 13C, and 13D of the high-frequency amplifier 1B according to the second embodiment. [Figure 18] FIG. 18 is a diagram showing the relationship between the phase variation of the load impedance and the power efficiency in the second embodiment. [Figure 19] FIG. 19 is a plan view showing a matching circuit 60X of a high-frequency amplifier according to a comparative example. [Figure 20] FIG. 20 is a Smith chart showing the load impedances of the transistors 13A, 13B, 13C, and 13D of the high-frequency amplifier according to the comparative example. [Figure 21] FIG. 21 is a diagram showing the relationship between the phase variation of the load impedance and the power efficiency in the comparative example. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0006] [Problem that this disclosure aims to solve] In the high-frequency amplifier disclosed in Patent Document 1, a plurality of transistors are arranged in parallel in the amplifying element, and each of the transistors is connected to a matching circuit by a wire such as a bonding wire. In such a case, the mutual inductance components of the plurality of wires are affected by other nearby wires. As a result, the effective impedance may vary between the plurality of wires depending on the degree to which the wires are affected by the other wires. In this case, the phase of the RF signal varies between the plurality of wires, so that the phase of the RF signal may also vary between the plurality of transistors. In order to improve the power efficiency of the high-frequency amplifier, it is desirable to reduce the phase variation of the RF signal.
[0007] [Effects of this disclosure] According to a high-frequency amplifier according to an embodiment of the present disclosure, it is possible to reduce the variation in phase of an RF signal between a plurality of transistors.
[0008] [Description of the embodiments of the present disclosure] First, the contents of the embodiments of the present disclosure will be listed and described. a first transmission line electrically connected to the first drain pad, a second transmission line electrically connected to the second drain pad, and a third transmission line electrically connected to the third drain pad, the first transmission line being electrically connected to the first drain pad, and the third transmission line being electrically connected to the third drain pad, the first transmission line being electrically connected to the first drain pad, the second transmission line being electrically connected to the second drain pad, and the third transmission line being electrically connected to the third drain pad, the second transmission line being electrically connected to the third drain pad, the third transmission line being electrically connected to the third drain pad, the second transmission line being electrically connected to the first drain pad, and the third transmission line being electrically connected to the third drain pad, the second transmission line being electrically connected to the third drain pad, the second transmission line being electrically connected to the third drain pad, and the third transmission line being electrically connected to the third drain pad; The effective impedance of the second wire is greater than the effective impedance of the first wire. The matching circuit pattern has an asymmetrical outer shape with respect to a second virtual line that is perpendicular to a first virtual line connecting a first connection point to which the first wire of the first transmission line is connected and a second connection point to which the second wire of the second transmission line is connected and that passes through a midpoint between the first connection point and the second connection point. The electrical length of the second transmission line is shorter than the electrical length of the first transmission line.
[0009] In this high-frequency amplifier, a first transistor, a second transistor, and a third transistor as a plurality of transistors are connected to the matching circuit pattern by a first wire, a second wire, and a third wire as a plurality of wires, respectively. In this configuration, the effective impedance may vary among the plurality of wires depending on the degree of influence from another wire. Such variation in the effective impedance among the plurality of wires appears as variation in the electrical length among the plurality of wires. Here, in this high-frequency amplifier, since the matching circuit pattern has an asymmetrical outer shape, the electrical length of the second transmission line is shorter than the electrical length of the first transmission line. As a result, in a configuration in which the electrical length of the second wire is longer than the electrical length of the first wire, the electrical length difference between the wires can be canceled out by the electrical length difference between the transmission lines. Therefore, the variation between the electrical length from the first transistor to the wiring pattern and the electrical length from the second transistor to the wiring pattern is reduced. Since the smaller the variation in electrical length among the plurality of wires, the smaller the variation in effective impedance among the plurality of wires, this configuration can reduce the phase variation between the first transistor and the second transistor.
[0010] In the above-mentioned high-frequency amplifier, the matching circuit pattern has a first pad including a first corner and a second pad including a second corner, the first transmission line is constituted by a portion of the first pad including the first corner, and the second transmission line is constituted by a portion of the second pad including the second corner, and the second corner may be chamfered by an amount greater than that of the first corner. According to the knowledge of the present inventors, at high frequencies, current (electric field) tends to concentrate on the outer edge of a conductor. Therefore, in the first transmission line, a high-frequency signal propagates along the first corner of the first pad, and the second transmission line is constituted by a portion of the second pad including the second corner, and the second corner is chamfered by an amount greater than that of the first corner. lineIn the second pad, the high-frequency signal propagates along the second corner of the second pad. Here, the second corner is chamfered by a larger amount than the amount of chamfering of the first corner, so that the path including the second corner is shorter than the path including the first corner, and a configuration can be realized in which the electrical length of the second transmission line is shorter than the electrical length of the first transmission line. Furthermore, with this configuration, it is easy to match the electrical length from the first transistor to the wiring pattern and the electrical length from the second transistor to the wiring pattern by adjusting the amount of chamfering. Therefore, it is possible to more reliably reduce the phase variation between the first transistor and the second transistor.
[0011] In the above-mentioned high-frequency amplifier, the first transmission line may be thinner than the second transmission line. In this case, the impedance per unit length of the first transmission line can be made larger than the impedance per unit length of the second transmission line, so that a configuration in which the electrical length of the second transmission line is shorter than the electrical length of the first transmission line can be realized. In addition, by making the first transmission line thinner, it is easy to match the electrical length from the first transistor to the wiring pattern and the electrical length from the second transistor to the wiring pattern. Therefore, it is possible to more reliably reduce the phase variation between the first transistor and the second transistor.
[0012] In the above-mentioned high-frequency amplifier, a plurality of transistor groups each including a first transistor, a second transistor, and a third transistor may be provided, whereby a high-output high-frequency amplifier can be realized.
[0013] In the above-mentioned high-frequency amplifier, a first wire may be disposed on the outermost side of the transistor group, a second wire may be disposed between the third wire and the first wire, the length of the first wire may be equal to the length of the second wire, and an effective electrical length from the first drain pad to the wiring pattern may be substantially equal to an effective electrical length from the second drain pad to the wiring pattern. In this case, the phases of the first transistor and the second transistor are substantially the same, which is particularly advantageous in terms of improving the power efficiency of the high-frequency amplifier.
[0014] [Details of the embodiment of the present disclosure] Specific examples of a high-frequency amplifier according to an embodiment of the present disclosure will be described below with reference to the drawings. Note that the present disclosure is not limited to these examples, but is indicated by the claims, and is intended to include all modifications within the meaning and scope equivalent to the claims. In the following description, the same elements or elements having the same functions are given the same reference numerals, and duplicated descriptions may be omitted. In the description, the XYZ Cartesian coordinate system shown in the drawings may be referenced.
[0015] FIG. 1 is a plan view showing the internal configuration of a high-frequency amplifier 1A according to the first embodiment and a high-frequency amplifier 1B according to the second embodiment. First, the high-frequency amplifier 1A according to the first embodiment will be described. The high-frequency amplifier 1A includes one input terminal 2, one output terminal 3, an amplifying element section 10, a branch circuit board 20, a composite circuit board 30, a matching circuit 40, and a matching circuit 50. In this embodiment, the high-frequency amplifier 1A includes, as an example, two matching circuits 40 and 50. The amplifying element section 10 includes two amplifying elements 11. The output of each amplifying element 11 is, for example, 30 W, and the output of the entire amplifying element section 10 is, for example, 60 W. The high-frequency amplifier 1A includes a package 4 that houses the amplifying element section 10, the branch circuit board 20, the composite circuit board 30, and the matching circuits 40 and 50, and bonding wires W1, W2, W3, W4, W5, and W6.
[0016] The package 4 is made of metal and is connected to a reference potential. The planar shape of the package 4 is substantially rectangular. The package 4 has side walls 4c, 4d facing each other in a first direction and end walls 4a, 4b facing each other in a second direction. The first direction and the second direction intersect with each other, and in one example, are perpendicular to each other. In this embodiment, the first direction is the X-axis direction, and the second direction is the Y-axis direction.
[0017] The package 4 has a flat rectangular bottom plate 4e. The bottom plate 4e extends along a plane defined by the Y-axis direction and the X-axis direction. The end walls 4a, 4b stand along a pair of sides (sides extending along the X-axis direction) of the bottom plate 4e, and the side walls 4c, 4d stand along another pair of sides (sides extending along the Y-axis direction) of the bottom plate 4e. The package 4 further has a lid (not shown). The lid seals an upper opening formed by the end walls 4a, 4b and the side walls 4c, 4d.
[0018] The input terminal 2 is a metal wiring pattern that inputs a high-frequency signal from outside the high-frequency amplifier 1A. The high-frequency signal is a signal based on a multi-carrier transmission method, and is formed by superimposing a plurality of signals with different carrier signal frequencies. The frequency band of the carrier signal is, for example, 500 MHz or less. The input terminal 2 is provided at the center of the end wall 4a in the X-axis direction, and extends from the outside to the inside of the package 4.
[0019] The output terminal 3 is a metallic wiring pattern that outputs the amplified high-frequency signal to the outside of the high-frequency amplifier 1 A. The output terminal 3 is provided at the center of the end wall 4 b in the X-axis direction, and extends from the inside of the package 4 to the outside.
[0020] The amplifying element section 10 is disposed on the bottom plate 4e of the package 4, approximately in the center of the package 4 in the Y-axis direction. The two amplifying elements 11 in the amplifying element section 10 are disposed side by side in the X-axis direction. The amplifying element 11 includes a plurality of transistors 13 (see FIG. 2). The plurality of transistors 13 are, for example, field effect transistors (FETs), and in one embodiment, high electron mobility transistors (HEMTs). Each of the plurality of transistors 13 has a gate electrode, a source electrode, and a drain electrode. Each transistor 13 amplifies an input high frequency signal, and outputs the amplified high frequency signal. A more specific configuration of the amplifying element 11 will be described later.
[0021] The branch circuit board 20 is disposed on the bottom plate 4e of the package 4. The branch circuit board 20 is disposed alongside the input terminal 2 and the amplifier element section 10 along the Y-axis direction, and is located between the input terminal 2 and the amplifier element section 10. The branch circuit board 20 has a ceramic substrate 21 and a branch circuit 22 provided on the main surface of the substrate 21. The substrate 21 has a planar shape of, for example, a rectangle, with one long side 21a facing the input terminal 2 and the other long side 21b facing the amplifier element section 10 via the matching circuit 40. The back surface of the substrate 21 faces the bottom plate 4e of the package 4. One short side 21c of the substrate 21 is located near the side wall 4c of the package 4, and the other short side 21d of the substrate 21 is located near the side wall 4d of the package 4. That is, the substrate 21 extends from near one end of the package 4 to near the other end in the X-axis direction.
[0022] The branch circuit 22 includes a wiring pattern 23 provided on the main surface of the substrate 21. The wiring pattern 23 is electrically connected to the input terminal 2 via a bonding wire W1. A high-frequency signal is input to the wiring pattern 23 from the center of the substrate 21 in the X-axis direction. The wiring pattern 23 has a shape that is linearly symmetrical with respect to the center line of the substrate 21 along the Y-axis direction. The wiring pattern 23 branches into two branches starting from the connection point with the bonding wire W1, and finally reaches eight metal pads 23a. The eight metal pads 23a are arranged side by side along the long side 21b. The adjacent metal pads 23a are connected to each other via a film resistor, forming a Wilkinson type coupler. This ensures isolation between a plurality of gate pads 14 (described later) of the amplifier element section 10, while matching the input impedance of the amplifier element section 10 as viewed from the input terminal 2. Note that only one film resistor 23b is shown in the figure as a representative. The eight metal pads 23a are electrically connected to the matching circuit 40 via bonding wires W2.
[0023] The matching circuit 40 is disposed on the bottom plate 4e of the package 4, and is disposed between the branch circuit board 20 and the amplifying element section 10 in the Y-axis direction. The matching circuit 40 is, for example, a die capacitor, and has a dielectric substrate and a circuit pattern (not shown) provided on the main surface of the dielectric substrate. The circuit pattern has a plurality of metal pads (not shown). The number of the metal pads is, for example, the same as the number of the metal pads 23a. The plurality of metal pads include X-axis direction The metal pads are arranged in a row along the line 11. Each metal pad is electrically connected to a corresponding metal pad 23a via a bonding wire W2, and is also electrically connected to a corresponding gate pad 14 of the amplifier element section 10 via a bonding wire W3.
[0024] In the matching circuit 40, a T-type filter circuit is formed by the inductance components of the bonding wires W2 and W3 and the capacitance of a metal pad connected between a node between these inductance components and a reference potential (bottom plate 4e). The matching circuit 40 performs impedance conversion using this T-type filter circuit. Normally, the impedance seen from the gate pad 14 to the inside of the transistor 13 in the amplifying element section 10 differs from the characteristic impedance of the transmission line (e.g., 50Ω). The matching circuit 40 converts this impedance to 50Ω seen from the input terminal 2 to the inside of the package 4 using the T-type filter circuit.
[0025] The matching circuit 50 is disposed on the bottom plate 4e of the package 4, and is disposed between the amplifying element section 10 and the composite circuit board 30 in the Y-axis direction. The matching circuit 50 is, for example, a parallel plate type capacitor (die capacitor) like the matching circuit 40. As shown in FIG. 2, each matching circuit 50 has a dielectric substrate 51 (see FIG. 2) and a plurality of matching circuit patterns 52 (see FIG. 2) provided on the dielectric substrate 51. The planar shape of the dielectric substrate 51 is a rectangle with the X-axis direction as the longitudinal direction. The thickness of the dielectric substrate 51 (here, the dimension in the Z-axis direction) is, for example, about 200 μm. The relative dielectric constant (εr) of the dielectric substrate 51 is, for example, εr=150. Each matching circuit pattern 52 has a plurality of metal pads 53. Each metal pad 53 is electrically connected to a corresponding drain pad 15 (described later) of the amplifying element section 10 via a bonding wire W4 (wire), and is also electrically connected to a corresponding metal pad 33a (described later) of the composite circuit board 30 via a bonding wire W5. A more specific configuration of the matching circuit 50 will be described later.
[0026] In the matching circuit 50, a T-type filter circuit (matching circuit) is also formed by the inductance components of the bonding wires W4 and W5 and the capacitance of the metal pad 53 connected between the node between these inductance components and the reference potential (bottom plate 4e). The matching circuit 50 performs impedance matching for the amplifying element section 10 by performing impedance conversion using this T-type filter circuit. Normally, the impedance in the amplifying element section 10 looking from the drain pad 15 to the inside of the transistor 13 is different from the characteristic impedance of the transmission line (e.g., 50Ω) and is generally smaller than 50Ω. The matching circuit 50 matches this impedance to 50Ω looking from the output terminal 3 to the inside of the package 4 using the T-type filter circuit.
[0027] The composite circuit board 30 is disposed on the bottom plate 4e of the package 4. The composite circuit board 30 is disposed alongside the amplifier section 10 and the output terminal 3 along the Y-axis direction, and is located between the amplifier section 10 and the output terminal 3. The composite circuit board 30 has a ceramic substrate 31 and a composite circuit 32 provided on the main surface of the substrate 31. The substrate 31 has a planar shape of, for example, a rectangle, with one long side 31a facing the amplifier section 10 via the matching circuit 50, and the other long side 31b facing the output terminal 3. The back surface of the substrate 31 faces the bottom plate 4e of the package 4. One short side 31c of the substrate 31 is located near the side wall 4c of the package 4, and the other short side 31d of the substrate 31 is located near the side wall 4d of the package 4. That is, the substrate 31 extends from near one end of the package 4 to near the other end in the X-axis direction.
[0028] The combining circuit 32 combines signals output from a plurality of drain pads 15 of the amplifier element section 10 to generate one output signal. The combining circuit 32 includes a wiring pattern 33 provided on the main surface of the substrate 31. The wiring pattern 33 has a shape that is linearly symmetrical with respect to the center line of the substrate 31 along the Y-axis direction. The wiring pattern 33 includes four metal pads 33a. The four metal pads 33a are arranged in a line along the long side 31a. The adjacent metal pads 33a are connected to each other via a film resistor to form a Wilkinson type coupler. This ensures isolation between the plurality of drain pads 15 of the amplifier element section 10 while matching the output impedance of the amplifier element section 10 as viewed from the output terminal 3. Each metal pad 33a is electrically connected to two corresponding metal pads 53 of the matching circuit 50 via a bonding wire W5. The wiring pattern 33 is repeatedly coupled from the four metal pads 33a and finally reaches a connection point with the bonding wire W6. The wiring pattern 33 is electrically connected to the output terminal 3 via a bonding wire W6. The amplified high frequency signal is output to the output terminal 3 from the center of the substrate 31 in the X-axis direction.
[0029] Next, the amplifying element 11 and the matching circuit 50 will be described in more detail with reference to FIG. 2. FIG. 2 is a plan view showing the amplifying element 11 and the matching circuit 50 of FIG. 1. The amplifying element 11 includes a semiconductor substrate 12, a plurality of transistors 13, a plurality of gate pads 14, a plurality of drain pads 15, and a plurality of source pads 16. The planar shape of the semiconductor substrate 12 is a rectangle whose longitudinal direction is the X-axis direction. The plurality of transistors 13 are arranged side by side in the X-axis direction on the semiconductor substrate 12. The number of the gate pads 14, the drain pads 15, and the source pads 16 is the same as the number of the transistors 13. Each of the plurality of gate pads 14, the plurality of drain pads 15, and the plurality of source pads 16 is a metal film (e.g., Au film) formed on the main surface of the semiconductor substrate 12.
[0030] The gate pads 14 are electrically connected to the gate electrodes of the transistors 13. The gate pads 14 are arranged in a line along the edge of each amplifier element 11 on the input terminal 2 side. The drain pads 15 are electrically connected to the drain electrodes of the transistors 13. The drain pads 15 are arranged in a line along the edge of each amplifier element 11 on the output terminal 3 side. The source pads 16 are electrically connected to the source electrodes of the transistors 13. The source pads 16 are arranged alternately with the gate pads 14 along the edge of each amplifier element 11 on the input terminal 2 side. The source pads 16 are electrically connected to the bottom plate 4e of the package 4 through via holes that penetrate the amplifier element 11 in the thickness direction (here, the Z-axis direction) and are set to a reference potential. Each transistor 13 amplifies a high-frequency signal input to the gate pad 14 and outputs the amplified high-frequency signal from each drain pad 15.
[0031] FIG. 2 shows one amplifying element 11. In this embodiment, one amplifying element 11 has a transistor group composed of eight transistors 13. That is, in the high-frequency amplifier 1A, a plurality of transistor groups (two in this embodiment) are provided on the bottom plate 4e of the package 4. The eight transistors 13 include transistors 13A, 13B, 13C, 13D, 13E, 13F, 13G, and 13H arranged in this order in the X-axis direction. In other words, the transistor 13D is arranged in the X-axis direction with respect to the transistor 13A, and the transistors 13E, 13F, 13G, and 13H are arranged in the X-axis direction opposite to the transistor 13A with respect to the transistor 13D. The transistor 13A is an example of a first transistor in this embodiment, and the transistor 13D is an example of a second transistor in this embodiment. The transistors 13E, 13F, 13G, and 13H are each an example of a third transistor in this embodiment. Of the eight transistors 13, transistors 13A and 13H are disposed at the outermost positions in the X-axis direction. Transistors 13B to 13G are disposed between transistors 13A and 13H. That is, transistors 13A and 13H are adjacent to another transistor 13 on only one side in the X-axis direction. Moreover, transistors 13B to 13G are adjacent to other transistors 13 on both sides in the X-axis direction.
[0032] Hereinafter, the drain pads 15 electrically connected to the drain electrodes of the transistors 13A to 13H are referred to as drain pads 15A, 15B, 15C, 15D, 15E, 15F, 15G, and 15H, respectively. The drain pad 15A is an example of a first drain pad in this embodiment, and the drain pad 15D is an example of a second drain pad in this embodiment. The drain pads 15E, 15F, 15G, and 15H are each an example of a third drain pad in this embodiment.
[0033] In addition, hereinafter, the eight bonding wires W4 are referred to as bonding wires W41, W42, W43, W44, W45, W46, W47, and W48, respectively, corresponding to the drain pads 15A to 15H. The bonding wire W41 is an example of the first wire in this embodiment, and the bonding wire W44 is an example of the second wire in this embodiment. The bonding wires W45, W46, W47, and W48 are an example of the third wire in this embodiment. The eight bonding wires W4 have the same length. In the eight bonding wires W4, the bonding wires W41 and W48 are respectively arranged at the outermost positions in the X-axis direction. Moreover, the bonding wires W42 to W47 are arranged between the bonding wires W41 and W48.
[0034] In this embodiment, one matching circuit 50 has two matching circuit patterns 52. Each matching circuit pattern 52 has a plurality of transmission lines 54. In the two matching circuit patterns 52, the number of the plurality of transmission lines 54 is the same as the number of transistors 13. In this embodiment, one matching circuit pattern 52 has four transmission lines 54. The four transmission lines 54 include transmission lines 54A, 54B, 54C, and 54D.
[0035] The two matching circuit patterns 52 have a configuration that is inverted from each other in the X-axis direction. In one matching circuit pattern 52, the transmission lines 54A, 54B, 54C, and 54D are arranged in this order in the X-axis direction corresponding to the transistors 13A, 13B, 13C, and 13D. The transmission line 54A in the one matching circuit pattern 52 is an example of the first transmission line in this embodiment. The transmission line 54D in the one matching circuit pattern 52 is an example of the second transmission line in this embodiment. In the other matching circuit pattern 52, the transmission lines 54D, 54C, 54B, and 54A are arranged in this order in the X-axis direction corresponding to the transistors 13E, 13F, 13G, and 13H. The transmission lines 54D, 54C, 54B, and 54A in the other matching circuit pattern 52 are an example of the third transmission line in this embodiment. Of the eight transmission lines 54, two transmission lines 54A are disposed at the outermost positions in the X-axis direction. Also, each of the transmission lines 54B, 54C, and 54D is disposed between two transmission lines 54A.
[0036] The transmission lines 54A and 54B are configured by a metal pad 53A (first pad) as the above-mentioned metal pad 53. The transmission lines 54C and 54D are configured by a metal pad 53B (second pad) as the above-mentioned metal pad 53. The metal pad 53A is an example of a first pad in this embodiment. The metal pad 53B is an example of a second pad in this embodiment. The metal pad 53A and the metal pad 53B are connected to each other via a film resistor 55a. The width (here, the maximum dimension in the X-axis direction) of the metal pad 53A and the width (here, the maximum dimension in the X-axis direction) of the metal pad 53B are equal to each other. In addition, the length (here, the maximum dimension in the Y-axis direction) of the metal pad 53A and the length (here, the maximum dimension in the Y-axis direction) of the metal pad 53B are equal to each other.
[0037] 3 is an enlarged plan view of one matching circuit pattern 52 in FIG. 2. One end of the metal pad 53A in the Y-axis direction is branched into an input end 53a of a transmission line 54A and an input end 53b of a transmission line 54B. The transmission line 54A is connected to a bonding wire W41 or a bonding wire W48 at a connection point P1 (first connection point) at the input end 53a. The transmission line 54B is connected to a bonding wire W42 or a bonding wire W47 at a connection point P2 at the input end 53b. The input end 53a and the input end 53b are connected to each other via a film resistor 55b. The transmission line 54A and the transmission line 54B are coupled to a connection point Q1 at the other end of the metal pad 53A in the Y-axis direction and connected to a bonding wire W5.
[0038] Moreover, one end of the metal pad 53B in the Y-axis direction is branched into an input end 53c of a transmission line 54C and an input end 53d of a transmission line 54D. The transmission line 54C is connected to a bonding wire W43 or a bonding wire W46 at a connection point P3 at the input end 53c. The transmission line 54D is connected to a bonding wire W44 or a bonding wire W45 at a connection point P4 (second connection point) at the input end 53d. The input end 53c and the input end 53d are connected to each other via a film resistor 55c. The transmission line 54C and the transmission line 54D are joined at a connection point Q2 at the other end of the metal pad 53B in the Y-axis direction and are connected to a bonding wire W5.
[0039] The contours of the metal pads 53A and 53B are asymmetrical with respect to a virtual line N1 (second virtual line) that is perpendicular to a virtual line (first virtual line) that connects the connection points P1 and P4 and passes through the midpoint between the connection points P1 and P4. In other words, the matching circuit pattern 52 has an asymmetrical contour with respect to the virtual line N1. Due to such a contour, the length of a portion of the outer edge of the matching circuit pattern 52 on one side of the virtual line N1 (here, the side including the connection point P1) is longer than the length of a portion of the outer edge of the matching circuit pattern 52 on the other side of the virtual line N1 (here, the side including the connection point P4). Specifically, the length of the portion on one side of the virtual line N1 is the length of a portion of the outer edge of the matching circuit pattern 52 including the metal pad 53A, the length starting from a position projected from the connection point P1 in the X-axis direction and ending from a position projected from the connection point Q1 in the Y-axis direction. The length of the portion on the other side of the virtual straight line N1 is the length of the portion of the outer edge of the matching circuit pattern 52 including the metal pad 53B, which has its start point projected in the X-axis direction from the connection point P4 and its end point projected in the Y-axis direction from the connection point Q2.
[0040] The metal pad 53A has a substantially rectangular shape in a plan view and has four corners. The four corners include one chamfered corner C1 (first corner). Of the four corners, three corners other than the corner C1 are not chamfered. The metal pad 53A is connected to the metal pad 53B at one long side 53r. The corner C1 is located at the intersection of the other long side 53s of the metal pad 53A and the short side 53t facing the composite circuit board 30 (see FIG. 1). The transmission line 54A is formed by a portion of the metal pad 53A that does not include the corner C1. The transmission line 54B is formed by a portion of the metal pad 53A that does not include the corner C1.
[0041] The metal pad 53B has a substantially rectangular shape in a plan view and has four corners. The four corners include one chamfered corner C2 (second corner). The corner C2 is chamfered by a chamfering amount L2 that is larger than the chamfering amount L1 of the corner C1. The chamfering amount L2 is, for example, about three times the chamfering amount L1. Of the four corners, the three corners except the corner C2 are not chamfered. The metal pad 53B is connected to the metal pad 53A at one long side 53u. The corner C2 is located at the intersection of the other long side 53v of the metal pad 53B and the short side 53w facing the composite circuit board 30 (see FIG. 1). The transmission line 54C is formed by a portion of the metal pad 53B that does not include the corner C2. The transmission line 54D is formed by a portion of the metal pad 53B that does not include the corner C2. Therefore, the length of the portion of the outer edge of metal pad 53B that constitutes transmission line 54D is shorter than the length of the portion of the outer edge of metal pad 53A that constitutes transmission line 54A.
[0042] Next, the electrical length in the configuration from the amplifying element 11 to the combining circuit 32 will be described. FIG. 4 is a perspective view for explaining the electrical length of each of the bonding wires W41 to W48. As shown in FIG. 4, the bonding wire W41 is mainly affected by the magnetic field coupling from the closest bonding wire W42 and the next closest bonding wire W43. In contrast, the bonding wire W44 is mainly affected by the magnetic field coupling from the closest bonding wires W43 and W45 and the next closest bonding wires W42 and W46. As a result, the mutual inductance component of the bonding wire W41 increases in response to the influence from only one side, whereas the mutual inductance component of the bonding wire W44 increases in response to the influence from both sides. That is, the mutual inductance components of the bonding wires W41 to W48 increase in response to the other bonding wire W4 that is close to them.
[0043] FIG. 5 is a Smith chart showing the impedances of the bonding wires W41 to W48. FIG. 5 shows the S-parameters (S11) of the transmission line terminations corresponding to the impedances of the bonding wires W41 to W48 at frequencies from 10.700 GHz to 12.700 GHz. The impedances of the bonding wires W41 to W48 refer to the effective impedances of the bonding wires W41 to W48. The effective impedance refers to the impedance of one bonding wire W4 that takes into account the influence of other bonding wires W4. In the Smith chart of FIG. 5, the display range is set to a reflection coefficient Γ=0.3, and the impedance 50Ω at the center of the Smith chart is normalized by the characteristic impedance Z0=50Ω of the calculation port. Since the inductance increases in the direction of the arrow in FIG. 5, it can be seen from FIG. 5 that the inductance of the bonding wires W44 and W45 is greater than the inductance of the bonding wires W41 and W48.
[0044] FIG. 6 is a graph showing the inductance of each of the bonding wires W41 to W48. In the graph of FIG. 6, the horizontal axis is the frequency, and the vertical axis is the inductance value. From FIG. 6, it can be seen that the inductance of the bonding wires W44 and W45 is larger than that of the bonding wires W41 and W48. In this way, in a plurality of bonding wires W4 arranged in a row, the inductance is larger toward the center of the row, and the inductance is smaller toward the end of the row. Therefore, the effective electrical length of the bonding wires W41 to W48 is longer for the wires closer to the center of the row, and shorter for the wires closer to the end of the row. Specifically, the effective electrical length of the bonding wire W44 is longer than that of the bonding wire W41. The effective electrical length is the electrical length of a predetermined path through which an electromagnetic wave passes, and means an electrical length that takes into account the influence of mutual inductance from a bonding wire W4 other than the bonding wire W4 included in the path.
[0045] Next, a description will be given of the electrical length of the transmission lines 54A and 54D. Since high-frequency signals propagate through the transmission lines 54A and 54D as electromagnetic waves, the electrical length of the transmission lines 54A and 54D corresponds to the length of the path along which the electromagnetic waves travel in the transmission lines 54A and 54D.
[0046] First, the relationship between the length of the physical path through which the electromagnetic wave passes and the phase will be described with reference to Figs. 7 and 8. Figs. 7 and 8 are diagrams for explaining the relationship between the length of the physical path through which the electromagnetic wave passes and the phase. Fig. 7 shows a path D1 through which the electromagnetic wave E passes, and Fig. 8 shows a path D2 through which the electromagnetic wave E passes. The path D2 is longer than the path D1. In the example of Figs. 7 and 8, the length of the path D2 is about 1.4 times the length of the path D1. This causes a phase difference between the electromagnetic wave E that has passed through the path D1 and the electromagnetic wave E that has passed through the path D2. Here, the phase of the electromagnetic wave E at the end of the path D1 is 0 deg, whereas the phase of the electromagnetic wave E at the end of the path D2 is about 90 deg, resulting in a phase difference of about 90 deg.
[0047] Here, the path through which high frequency waves (microwaves) as electromagnetic waves pass will be specifically described. FIG. 9 is a diagram for explaining the distribution state of current in microwaves. Part (a) of FIG. 9 is a cross-sectional view showing a dielectric substrate 90 and a metal conductor 80 provided on the dielectric substrate 90, and part (b) of FIG. 9 is a graph showing the magnitude of current flowing through the metal conductor 80 in the Y-axis direction. The metal conductor 80 is, for example, a microstrip conductor. In the graph shown in FIG. 9, the horizontal axis represents the position in the X-axis direction with the center of the metal conductor 80 in the X-axis direction set at 0, and the vertical axis represents the current value I.
[0048] Fig. 10 is a simulation diagram showing the state of current concentration in microwaves. In Fig. 10, the darker the color, the more concentrated the current is. From Fig. 10, it can be seen that the current is particularly concentrated at both ends of the metal conductor 80 in the X-axis direction. Therefore, as shown in Fig. 9, it can be seen that in microwaves, when a current flows through the metal conductor 80, the current value I becomes large at both ends of the metal conductor 80 in the X-axis direction.
[0049] FIG. 11 is a simulation diagram showing the distribution state of the electric field strength in microwaves. In FIG. 11, the darker the color, the greater the electric field strength. From FIG. 11, it can be seen that the electric field strength is particularly large at both ends of the metal conductor 80 in the X-axis direction. Therefore, it can be seen that microwaves as electromagnetic waves mainly pass through the outer edge of a pattern such as the above-mentioned metal conductor 80. In other words, the physical path through which the electromagnetic waves pass corresponds to the outer edge of the pattern. Therefore, by changing the length of the outer edge of the pattern, the length of the physical path through which the electromagnetic waves pass changes. The length of the outer edge of the pattern can be changed by changing the shape of the outer edge.
[0050] Fig. 12 is a plan view showing one reference example of a pattern 80A through which an electromagnetic wave passes, and Fig. 13 is a plan view showing another reference example of a pattern 80B through which an electromagnetic wave passes. Both of the patterns 80A and 80B have two input ports 81 and 82 arranged in parallel, and one output port 83. Each of the patterns 80A and 80B has a substantially rectangular shape, and is bifurcated at one end in the longitudinal direction into a portion constituting the input port 81 and a portion constituting the input port 82. The output port 83 is located at the other end in the longitudinal direction of the patterns 80A and 80B. The distance from the input port 81 to the output port 83 is equal to the distance from the input port 82 to the output port 83.
[0051] Here, the outer shape of the pattern 80A is symmetrical with respect to a virtual line N82 that is perpendicular to the virtual line N81 connecting the input ports 81 and 82 and passes through the midpoints of the input ports 81 and 82. In contrast, the outer shape of the pattern 80B is asymmetrical with respect to the virtual line N82. In the examples of Figs. 12 and 13, all four corners of the pattern 80A are right angles, whereas the four corners of the pattern 80B include three right-angled corners and one chamfered corner. Specifically, the outer shape of the pattern 80B has a shape in which the right-angled corners are cut in a part of the portion extending from the input port 81 to the output port 83. Therefore, in the outer edge of the pattern 80B, the length from the input port 81 to the output port 83 is shorter than the length from the input port 82 to the output port 83.
[0052] As described above, microwaves as electromagnetic waves mainly pass through the outer edges of the patterns 80A and 80B. Therefore, the path DA from the input port 81 to the output port 83 is 82 A shortcut is created with respect to the path DB from the path DA to the output port 83. In this way, when a difference in physical length occurs between the paths DA and DB, as described above, a phase difference occurs between the electromagnetic wave that has passed through the path DA and the electromagnetic wave that has passed through the path DB.
[0053] FIG. 14 is a graph showing the phase difference occurring between two electromagnetic waves that have passed through paths DA and DB in patterns 80A and 80B, respectively. In the graph of FIG. 14, the horizontal axis represents frequency and the vertical axis represents phase difference. It can be seen from FIG. 14 that no phase difference occurs in pattern 80A, whereas a phase difference occurs in pattern 80B. Also, in FIG. 14, the absolute value of the phase difference occurring in pattern 80B is frequency Since the higher the frequency, it can be seen that the effect on the phase difference caused by changing the physical lengths of the paths DA and DB is greater as the frequency is higher.
[0054] In light of the above, the electrical lengths of the transmission lines 54A and 54D in this embodiment will be described with reference to Fig. 3 again. As described above, the length of the portion of the outer edge of the metal pad 53B that constitutes the transmission line 54D is shorter than the length of the portion of the outer edge of the metal pad 53A that constitutes the transmission line 54A. Therefore, the electrical length of the transmission line 54D is shorter than the electrical length of the transmission line 54A.
[0055] In this embodiment, the absolute value of the electrical length difference between the transmission lines 54A and 54D is substantially equal to, for example, the absolute value of the electrical length difference between the bonding wires W41 and W44 described above. In this specification, "substantially equal" means that the difference between the two values is small enough to be negligible. As a result, the combined electrical length of the effective electrical length of the bonding wire W41 and the electrical length of the transmission line 54A is substantially equal to the combined electrical length of the effective electrical length of the bonding wire W44 and the electrical length of the transmission line 54D. In other words, the effective electrical length from the drain pad 15A to the wiring pattern 33 is substantially equal to the effective electrical length from the drain pad 15D to the wiring pattern 33. In this embodiment, the effective electrical length difference between the bonding wires W41 and W44 and the electrical length difference between the transmission lines 54A and 54D are offset by each other. As a result, no phase difference occurs between the electromagnetic wave that has passed through the path from the drain pad 15A to the wiring pattern 33 and the electromagnetic wave that has passed through the path from the drain pad 15D to the wiring pattern 33.
[0056] The effects of the high-frequency amplifier 1A described above will be described. In this high-frequency amplifier 1A, each of the multiple transistors 13 is connected to the matching circuit pattern 52 by a bonding wire W4. In this configuration, the effective impedance may vary between the multiple bonding wires W4 depending on the degree of influence from another bonding wire W4. Such variation in effective impedance between the multiple bonding wires W4 appears as variation in effective electrical length between the multiple bonding wires W4. Specifically, the effective electrical length of the bonding wire W44 is longer than the effective electrical length of the bonding wire W41. Here, in this high-frequency amplifier 1A, since the matching circuit pattern 52 has an asymmetrical outer shape, the electrical length of the transmission line 54D is shorter than the electrical length of the transmission line 54A. As a result, in a configuration in which the effective electrical length of the bonding wire W44 is longer than the effective electrical length of the bonding wire W41, the electrical length difference between the bonding wires W41 and W44 can be offset by the electrical length difference between the transmission lines 54A and 54D. Therefore, the variation between the effective electrical length from the transistor 13A (more specifically, the drain pad 15A) to the wiring pattern 33 and the effective electrical length from the transistor 13D (more specifically, the drain pad 15D) to the wiring pattern 33 is reduced. The smaller the variation in electrical length between the multiple bonding wires W4, the smaller the variation in effective impedance between the multiple bonding wires W4. Therefore, this configuration can reduce the phase variation between the transistors 13A and 13D. The same is true between the transistors 13E and 13H.
[0057] In the above-mentioned high-frequency amplifier 1A, the matching circuit pattern 52 has a metal pad 53A including a corner C1 and a metal pad 53B including a corner C2. The transmission line 54A is formed by a portion of the metal pad 53A including the corner C1. The transmission line 54D is formed by a portion of the metal pad 53B including the corner C2. The corner C2 is chamfered by a chamfering amount L2 that is larger than the chamfering amount L1 of the corner C1. According to the knowledge of the present inventors, as described above, a current (electric field) at a high frequency is likely to concentrate on the outer edge of a conductor. Therefore, in the transmission line 54A, a high-frequency signal propagates along the corner C1 of the metal pad 53A, and in the transmission line 54D, a high-frequency signal propagates along the corner C2 of the metal pad 53B. Here, since the corner C2 is chamfered by a chamfering amount L2 larger than the chamfering amount L1 of the corner C1, the path including the corner C2 is shorter than the path including the corner C1, and a configuration can be realized in which the electrical length of the transmission line 54D is shorter than the electrical length of the transmission line 54A. Furthermore, with this configuration, by adjusting the chamfering amounts L1 and L2, it is easy to match the effective electrical length from the transistor 13A to the wiring pattern 33 and the effective electrical length from the transistor 13D to the wiring pattern 33. Therefore, it is possible to more reliably reduce the phase variation between the transistors 13A and 13D. The same is true between the transistors 13E and 13H.
[0058] The high-frequency amplifier 1A described above includes a plurality of transistor groups each including a transistor 13A, a transistor 13D, and transistors 13E, 13F, 13G, and 13H. This configuration makes it possible to realize a high-output high-frequency amplifier 1A.
[0059] In the above-mentioned high-frequency amplifier 1A, a bonding wire W41 is disposed on the outermost side of the transistor group, and a bonding wire W44 is disposed between the bonding wire W41 and the bonding wires W45, W46, W47, and W48. The length of the bonding wire W41 is equal to the length of the bonding wire W44, and the effective electrical length from the drain pad 15A to the wiring pattern 33 is substantially equal to the effective electrical length from the drain pad 15D to the wiring pattern 33. Therefore, the phases between the transistors 13A and 13D are substantially the same. The same is true between the transistors 13E and 13H. This is particularly advantageous in terms of improving the power efficiency of the high-frequency amplifier 1A.
[0060] Next, a high-frequency amplifier 1B according to the second embodiment will be described. As described above, Fig. 1 is a plan view showing the internal configuration of the high-frequency amplifier 1A according to the first embodiment and the high-frequency amplifier 1B according to the second embodiment. The high-frequency amplifier 1B differs from the high-frequency amplifier 1A in that it includes a matching circuit 60 instead of the matching circuit 50, but is otherwise configured similarly to the high-frequency amplifier 1A. In this embodiment, the high-frequency amplifier 1B includes two matching circuits 60, as an example. The differences from the high-frequency amplifier 1A will be described below.
[0061] 15 is a plan view showing a matching circuit 60. The matching circuit 60 differs from the matching circuit 50 in that the matching circuit 60 includes two matching circuit patterns 62 instead of the two matching circuit patterns 52, and is otherwise configured similarly to the matching circuit 50. Each matching circuit pattern 62 has a plurality of metal pads 63. Like the metal pad 53, each metal pad 63 is electrically connected to the corresponding drain pad 15 of the amplifier element section 10 via a bonding wire W4, and is also electrically connected to the corresponding metal pad 33a of the composite circuit board 30 via a bonding wire W5.
[0062] Furthermore, the matching circuit pattern 62 has a plurality of transmission lines 64. In the two matching circuit patterns 62, the number of the plurality of transmission lines 64 is the same as the number of transistors 13 (see FIG. 2). One matching circuit pattern 62 has four transmission lines 64. The four transmission lines 64 include transmission lines 64A, 64B, 64C, and 64D.
[0063] The two matching circuit patterns 62 have configurations that are inverted from each other in the X-axis direction. In one matching circuit pattern 62, the transmission lines 64A, 64B, 64C, and 64D are arranged in this order in the X-axis direction corresponding to the transistors 13A, 13B, 13C, and 13D (see FIG. 2). The transmission line 64A in the one matching circuit pattern 62 is an example of a first transmission line in this embodiment. The transmission line 64D in the one matching circuit pattern 62 is an example of a second transmission line in this embodiment. The transmission lines 64B and 64C in the one matching circuit pattern 62 may be an example of a second transmission line in this embodiment. In addition, in the other matching circuit pattern 62, the transmission lines 64D, 64C, 64B, and 64A are arranged in this order in the X-axis direction corresponding to the transistors 13E, 13F, 13G, and 13H. The transmission lines 64D, 64C, 64B, and 64A in the other matching circuit pattern 62 are an example of a third transmission line in this embodiment. Of the eight transmission lines 64, two transmission lines 64A are disposed at the outermost positions in the X-axis direction. Also, each of the transmission lines 64B, 64C, and 64D is disposed between two transmission lines 64A.
[0064] The transmission lines 64A and 64B are configured by a metal pad 63A as the above-mentioned metal pad 63. The transmission lines 64C and 64D are configured by a metal pad 63B as the metal pad 63. The metal pad 63A is an example of a first pad in this embodiment. The metal pad 63B is an example of a second pad in this embodiment. The metal pads 63A and 63B are connected to each other via a film resistor 65a. The width of the metal pad 63A (here, the maximum dimension in the X-axis direction) is smaller than the width of the metal pad 63B (here, the maximum dimension in the X-axis direction). The length of the metal pad 63A (here, the maximum dimension in the Y-axis direction) and the length of the metal pad 63B (here, the maximum dimension in the Y-axis direction) are equal to each other.
[0065] 16 is an enlarged plan view of one matching circuit pattern 62 in FIG. 15. One end of the metal pad 63A in the Y-axis direction is branched into an input end 63a of a transmission line 64A and an input end 63b of a transmission line 64B. The transmission line 64A is connected to a bonding wire W41 or a bonding wire W48 at a connection point P5 (first connection point) at the input end 63a. The transmission line 64B is connected to a bonding wire W42 or a bonding wire W47 at a connection point P6 at the input end 63b. The input end 63a and the input end 63b are connected to each other via a film resistor 65b. The transmission line 64A and the transmission line 64B are coupled to a connection point Q3 at the other end of the metal pad 63A in the Y-axis direction and connected to a bonding wire W5.
[0066] One end of the metal pad 63B in the Y-axis direction is branched into an input end 63c of a transmission line 64C and an input end 63d of a transmission line 64D. The transmission line 64C is connected to the bonding wire W43 or the bonding wire W46 at a connection point P7 at the input end 63c. 64D is the input terminal 63dThe input terminal 63c and the input terminal 63d are connected to a bonding wire W44 or a bonding wire W45 at a connection point P8 (second connection point) at the metal pad 63B. The input terminal 63c and the input terminal 63d are connected to each other via a film resistor 65c. The transmission line 64C and the transmission line 64D are joined at a connection point Q4 at the other end of the metal pad 63B in the Y-axis direction and are connected to a bonding wire W5.
[0067] The contours of the metal pads 63A and 63B are asymmetrical with respect to a virtual line N2 (second virtual line) that is perpendicular to a virtual line (first virtual line) that connects the connection points P5 and P8 and passes through the midpoint between the connection points P5 and P8. In other words, the matching circuit pattern 62 has an asymmetrical contour with respect to the virtual line N2. Due to such a contour, the length of a portion of the outer edge of the matching circuit pattern 62 on one side of the virtual line N2 (here, the side including the connection point P5) is longer than the length of a portion of the outer edge of the matching circuit pattern 62 on the other side of the virtual line N2 (here, the side including the connection point P8). Specifically, the length of the portion on one side of the virtual line N2 is the length of a portion of the outer edge of the matching circuit pattern 62 including the metal pad 63A, the length starting from a position projected from the connection point P5 in the X-axis direction and ending from a position projected from the connection point Q3 in the Y-axis direction. The length of the portion on the other side of the virtual straight line N2 is the length of the portion of the outer edge of the matching circuit pattern 62 including the metal pad 63B, which has its start point projected in the X-axis direction from the connection point P8 and its end point projected in the Y-axis direction from the connection point Q4.
[0068] The metal pad 63A, like the metal pad 53A, has a substantially rectangular shape in a plan view and has four corners including one corner C1 (first corner). The metal pad 63A is connected to the metal pad 63B at one long side 63r. The corner C1 is 63A The transmission line 64A is located at the intersection of the other long side 63s of the metal pad 63A and the short side 63t facing the composite circuit board 30 (see FIG. 1). The transmission line 64A is formed by a portion of the metal pad 63A that includes the corner C1. The transmission line 64B is formed by a portion of the metal pad 63A that does not include the corner C1.
[0069] The metal pad 63B, like the metal pad 53B, has a substantially rectangular shape in a plan view and has four corners including one corner C2 (second corner). The metal pad 63B is connected to the metal pad 63A at one long side 63u. The corner C2 is located at the intersection of the other long side 63v of the metal pad 63B and a short side 63w facing the composite circuit board 30 (see FIG. 1). The transmission line 64C is formed by a portion of the metal pad 63B that does not include the corner C2. The transmission line 64D is formed by a portion of the metal pad 63B that does not include the corner C2.
[0070] In this embodiment, the length of the portion of the outer edge of the metal pad 63B that constitutes the transmission line 64D is shorter than the length of the portion of the outer edge of the metal pad 63A that constitutes the transmission line 64A. As an example, in the metal pad 63A, the length L3 of the long side 63s is about 485 μm, the length L4 of the short side 63t is about 325 μm, and the length L5 of the chamfered portion at the corner C1 is about 78 μm. The length of the portion of the outer edge of the metal pad 63A that constitutes the transmission line 64A (i.e., the total value of the lengths L3, L4, and L5) is about 888 μm here. In addition, in the metal pad 63B, the length L6 of the long side 63v is about 360 μm, the length L7 of the short side 63w is about 255 μm, and the length L8 of the chamfered portion at the corner C2 is 255 μm. Of the outer edge of the metal pad 63B, the length of the portion that constitutes the transmission line 64D (that is, the total value of the lengths L6, L7, and L8) is about 870 μm here.
[0071] Furthermore, the metal pad 63A has an asymmetrical outer shape with respect to a virtual line N3 (second virtual line) that is perpendicular to the virtual line (first virtual line) connecting the connection points P5 and P6 and passes through the midpoint between the connection points P5 and P6. The metal pad 63B has an asymmetrical outer shape with respect to a virtual line N4 (second virtual line) that is perpendicular to the virtual line (first virtual line) connecting the connection points P7 and P8 and passes through the midpoint between the connection points P7 and P8. Specifically, the width L11 (here, the dimension in the X-axis direction) of the input end 63a and the width L12 (here, the dimension in the X-axis direction) of the input end 63b are different from each other, and the width L13 (here, the dimension in the X-axis direction) of the input end 63c and the width L14 (here, the dimension in the X-axis direction) of the input end 63d are different from each other. The width L11 is smaller than the width L12. That is, the transmission line 64A is thinner than the transmission line 64B. Furthermore, the width L13 is smaller than the width L14. That is, the transmission line 64C is narrower than the transmission line 64D.
[0072] In this embodiment, the transmission lines 64A and 64B are narrower than the transmission lines 64C and 64D. The widths L11, L12, L13, and L14 are different from one another, in this order. Large In other words, among the multiple input terminals arranged in the X-axis direction, the input terminal closer to the center of the dielectric substrate 51 in the X-axis direction has a larger width (here, the dimension in the X-axis direction). For example, the width L11 is about 0.6 times the width L14, the width L12 is about 0.85 times the width L14, and the width L13 is about 0.9 times the width L14. As an example, the width L11 is about 120 μm, the width L12 is about 170 μm, the width L13 is about 180 μm, and the width L14 is about 200 μm. In this example, considering the thickness (specifically, 200 μm) and the relative dielectric constant (specifically, εr=150) of the dielectric substrate 51 described above, when the frequency is 11.7 GHz, the effective wavelength in the transmission line 64A is 123.4 deg, and the effective wavelength in the transmission line 64D is 120.9 deg.
[0073] The electrical lengths of the transmission lines 64A and 64D will be described. As described above, the length of the portion of the outer edge of the metal pad 63B that constitutes the transmission line 64D is shorter than the length of the portion of the outer edge of the metal pad 63A that constitutes the transmission line 64A. Furthermore, in this embodiment, since the transmission line 64A is thinner than the transmission line 64D, the impedance of the transmission line 64A is larger than the impedance of the transmission line 64D. Therefore, the electrical length of the transmission line 64D is shorter than the electrical length of the transmission line 64A.
[0074] In this embodiment, the absolute value of the electrical length difference between the transmission lines 64A and 64D is equal to, for example, the absolute value of the effective electrical length difference between the above-mentioned bonding wires W41 and W44. As a result, the combined electrical length of the effective electrical length of the bonding wire W41 and the electrical length of the transmission line 64A is substantially equal to the combined electrical length of the effective electrical length of the bonding wire W44 and the electrical length of the transmission line 64D. In other words, the effective electrical length from the drain pad 15A to the wiring pattern 33 is substantially equal to the effective electrical length from the drain pad 15D to the wiring pattern 33. In this embodiment, the effective electrical length difference occurring between the bonding wires W41 and W44 and the electrical length difference occurring between the transmission lines 64A and 64D are offset by each other. As a result, no phase difference occurs between the electromagnetic wave that has passed through the path from the drain pad 15A to the wiring pattern 33 and the electromagnetic wave that has passed through the path from the drain pad 15D to the wiring pattern 33.
[0075] The effects of the high-frequency amplifier 1B described above will now be described. First, a comparative example will be described. The high-frequency amplifier according to the comparative example differs from the high-frequency amplifier 1B in that it includes a matching circuit 60X according to the comparative example instead of the matching circuit 60. FIG. 19 is a plan view showing the matching circuit 60X of the high-frequency amplifier according to the comparative example. The matching circuit 60X differs from the matching circuit 60 in that it includes two matching circuit patterns 62X instead of the two matching circuit patterns 62, but is otherwise configured similarly to the matching circuit 60.
[0076] Each matching circuit pattern 62X has two metal pads 63X instead of the metal pads 63A and 63B. The metal pad 63X is different from the metal pads 63A and 63B in that it does not have the chamfered corners C1 and C2. All four corners of the metal pad 63X are configured at right angles. The two metal pads 63X in the matching circuit pattern 62X are different from the metal pads 63A and 63B in that it has four input ends 63y instead of the input ends 63a, 63b, 63c, and 63d. The widths (here, dimensions in the X-axis direction) of the four input ends 63y are equal to each other. In other respects, the metal pad 63X is configured similarly to the metal pads 63A and 63B. That is, the outer shape of the matching circuit pattern 62X according to the comparative example is line-symmetrical with respect to the virtual straight line N2.
[0077] Fig. 20 is a Smith chart showing the load impedances of the transistors 13A, 13B, 13C, and 13D of the high-frequency amplifier according to the comparative example. In Fig. 20, the load impedances of the transistors 13A, 13B, 13C, and 13D are shown in the frequency range from 10.700 GHz to 12.700 GHz. From Fig. 20, it can be seen that the range of phase variation occurring between the transistors 13A, 13B, 13C, and 13D (here, the phase difference ΔZ1 occurring between the transistors 13A and 13D) becomes large. In the example of Fig. 20, the phase difference ΔZ1 is about 9 deg.
[0078] It is considered that the load impedances of transistors 13H, 13G, 13F, and 13E are approximately the same as the load impedances of transistors 13A, 13B, 13C, and 13D. From the above, it can be seen that in the high-frequency amplifier according to the comparative example, a large phase variation (a variation in the range of about 9 degrees in the example of FIG. 20) occurs between transistor 13A and transistor 13H.
[0079] FIG. 21 is a diagram showing the relationship between the variation in the phase of the load impedance and the power efficiency in the comparative example. In the contour diagram of FIG. 21, the darker the color, the higher the power efficiency. Also, the Smith chart of FIG. 21 shows a peak point Z10 of the phase of the load impedance at which the maximum power efficiency (here, ηd=52.9%) is obtained at a frequency of 11.700 GHz. Here, the peak point Z10 is 99 deg. From FIG. 21, it can be seen that at a point Z11 (here, 108 deg), which is shifted in phase by about 9 deg from the peak point Z10, the power efficiency is lower than that of the peak point Z10 (here, ηd=about 50.0%). Therefore, it can be seen that when the matching circuit 60X is used, the power efficiency is about 3% lower than that of the peak point Z10.
[0080] In contrast, the high-frequency amplifier 1B according to this embodiment has a similar configuration to the high-frequency amplifier 1A described above, and thus has the same effects as the high-frequency amplifier 1A. Furthermore, in the high-frequency amplifier 1B, the transmission line 64A is thinner than the transmission line 64D. This allows the impedance per unit length of the transmission line 64A to be reduced by 1 / 2 the transmission line 64D. 64D Since the impedance per unit length of the transmission line 64D can be made large, a configuration can be realized in which the electrical length of the transmission line 64D is shorter than the electrical length of the transmission line 64A. In addition, by thinning the transmission line 64A, it is easy to match the effective electrical length from the transistor 13A to the wiring pattern 33 with the effective electrical length from the transistor 13D to the wiring pattern 33. Therefore, it is possible to more reliably reduce the phase variation between the transistor 13A and the transistor 13D. For the same reason, it is possible to more reliably reduce the phase variation between the transistor 13A and the transistor 13H.
[0081] Fig. 17 is a Smith chart showing the load impedances of the transistors 13A, 13B, 13C, and 13D of the high-frequency amplifier 1B according to the second embodiment. In Fig. 17, the load impedances of the transistors 13A, 13B, 13C, and 13D are shown in the frequency range of 10.700 GHz to 12.700 GHz. From Fig. 17, it can be seen that the range of the phase variation occurring between the transistors 13A, 13B, 13C, and 13D (here, the phase difference ΔZ2 occurring between the transistors 13A and 13D) is small. In the example of Fig. 17, the phase difference ΔZ2 is about 4 deg.
[0082] It is considered that the load impedances of the transistors 13H, 13G, 13F, and 13E are approximately equal to the load impedances of the transistors 13A, 13B, 13C, and 13D. From the above, it can be seen that, according to the high-frequency amplifier 1B, only a small variation in phase occurs between the transistors 13A and 13H (a variation in the range of about 4 degrees in the example of FIG. 17). Therefore, it can be seen that the high-frequency amplifier 1B can reduce the variation in phase occurring between the transistors 13A and 13H more than the high-frequency amplifier according to the comparative example. It can be seen that, in the example of FIG. 17, the range of the variation can be made smaller by about 5 degrees than in the example of FIG. 20.
[0083] FIG. 18 is a diagram showing the relationship between the variation in the phase of the load impedance and the power efficiency in the second embodiment. In the contour diagram of FIG. 18, the darker the color, the higher the power efficiency. Also, FIG. 18 shows a peak point Z10, as in FIG. 21. From FIG. 18, it can be seen that at point Z12 (here, 103 deg), which is shifted in phase by about 4 deg from peak point Z10, the power efficiency is almost the same as that of peak point Z10 (here, ηd=about 52.1%). Therefore, it can be seen that by using the matching circuit 60, it is possible to suppress the decrease in power efficiency relative to peak point Z10 to less than 1%.
[0084] The above embodiment has been described as an embodiment of the high frequency amplifier according to the present disclosure. The high frequency amplifier according to the present disclosure may be any of the above-described embodiments modified in any way.
[0085] For example, the high-frequency amplifier 1A according to the above embodiment includes two matching circuits 40, 50, and the amplifying element section 10 includes two amplifying elements 11, but this configuration is not limited to this. The high-frequency amplifier 1A may include one matching circuit 40, 50, or may include three or more matching circuits 40, 50. The amplifying element section 10 may include a single amplifying element 11, or may include three or more amplifying elements 11. The same applies to the high-frequency amplifier 1B. Moreover, the high-frequency amplifiers 1A, 1B may each include a matching circuit 50 or a matching circuit 60 instead of the matching circuit 40.
[0086] Furthermore, the number of transistors 13 and the number of bonding wires W4 are arbitrary and may be less than 8 or may be 9 or more. The numbers of matching circuit patterns 52, 62 and the numbers of transmission lines 54, 64 may also be changed arbitrarily in response to the numbers of transistors 13 and the numbers of bonding wires W4.
[0087] In the above embodiment, the matching circuit pattern 52 has two metal pads 53, but is not limited to this configuration. The matching circuit pattern 52 may have only one metal pad 53, or may have three or more metal pads 53. The matching circuit pattern 52 may have both metal pads 53 and 63. The same applies to the matching circuit pattern 62. [Explanation of symbols]
[0088] 1A, 1B...High frequency amplifier 2...Input terminal 3...Output terminal 4. Package 4a, 4b...End walls 4c,4d…side wall 4e…Bottom plate 10...Amplification element section 11...Amplification element 12...Semiconductor substrate 13, 13B, 13C...Transistor 13A…Transistor (first transistor) 13D...Transistor (second transistor) 13E, 13F, 13G, 13H ...Transistor (third transistor) 14…Gate pad 15, 15B, 15C...Drain pad 15A…Drain pad (first drain pad) 15D…Drain pad (second drain pad) 15E, 15F, 15G, 15H...Drain pad (third drain pad) 16...Sauce pad 20...Branch circuit board 21...Substrate 21a, 21b...Long side 21c, 21d…Short side 22...Branch circuit 23...Wiring pattern 23a…Metal pad 23b…Membrane resistance 30...Composite circuit board 31...Substrate 31a, 31b...Long side 31c, 31d…short side 32…Synthesis circuit 33...Wiring pattern 33a…Metal pad 40…Matching circuit 50…Matching circuit 51...Dielectric substrate 52…Matching circuit pattern 53...Metal pad 53A…Metal pad (first pad) 53B…Metal pad (second pad) 53a, 53b, 53c, 53d...input terminal 53r, 53s, 53u, 53v...long side 53t, 53w...short side 54, 54B, 54C...Transmission lines 54A…Transmission line (first transmission line) 54D…Transmission line (second transmission line) 55a,55b,55c...Membrane resistance 60,60X…Matching circuit 62,62X…Matching circuit pattern 63A…Metal pad (first pad) 63B…Metal pad (second pad) 63X…Metal pad 63a, 63b, 63c, 63d, 63y...input terminal 63r, 63s, 63u, 63v...long side 63t ,63w…short side 64, 64B, 64C...Transmission lines 64A…Transmission line (first transmission line) 64D…Transmission line (second transmission line) 65a,65b,65c…Membrane resistance 80...Metal conductor 80A, 80B…Pattern 81,82...Input port 83...Output port 90...Dielectric substrate C1...Corner (first corner) C2: Corner (second corner) E...electromagnetic wave D1, D2…Route I…Current value L1, L2... Chamfering amount L11, L12, L13, L14…Width N1, N2, N3, N4...Virtual line (second virtual line) N81, N82: Virtual straight line P1, P5...Connection point (first connection point) P4, P8…Connection points (second connection points) P2, P3, P6, P7...Connection points Q1, Q2, Q3, Q4...Connection points W1, W2, W3, W5, W6... Bonding wire W4, W42, W43... Bonding wire (wire) W41…Bonding wire (first wire) W44…Bonding wire (second wire) W45, W46, W47, W48... Bonding wire (third wire) Z10…Peak point Z11,Z12…points ΔZ1, ΔZ2…Phase difference
Claims
1. a first transistor, a second transistor aligned with the first transistor in a first direction, and a third transistor aligned with the second transistor in the first direction on an opposite side to the first transistor; a first drain pad electrically connected to a drain electrode of the first transistor, a second drain pad electrically connected to a drain electrode of the second transistor, and a third drain pad electrically connected to a drain electrode of the third transistor; a matching circuit pattern including a first transmission line electrically connected to the first drain pad, a second transmission line electrically connected to the second drain pad, and a third transmission line electrically connected to the third drain pad, for performing impedance matching of a high frequency signal with respect to each of the first transistor, the second transistor, and the third transistor; a first wire electrically connecting the first transmission line and the first drain pad, a second wire electrically connecting the second transmission line and the second drain pad, and a third wire electrically connecting the third transmission line and the third drain pad; a wiring pattern electrically connected to the first drain pad via the first transmission line and the first wire, and electrically connected to the second drain pad via the second transmission line and the second wire; Equipped with an effective impedance of the second wire is greater than an effective impedance of the first wire; the matching circuit pattern has an asymmetrical outer shape with respect to a second virtual line that is perpendicular to a first virtual line connecting a first connection point to which the first wire of the first transmission line is connected and a second connection point to which the second wire of the second transmission line is connected, and that passes through a midpoint between the first connection point and the second connection point; The electrical length of the second transmission line is shorter than the electrical length of the first transmission line. High frequency amplifier.
2. the matching circuit pattern has a first pad including a first corner portion and a second pad including a second corner portion; the first transmission line is formed by a portion of the first pad including the first corner portion, the second transmission line is configured by a portion of the second pad including the second corner portion, The second corner portion is chamfered by a chamfering amount larger than the chamfering amount of the first corner portion.
2. The high frequency amplifier according to claim 1.
3. The first transmission line is thinner than the second transmission line.
3. The high frequency amplifier according to claim 1 or 2.
4. a plurality of transistor groups each including the first transistor, the second transistor, and the third transistor are provided; The high-frequency amplifier according to any one of claims 1 to 3.
5. the first wire is disposed on an outermost portion of the group of transistors, and the second wire is disposed between the third wire and the first wire; the length of the first wire is equal to the length of the second wire; an effective electrical length from the first drain pad to the wiring pattern is substantially equal to an effective electrical length from the second drain pad to the wiring pattern; 5. The high frequency amplifier according to claim 4.
Citation Information
Patent Citations
Internal matching type high output field effect transistor
JP1988086904A
High-frequency semiconductor device
JP2010161348A
Power amplifier
JP2014222836A
High frequency circuit
JP2018078487A
High-frequency amplifier
JP2020036153A
