Multimode filter, filter, and multiplexer

Abstract

To provide a multimode filter that can widen a passband width and reduce loss in the passband.SOLUTION: A multimode filter includes a support substrate, a piezoelectric layer having a thickness of 3 μm or less provided on the support substrate, one or more input IDTs provided on the piezoelectric layer, each having a plurality of electrode fingers formed by stacking a first layer and a second layer having a density greater than that of the first layer, and one or more output IDTs provided alternately with the one or more input IDTs on the piezoelectric layer, each having a plurality of electrode fingers formed by stacking the first layer and the second layer, and has three or more resonance points in the frequency characteristics of the real part of admittance within the passband width range at an attenuation of -10 dB in the pass characteristics.SELECTED DRAWING: Figure 13

Description

[Technical Field]

[0001] The present invention relates to multimode filters, filters and multiplexers. [Background technology]

[0002] In high-frequency communication systems, such as mobile phones, high-frequency filters are used to remove unnecessary signals outside the frequency band used for communication. A known high-frequency filter is a multimode filter in which input IDTs (Interdigital Transducers) and output IDTs, each having a plurality of electrode fingers, are alternately arranged in the propagation direction of surface acoustic waves (SAWs) (see, for example, Patent Document 1).

[0003] It is also known that in an elastic wave resonator having a piezoelectric film provided on a support substrate, the Q value can be increased by making the thickness of the piezoelectric film 1λ or less (for example, Patent Document 2). It is also known that insertion loss can be reduced by making the electrode fingers a laminate film of a platinum film, a gold film, a silver film, or a copper film and an aluminum film thereon (for example, Patent Document 3). It is also known that power durability can be improved by making the electrode fingers a laminate film of an aluminum film and a tungsten film or tantalum film thereon (for example, Patent Document 4).

[0004] It is also known that low loss can be achieved by making the sound velocity of the surface acoustic waves excited by the electrode fingers slower than the sound velocity of the bulk waves propagating within the piezoelectric substrate (e.g., Patent Document 5). It is also known that the sound velocity of the surface acoustic waves excited by the electrode fingers can be slowed by making the electrode fingers a laminated film of a platinum film and an aluminum film thereon (e.g., Non-Patent Document 1). It is also known that miniaturization can be achieved by keeping the normalized film thickness of the electrode fingers, which is defined by the density and thickness of each layer of the electrode fingers and the density of molybdenum, within a predetermined range (e.g., Patent Document 6). It is also known that the passband can be widened by modulating the pitch of the IDTs that form a multimode filter (e.g., Non-Patent Documents 2 and 3). [Prior art documents]

Charter Documents

[0005] [Patent Document 1] Japanese Patent Publication No. 2007-181195 [Patent Document 2] Japanese Patent Publication No. 2015-73331 [Patent Document 3] International Publication No. 2011 / 049060 [Patent Document 4] Japanese Patent Publication No. 2011-211460 [Patent Document 5] Japanese Patent Publication No. 2016-136712 [Patent Document 6] Japanese Patent Publication No. 2019-129508

Non-licensed literature

[0006]

Non-patent document 1

Non-patent document 2

Non-patent document 3

[0007] The multimode filter can be made smaller by reducing the pitch of the electrode fingers of the IDT that forms the multimode filter. The electrode finger pitch decreases as the acoustic velocity of the acoustic waves decreases. Configuring the electrode fingers to include a denser layer reduces the acoustic velocity of the acoustic waves excited by the electrode fingers. Therefore, it is preferable for the electrode fingers to be configured with a first layer stacked on a second layer that is denser than the first layer. However, in a configuration in which a thin piezoelectric layer is provided on a support substrate, using electrode fingers that are stacked with a first layer stacked on a second layer that is denser than the first layer can narrow the passband bandwidth and increase loss in the passband.

[0008] The present invention has been made in view of the above-mentioned problems, and has as its object to widen the passband width and reduce loss in the passband. [Means for solving the problem]

[0009] The present invention is a multi-mode filter comprising: a support substrate; a piezoelectric layer provided on the support substrate and having a thickness of 3 μm or less; one or more input IDTs provided on the piezoelectric layer and having a plurality of electrode fingers formed by stacking a first layer and a second layer having a density greater than that of the first layer; and one or more output IDTs provided alternately with the one or more input IDTs on the piezoelectric layer and having the plurality of electrode fingers formed by stacking the first layer and the second layer, wherein the pass characteristics have three or more resonance points in the frequency characteristics of the real part of admittance within the pass bandwidth range at an attenuation of -10 dB.

[0010] In the above configuration, the one or more input IDTs and the one or more output IDTs may be configured such that a first pitch of the electrode fingers in an end region located at an end of a central region located at the center in the direction in which the electrode fingers are arranged is smaller than a second pitch of the electrode fingers in the central region.

[0011] In the above configuration, the first pitch in the edge region may become smaller from the central region side toward the opposite side to the central region.

[0012] In the above configuration, the first pitch may be configured to decrease with each pitch.

[0013] In the above configuration, the value of the real part of the admittance of the three or more resonance points is 0.5×10 -1 s or more.

[0014] In the above configuration, the thickness of the second layer may be 20% or more of the thickness of the plurality of electrode fingers.

[0015] In the above configuration, the plurality of electrode fingers may be formed by stacking a plurality of layers including the first layer and the second layer, and the density of each of the plurality of layers is ρi, the film thickness of each layer is hi, the density of molybdenum is ρo, and λ is twice the average pitch of the plurality of electrode fingers in a central region located at the center in the direction in which the plurality of electrode fingers are arranged in the one or more input IDTs and the one or more output IDTs, so that the value of the following formula is 0.08λ or more and 0.24λ or less: TIFF2025145253000002.tif37170

[0016] In the above structure, the first layer may be a layer containing aluminum as a main component, and the second layer may be a layer containing molybdenum, tungsten, platinum, ruthenium, or tantalum as a main component.

[0017] The present invention is a filter including the multimode filter described above.

[0018] The present invention is a multiplexer including the multimode filter described above. [Effects of the Invention]

[0019] According to the present invention, it is possible to widen the passband and reduce the loss in the passband. [Brief explanation of the drawings]

[0020] [Figure 1] FIG. 1(a) is a plan view of a multimode filter according to a first embodiment, and FIG. 1(b) is a cross-sectional view taken along the line AA in FIG. 1(a). [Figure 2] FIG. 2(a) is a plan view of the input IDT and the output IDT in the first embodiment, and FIG. 2(b) is a cross-sectional view of the electrode fingers. [Figure 3] 3(a) is a plan view of the input IDT and output IDT in models 1 and 2 used in simulation 1, FIG. 3(b) is a cross-sectional view of the electrode fingers of model 1, and FIG. 3(c) is a cross-sectional view of the electrode fingers of model 2. [Figure 4]FIG. 4 is a diagram showing the pitch of electrode fingers relative to the number of pairs of electrode fingers in models 1 and 2. In FIG. [Figure 5] 5(a) and 5(b) are diagrams showing simulation results of the attenuation and the real part of the admittance Real(Y) versus frequency in Models 1 and 2. FIG. [Figure 6] 6(a) and 6(b) are diagrams showing simulation results of the symmetric mode Ye and the antisymmetric mode Yo with respect to frequency in Models 1 and 2. FIG. [Figure 7] 7(a) to 7(c) are diagrams (part 1) showing the simulation results of the attenuation and the real part of the admittance Real(Y) versus frequency when the number of pairs of electrode fingers in the end region is increased compared to Model 1. [Figure 8] 8(a) to 8(c) are graphs (part 2) showing the simulation results of the attenuation and the real part of the admittance Real(Y) versus frequency when the number of pairs of electrode fingers in the end region is increased compared to Model 1. [Figure 9] 9(a) to 9(c) are diagrams (part 1) showing simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency when the number of pairs of electrode fingers in the edge region is increased compared to Model 1. [Figure 10] 10(a) to 10(c) are diagrams (part 2) showing simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency when the number of pairs of electrode fingers in the end region is increased compared to model 1. [Figure 11] FIG. 11 is a diagram showing the pitch of electrode fingers relative to the number of pairs of electrode fingers in model 3. In FIG. [Figure 12] FIG. 12 is a diagram showing the pitch of electrode fingers relative to the number of pairs of electrode fingers in model 4. In FIG. [Figure 13] FIG. 13(a) is a diagram showing the simulation results of the attenuation and the real part of the admittance Real(Y) versus frequency in Model 3, and FIG. 13(b) is a diagram showing the simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency. [Figure 14] FIG. 14(a) is a diagram showing the simulation results of the attenuation and the real part of the admittance Real(Y) versus frequency in Model 4, and FIG. 14(b) is a diagram showing the simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency. [Figure 15] FIG. 15(a) is a diagram showing the simulation results of the attenuation and the real part of the admittance Real(Y) versus frequency in Model 5, and FIG. 15(b) is a diagram showing the simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency. [Figure 16] FIG. 16(a) is a diagram showing the simulation results of the attenuation and the real part of the admittance Real(Y) versus frequency in Model 6, and FIG. 16(b) is a diagram showing the simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency. [Figure 17] FIG. 17(a) is a circuit diagram of a filter according to the second embodiment, and FIG. 17(b) is a plan view of a filter according to a modified example of the second embodiment. [Figure 18] FIG. 18 is a circuit diagram of a duplexer according to the third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0021] Hereinafter, embodiments of the present invention will be described with reference to the drawings. [Example]

[0022] FIG. 1(a) is a plan view of a multimode filter 100 according to a first embodiment, and FIG. 1(b) is a cross-sectional view taken along line AA of FIG. 1(a). Note that FIG. 1(a) shows a simplified illustration of an input IDT 21 and an output IDT 22. The arrangement direction of the electrode fingers 18 is defined as the X direction, the extension direction of the electrode fingers 18 as the Y direction, and the stacking direction of the support substrate 10 and the piezoelectric layer 14 as the Z direction. The X direction and the Y direction do not necessarily correspond to the X-axis direction and the Y-axis direction of the crystal orientation of the piezoelectric layer 14. When the piezoelectric layer 14 is a rotated Y-cut X-propagation substrate, the X direction corresponds to the X-axis direction of the crystal orientation.

[0023] 1(a) and 1(b), a piezoelectric layer 14 is provided on a support substrate 10. A first insulating layer 11 is provided between the support substrate 10 and the piezoelectric layer 14. A second insulating layer 12 is provided between the first insulating layer 11 and the piezoelectric layer 14. The thickness of the first insulating layer 11 is T1, the thickness of the second insulating layer 12 is T2, and the thickness of the piezoelectric layer 14 is T4.

[0024] One or more input IDTs 21, one or more output IDTs 22, and a pair of reflectors 23 are provided on the piezoelectric layer 14. The input IDT 21 is connected to the input terminal Tin. The output IDT 22 is connected to the output terminal Tout. The input IDTs 21 and output IDTs 22 are arranged alternately in the X direction. The reflectors 23 are arranged to sandwich the input IDTs 21 and output IDTs 22 in the X direction. The input IDTs 21, output IDTs 22, and reflectors 23 are formed by a metal film 16 on the piezoelectric layer 14. The numbers of input IDTs 21 and output IDTs 22 are set appropriately, but the total number of input IDTs 21 and output IDTs 22 is an odd number greater than or equal to three. In other words, if the number of input IDTs 21 is odd, for example, the number of output IDTs 22 is even. If the number of input IDTs 21 is even, for example, the number of output IDTs 22 is odd.

[0025] FIG. 2(a) is a plan view of the input IDT 21 and the output IDT 22 in the first embodiment, and FIG. 2(b) is a cross-sectional view of the electrode fingers 18. As shown in FIG. 2(a), the input IDT 21 and the output IDT 22 each include a pair of opposing comb electrodes 20. The comb electrode 20 includes a plurality of electrode fingers 18 and a bus bar 19 to which the plurality of electrode fingers 18 are connected. The region where the electrode fingers 18 of the pair of comb electrodes 20 intersect is an intersection region 25. The length of the intersection region 25 in the Y direction is the aperture length. The pair of comb electrodes 20 are arranged such that the electrode fingers 18 alternate with each other in at least a portion of the intersection region 25. The acoustic waves excited primarily by the plurality of electrode fingers 18 propagate primarily in the X direction. Note that the comb electrode 20 may include dummy electrode fingers.

[0026] Each of the input IDT 21 and the output IDT 22 has a central region 30 located at the center in the X direction and an edge region 32 located at the end of the central region 30 in the X direction. The pitch D2 (the pitch between the centers of the electrode fingers 18) of the electrode fingers 18 located in the edge region 32 is smaller than the pitch D1 of the electrode fingers 18 located in the central region 30. The width W2 and pitch D2 of the electrode fingers 18 located in the edge region 32 decrease toward the adjacent input IDT 21 or output IDT 22 (in the direction of the arrow). The width W1 and pitch D1 of the electrode fingers 18 located in the central region 30 are, for example, the same as each other. The width W2 and pitch D2 of the electrode fingers 18 in the edge region 32 of the input IDT 21 are the same as the width W2 and pitch D2 of the corresponding electrode fingers 18 in the edge region 32 of the output IDT 22.

[0027] The pitch of the electrode fingers 18 in the central region 30 of one of the pair of comb electrodes 20 is approximately equal to the wavelength λ of the acoustic wave. The pitch of the electrode fingers 18 of one of the comb electrodes 20 in the central region 30 is the pitch D1 of two of the electrode fingers 18 in the central region 30. The average pitch D1 can be calculated by dividing the width of the central region 30 in the X direction by the number of electrode fingers 18 in the central region 30. The average pitch D1 of the multiple input IDTs 21 and the multiple output IDTs 22 may be the average of the respective average pitches D1, or may be the average pitch D1 of one of the IDTs. An insulating film may be provided to cover the electrode fingers 18. The insulating film may function as a protective film or a temperature compensation film.

[0028] As shown in FIG. 2( b), the electrode fingers 18 of each of the input IDT 21 and the output IDT 22 are formed by a metal film 16 including a first layer 41, a second layer 42 provided between the first layer 41 and the piezoelectric layer 14, and a third layer 43 provided between the second layer 42 and the piezoelectric layer 14. The first layer 41 is the thickest layer among the first layer 41, the second layer 42, and the third layer 43. The first layer 41 is primarily composed of, for example, aluminum (Al). The second layer 42 is thinner than the first layer 41 and has a thickness of 20% or more of the thickness of the metal film 16. The second layer 42 is primarily composed of a material with a higher density than the first layer 41. The second layer 42 is primarily composed of, for example, molybdenum (Mo), tungsten (W), platinum (Pt), ruthenium (Ru), or tantalum (Ta). The third layer 43 is a layer for adhesion and is a layer mainly composed of, for example, titanium (Ti). The third layer 43 is a layer thinner than the first layer 41 and the second layer 42. "Mainly composed of a certain element" means that the certain element is contained at 50 atomic % or more (or, for example, 80 atomic % or more). Because the electrode finger 18 includes the first layer 41 and the second layer 42, which has a higher density than the first layer 41, the sound velocity of the surface acoustic wave excited by the electrode finger 18 is slow, for example, to 3200 m / s or less.

[0029] The piezoelectric layer 14 is, for example, a single-crystal lithium tantalate (LiTaO3) layer or a single-crystal lithium niobate (LiNbO3) layer, such as a rotated Y-cut X-propagation lithium tantalate layer or a rotated Y-cut X-propagation lithium niobate layer. The piezoelectric layer 14 may also be, for example, a rotated Y-cut X-propagation lithium tantalate layer with a cut angle of 5° to 42°. The thickness T4 of the piezoelectric layer 14 is preferably 1.0λ or less, more preferably 0.5λ or less, from the viewpoint of improving the Q value and suppressing spurious and loss. If the piezoelectric layer 14 is too thin, it becomes difficult to excite an acoustic wave, so the thickness T4 is preferably 0.1λ or more. In other words, the thickness T4 of the piezoelectric layer 14 is preferably 3.0 μm or less, more preferably 1.5 μm or less. The thickness T4 is preferably 0.3 μm or more.

[0030] The support substrate 10 is, for example, a sapphire substrate, alumina substrate, silicon substrate, spinel substrate, quartz substrate, quartz substrate, or silicon carbide substrate. The sapphire substrate is a single-crystal Al2O3 substrate, the alumina substrate is a polycrystalline or amorphous Al2O3 substrate, and the silicon substrate is a single-crystal or polycrystalline silicon substrate. The spinel substrate is a polycrystalline or amorphous MgAl2O4 substrate, the quartz substrate is a single-crystal SiO2 substrate, the quartz substrate is a polycrystalline or amorphous SiO2 substrate, and the silicon carbide substrate is a polycrystalline or single-crystal SiC substrate. The linear expansion coefficient of the support substrate 10 in the X direction is smaller than the linear expansion coefficient of the piezoelectric layer 14 in the X direction. This reduces the frequency temperature dependence of the multimode filter 100.

[0031] The acoustic velocity of the bulk wave propagating through the first insulating layer 11 is faster than the acoustic velocity of the bulk wave propagating through the second insulating layer 12 and the piezoelectric layer 14. This confines the energy of the main response elastic wave within the piezoelectric layer 14 and the second insulating layer 12. The first insulating layer 11 is, for example, polycrystalline or amorphous, and is an aluminum oxide layer, silicon nitride layer, aluminum nitride layer, silicon carbide layer, or polysilicon layer. From the viewpoint of confining the elastic wave within the piezoelectric layer 14 and the second insulating layer 12, the thickness T1 of the first insulating layer 11 is preferably 0.5λ or more, more preferably 1.0λ or more. From the viewpoint of improving characteristics, the thickness T1 is preferably 10.0λ or less. In other words, the thickness T1 of the first insulating layer 11 is preferably 1.5 μm or more, more preferably 3.0 μm or more. The thickness T1 is preferably 30 μm or less.

[0032] The second insulating layer 12 is, for example, a temperature compensation film, and has a temperature coefficient of elastic constant with the opposite sign to that of the piezoelectric layer 14. For example, the temperature coefficient of elastic constant of the piezoelectric layer 14 is negative, and the temperature coefficient of elastic constant of the second insulating layer 12 is positive. The second insulating layer 12 is, for example, a silicon oxide (SiO2) layer that is undoped or contains an additive element such as fluorine, phosphorus, or boron, and is, for example, polycrystalline or amorphous. This reduces the frequency temperature dependence of the multimode filter 100. When the second insulating layer 12 is a silicon oxide layer, the sound velocity of bulk waves propagating through the second insulating layer 12 is slower than that of bulk waves propagating through the piezoelectric layer 14.

[0033] The sound velocity of the bulk wave in each layer is the sound velocity of the shear wave V S and is expressed by equation 1, where G is the modulus of rigidity and ρ is the density.

number

number

[0034] For the second insulating layer 12 to have a temperature compensation function, it is necessary that a certain amount of energy of the main response acoustic wave be present within the second insulating layer 12. The range in which the energy of the surface acoustic wave is concentrated depends on the type of surface acoustic wave, but it is typically concentrated within a range of 2.0λ from the top surface of the piezoelectric layer 14, and particularly within a range of 1.0λ from the top surface of the piezoelectric layer 14. Therefore, the distance from the bottom surface of the second insulating layer 12 to the top surface of the piezoelectric layer 14 (thickness T2+T4) is preferably 2.0λ or less, and more preferably 1.0λ or less. In other words, thickness T2+T4 is preferably 6.0 μm or less, and more preferably 3.0 μm or less.

[0035] The input IDT 21, output IDT 22, and reflector 23 are arranged in the X direction, which is the propagation direction of the acoustic waves excited by the input IDT 21 and output IDT 22. The acoustic waves excited by the input IDT 21 and output IDT 22 are reflected by the reflector 23. As a result, the energy of the acoustic waves is confined within the input IDT 21 and output IDT 22. A bandpass filter is formed by utilizing a resonance mode generated by acoustic coupling between the input IDT 21 and output IDT 22.

[0036] [Simulation 1] FIG. 3(a) is a plan view of the input IDT 21 and the output IDT 22 in Models 1 and 2 used in Simulation 1. FIG. 3(b) is a cross-sectional view of the electrode fingers 18 in Model 1. FIG. 3(c) is a cross-sectional view of the electrode fingers 18 in Model 2. As shown in FIG. 3(a), Models 1 and 2 used in Simulation 1 have the same pitch D2 in the edge region 32 in the input IDT 21 and the output IDT 22. FIG. 3(a) illustrates a case where the pitch D2 of two pairs of electrode fingers 18 is the same in the edge region 32. As shown in FIG. 3(b), Model 1 has the same electrode fingers 18 as Example 1, with the electrode fingers 18 being a laminated film including a first layer 41, a second layer 42, and a third layer 43. As shown in FIG. 3(c), Model 2 has the same electrode fingers 18 as Example 1. The other configurations of Models 1 and 2 are the same as Example 1.

[0037] The simulation conditions other than the pitch of the electrode fingers 18 of the input IDT 21 and the output IDT 22 are as follows. Common conditions for Model 1 and Model 2 Support substrate 10: 30 μm thick sapphire substrate First insulating layer 11: Aluminum oxide (Al2O3) layer with a thickness of 6 μm Number of pairs of electrode fingers 18 in the end region 32: 2 pairs Aperture length of comb-shaped electrode 20: 25λ Duty ratio of comb electrode 20: 50% Number of input IDT21: 3 Number of output IDT22: 4 Model 1 conditions Second insulating layer 12: 1.35 μm thick silicon oxide (SiO2) layer Piezoelectric layer 14: 0.78 μm thick 42° rotated Y-cut X-propagation lithium tantalate (LiTaO3) layer First layer 41: Aluminum (Al) layer with a thickness of 0.250 μm Second layer 42: Molybdenum (Mo) layer with a thickness of 0.245 μm Third layer 43: Titanium (Ti) layer with a thickness of 0.110 μm Model 2 conditions Second insulating layer 12: 1.50 μm thick silicon oxide (SiO2) layer Piezoelectric layer 14: 1.00 μm thick 42° rotated Y-cut X-propagation lithium tantalate (LiTaO3) layer First layer 41: Aluminum (Al) layer with a thickness of 0.250 μm Third layer 43: Titanium (Ti) layer with a thickness of 0.110 μm

[0038] FIG. 4 is a diagram showing the pitch of the electrode fingers 18 relative to the number of pairs of the electrode fingers 18 in Models 1 and 2. The number of pairs of the electrode fingers 18 on the horizontal axis is counted starting from 0 at the end of one reflector 23. As shown in FIG. 4, in Model 1, the pitch D1 in the central region 30 of the input IDT 21 and the output IDT 22 was set to 3.9 μm. The pitch D2 in the edge region 32 was set to 3.471 μm, which is 11% smaller than the pitch D1. The pitch of the reflectors 23 was set to 3.978 μm. In Model 2, the pitch D1 in the central region 30 of the input IDT 21 and the output IDT 22 was set to 5.0 μm. The pitch D2 in the edge region 32 was set to 4.450 μm, which is 11% smaller than the pitch D1. The pitch of the reflectors 23 was set to 5.100 μm.

[0039] 5(a) and 5(b) show simulation results of the attenuation and the real part of admittance Real(Y) versus frequency for Models 1 and 2. As shown in FIGS. 5(a) and 5(b), both Models 1 and 2 have filter characteristics utilizing two resonance modes. That is, the passband is formed using two resonance points of the real part of admittance Real(Y). Model 1 has a narrower frequency interval F between the resonance points of the real part of admittance Real(Y) than Model 2. That is, the frequency interval F for Model 2 is approximately 40 MHz, while the frequency interval F for Model 1 is approximately 32 MHz. This is thought to be because Model 1 has a second layer 42 made of molybdenum, which is denser (heavier) than aluminum, between the first layer 41 made of aluminum and the piezoelectric layer 14, compared to Model 2. In Model 1, the passband bandwidth is narrower due to the narrower frequency interval F.

[0040] Figures 6(a) and 6(b) show simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency for Models 1 and 2. The passband characteristics of the symmetric mode Ye and the antisymmetric mode Yo were calculated by converting them into Y parameters. For the symmetric mode Ye, Ye = (Y11 + Y22 + 2Y12) / 2 was calculated, and for the antisymmetric mode Yo, Yo = (Y11 + Y22 - 2Y12) / 2 was calculated. Multimode filters are generally designed so that the resonant frequency of the symmetric mode Ye matches the antiresonant frequency of the antisymmetric mode Yo, and so that the antiresonant frequency of the symmetric mode Ye matches the resonant frequency of the antisymmetric mode Yo. As shown in Figures 6(a) and 6(b), Model 1 has a larger difference between the antiresonant frequency of the symmetric mode Ye and the resonant frequency of the antisymmetric mode Yo than Model 2. Therefore, as shown in Figures 5(a) and 5(b), Model 1 exhibits larger loss in the passband than Model 2.

[0041] In this way, in a structure in which a thin piezoelectric layer 14 is provided on a support substrate 10, miniaturization can be achieved by using a first layer 41 and a second layer 42 that is denser than the first layer 41, but the bandwidth of the passband becomes narrower and the loss in the passband becomes larger.

[0042] [Simulation 2] A simulation was performed on the case where the number of pairs of electrode fingers 18 in the edge region 32 was increased in comparison with Model 1. The simulation conditions were the same as those for Model 1 of Simulation 1, except that the number of pairs of electrode fingers 18 in the edge region 32 was changed.

[0043] 7(a) to 8(c) show simulation results of attenuation and the real part of admittance Real(Y) versus frequency when the number of pairs of electrode fingers 18 in the edge region 32 is increased relative to Model 1. FIGS. 9(a) to 10(c) show simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency when the number of pairs of electrode fingers 18 in the edge region 32 is increased relative to Model 1. Simulation results for the case where the number of pairs of electrode fingers 18 in the edge region 32 is three are shown in FIGS. 7(a) and 9(a). Simulation results for the case where the number of pairs is four are shown in FIGS. 7(b) and 9(b). Simulation results for the case where the number of pairs is five are shown in FIGS. 7(c) and 9(c). Simulation results for the case where the number of pairs is six are shown in FIGS. 8(a) and 10(a). Simulation results for the case where the number of pairs is seven are shown in FIGS. 8(b) and 10(b). Simulation results for the case where the number of pairs is eight are shown in FIGS. 8(c) and 10(c).

[0044] As shown in Figures 7(a) to 8(c), increasing the number of pairs of electrode fingers 18 in the edge region 32 tends to disrupt the passband characteristics, making it difficult to obtain bandpass characteristics. As shown in Figures 9(a) to 10(c), increasing the number of pairs of electrode fingers 18 in the edge region 32 tends to increase the number of resonance points and antiresonance points of the symmetric mode Ye and the antiresonance points of the antisymmetric mode Yo. Furthermore, increasing the number of pairs of electrode fingers 18 in the edge region 32 tends to increase the difference between the resonance frequency of the symmetric mode Ye and the antiresonance frequency of the antisymmetric mode Yo, and therefore tends to increase the difference between the antiresonance frequency of the symmetric mode Ye and the resonance frequency of the antisymmetric mode Yo. For this reason, as shown in Figures 7(a) to 8(c), it is thought that increasing the number of pairs of electrode fingers 18 in the edge region 32 makes it difficult to obtain bandpass characteristics.

[0045] [Simulation 3] In Simulation 3, models 3 and 4 were used in which the pitch D2 of the electrode fingers 18 in the end regions 32 of the input IDT 21 and the output IDT 22 decreases from the central region 30 side toward the opposite side to the central region 30, as shown in FIG. 2(a).

[0046] FIG. 11 is a diagram showing the pitch of the electrode fingers 18 relative to the number of pairs of the electrode fingers 18 in Model 3. The number of pairs of the electrode fingers 18 on the horizontal axis is counted starting from 0 at the end of one reflector 23. As shown in FIG. 11 , in Model 3, the pitch D1 in the central region 30 of the input IDT 21 and the output IDT 22 was set to 3.9 μm. The pitch D2 in the edge regions 32 of the input IDT 21 and the output IDT 22 was varied so as to decrease by 2.2% from the central region 30 side toward the opposite side from the central region 30. Specifically, the pitch D2 located closest to the central region 30 in the edge regions 32 was set to 3.8142 μm, which is 2.2% smaller than the pitch D1. The adjacent pitch D2 was set to 3.7284 μm, which is 4.4% smaller. The adjacent pitch D2 was set to 3.6426 μm, which is 6.6% smaller. The adjacent pitch D2 was set to 3.5568 μm, which is 8.8% smaller. The pitch D2 located at the endmost side was further reduced by 11.0% to 3.4710 μm. The pitch of the reflectors 23 was set to 4.0560 μm.

[0047] The other simulation conditions for Model 3 are the same as those for Model 1 of Simulation 1, except that the aperture length of comb-shaped electrode 20 is set to 40λ.

[0048] FIG. 12 is a diagram showing the pitch of the electrode fingers 18 relative to the number of pairs of the electrode fingers 18 in Model 4. The number of pairs of the electrode fingers 18 on the horizontal axis is counted starting from 0 at the end of one reflector 23. As shown in FIG. 12, in Model 4, the pitch D1 in the central region 30 of the input IDT 21 and the output IDT 22 was set to 3.9 μm. The pitch D2 in the edge regions 332 of the input IDT 21 and the output IDT 22 was varied so as to decrease by 1.4% from the central region 30 side toward the opposite side from the central region 30. Specifically, the pitch D2 located closest to the central region 30 in the edge region 32 was set to 3.8464 μm, approximately 1.4% smaller than the pitch D1. The adjacent pitch D2 was set to 3.7928 μm, approximately 2.8% smaller. The adjacent pitch D2 was set to 3.7391 μm, approximately 4.1% smaller. The adjacent pitch D2 was set to 3.6855 μm, approximately 5.5% smaller. The pitch D2 next to that was set to 3.6319 μm, which is approximately 6.9% smaller. The pitch D next to that was set to 3.5783 μm, which is approximately 8.3% smaller. The pitch D2 next to that was set to 3.5246 μm, which is approximately 9.6% smaller. The pitch D2 located at the end was set to 3.4710 μm, which is 11.0% smaller. The pitch of the reflectors 23 was set to 4.0560 μm.

[0049] The other simulation conditions for Model 4 were the same as those for Model 1 in Simulation 1, except that the thickness of the second layer 42 was 0.230 μm, a 20° rotated Y-cut X-propagation lithium tantalate layer was used for the piezoelectric layer 14, and the aperture length of the comb electrode 20 was 20λ.

[0050] FIG. 13(a) shows the simulation results of the attenuation and the real part of the admittance Real(Y) versus frequency in Model 3. FIG. 13(b) shows the simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency in Model 3. As shown in FIG. 13(a), Model 3 has filter characteristics utilizing three resonance modes. That is, a passband is formed using three resonance points 50 of the real part of the admittance Real(Y). The three resonance points 50 of the real part of the admittance Real(Y) are located within a passband width 52 where the attenuation is −10 dB in the passband characteristics. Therefore, the passband width is wider. Furthermore, compared to Model 2 in which the electrode fingers 18 are formed on the first layer 41 shown in FIG. 5(b), the skirt characteristics are steeper. The passband width 52 is the width of the passband within which there is no point where the absolute value of the attenuation is greater than −10 dB.

[0051] As shown in Figure 13(b), in Model 3, two resonance points and two antiresonance points are formed in the symmetric mode Ye, and three resonance points and three antiresonance points are formed in the antisymmetric mode Yo. This is thought to be the reason for the filter characteristics that utilize three resonance modes, as shown in Figure 13(a). Furthermore, as shown in Figure 13(b), the difference between the resonance frequency of the symmetric mode Ye and the antiresonance frequency of the antisymmetric mode Yo, and the difference between the antiresonance frequency of the symmetric mode Ye and the resonance frequency of the antisymmetric mode Yo, are small. This is thought to be the reason for the suppression of loss deterioration in the pass band, as shown in Figure 13(a). Loss in the pass band is suppressed to -3 dB or less.

[0052] FIG. 14(a) is a diagram showing simulation results of the attenuation and the real part of admittance Real(Y) versus frequency in Model 4. FIG. 14(b) is a diagram showing simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency in Model 4. As shown in FIG. 14(a), Model 4 has filter characteristics that utilize four resonance modes. That is, four resonance points 50 of the real part of admittance Real(Y) exist within a passband width 52 where the attenuation is −10 dB in the passband characteristics, and a passband is formed using the four resonance points 50 of the real part of admittance Real(Y). This widens the passband width.

[0053] As shown in Figure 14(b), in Model 4, three resonance points and three antiresonance points are formed in the symmetric mode Ye, and four resonance points and four antiresonance points are formed in the antisymmetric mode Yo. This is thought to be the reason for the filter characteristics utilizing four resonance modes, as shown in Figure 14(a). Furthermore, as shown in Figure 14(b), the difference between the resonance frequency of the symmetric mode Ye and the antiresonance frequency of the antisymmetric mode Yo, as well as the difference between the antiresonance frequency of the symmetric mode Ye and the resonance frequency of the antisymmetric mode Yo, are small. This is thought to be the reason for the suppression of loss deterioration in the pass band, as shown in Figure 14(a). Loss in the pass band is suppressed to -3 dB or less.

[0054] [Simulation 4] In Models 5 and 6 used in Simulation 4, a material other than molybdenum (Mo) was used for the second layer 42 of the electrode fingers 18 of the input IDT 21 and the output IDT 22. In Model 5, the same simulation conditions as in Model 3 were used, except that the second layer 42 was a tungsten (W) layer with a thickness of 0.135 μm. In Model 6, the same simulation conditions as in Model 3 were used, except that the second layer 42 was a platinum (Pt) layer with a thickness of 0.120 μm.

[0055] FIG. 15(a) is a diagram showing simulation results of the attenuation and the real part of the admittance, Real(Y), versus frequency in Model 5. FIG. 15(b) is a diagram showing simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency in Model 5. FIG. 16(a) is a diagram showing simulation results of the attenuation and the real part of the admittance, Real(Y), versus frequency in Model 6. FIG. 16(b) is a diagram showing simulation results of the symmetric mode Ye and the antisymmetric mode Yo versus frequency in Model 6. As shown in FIGS. 15(a) and 16(a), Models 5 and 6 have filter characteristics utilizing three resonance modes. That is, three resonance points 50 of the real part of the admittance, Real(Y), exist within a passband width 52 where the attenuation is −10 dB in the passband characteristics, and a passband is formed using the three resonance points 50 of the real part of the admittance, Real(Y). This results in a wide passband width.

[0056] As shown in Figures 15(b) and 16(b), Models 5 and 6 have two resonance points and two antiresonance points in the symmetric mode Ye, and three resonance points and three antiresonance points in the antisymmetric mode Yo. This is thought to be the reason for the filter characteristics utilizing three resonance modes, as shown in Figures 15(a) and 16(a). Furthermore, as shown in Figures 15(b) and 16(b), the difference between the resonance frequency of the symmetric mode Ye and the antiresonance frequency of the antisymmetric mode Yo, as well as the difference between the antiresonance frequency of the symmetric mode Ye and the resonance frequency of the antisymmetric mode Yo, are small. This is thought to be why the deterioration of loss in the passband is suppressed, as shown in Figures 15(a) and 16(a). Loss in the passband is suppressed to less than -3 dB.

[0057] In the above simulation, molybdenum, tungsten, or platinum is used for the second layer 42. However, even if ruthenium or tantalum, which have similarly high densities, are used, it is believed that filter characteristics utilizing three or more resonance modes can be obtained by setting appropriate simulation conditions.

[0058] From the above, in the first embodiment, as shown in FIG. 13(a) and the like, three or more resonance points 50 in the frequency characteristics of the real part of admittance are set within the range of the passband width 52 at which the attenuation is −10 dB in the passband characteristics. As a result, even when the thickness of the piezoelectric layer 14 is set to 3 μm or less to improve the Q value and suppress spurious and loss, and when electrode fingers 18 including a first layer 41 and a second layer 42 having a higher density than the first layer 41 are used for miniaturization, the passband width can be widened and loss in the passband can be reduced. From the viewpoint of increasing the passband width and reducing loss in the passband, Real(Y) of at least one of the three or more resonance points 50 is set to 1×10 -1 s or more, and all Real(Y) is 0.5×10 -1 s or more. All Real(Y) is 0.8×10 -1 s or more is more preferable, and 1.0 × 10 -1 It is more preferable that the value is s or more.

[0059] 2, 11, and 12, in the input IDT 21 and the output IDT 22, the pitch D2 (first pitch) of the plurality of electrode fingers 18 in the edge region 32 is set smaller than the pitch D1 (second pitch) of the plurality of electrode fingers 18 in the central region 30. This makes it possible to widen the passband width and reduce loss in the passband, as shown in FIG. 13(a) and so on.

[0060] 2, 11, and 12, the pitch D2 in the end regions 32 of the input IDT 21 and the output IDT 22 is set to decrease from the central region 30 side toward the opposite side of the central region 30. This makes it possible to widen the passband width and reduce loss in the passband, as shown in FIG. 13(a) and other figures.

[0061] 2, 11, and 12, the pitch D2 in the end regions 32 of the input IDT 21 and the output IDT 22 is made smaller by one pitch each, thereby making it possible to widen the passband width and reduce loss in the passband, as shown in FIG.

[0062] In Example 1, as in the simulation, the thickness of the second layer 42 of the electrode finger 18 is set to 20% or more of the thickness of the electrode finger 18. This reduces the sound velocity of the surface acoustic waves excited by the electrode finger, thereby enabling the filter to be miniaturized. Furthermore, in the passband characteristics, three or more resonance points 50 in the frequency characteristics of the real part of the admittance can be present within the passband width 52 at an attenuation of −10 dB. The thickness of the second layer 42 may be 25% or more, or even 30% or more, of the thickness of the electrode finger 18. If the second layer 42 is too thick, the coupling coefficient decreases. Therefore, the thickness of the second layer 42 is preferably 70% or less, more preferably 60% or less, and even more preferably 50% or less of the thickness of the electrode finger 18.

[0063] In the first embodiment, the normalized film thickness of the electrode fingers 18 is defined as in Equation 3. In Equation 3, ρi is the density of each layer of the electrode fingers 18, hi is the film thickness of each layer, and ρo is the density of molybdenum.

number

[0064] In Model 3 of Simulation 3, the value of the number 3 of electrode fingers 18 is 0.114λ. In Model 4 of Simulation 3, the value of the number 3 of electrode fingers 18 is 0.110λ. In Model 5 of Simulation 4, the value of the number 3 of electrode fingers 18 is 0.099λ. In Model 6 of Simulation 4, the value of the number 3 of electrode fingers 18 is 0.096λ. Therefore, when a rotated Y-cut X-propagation lithium tantalate layer having a cut angle of 5° to 42° is used for piezoelectric layer 14, the value of the number 3 of electrode fingers 18 is preferably 0.080λ or more, and more preferably 0.096λ or more. In this case, the sound velocity of the surface acoustic wave excited by electrode fingers 18 is 3200 m / s or less. On the other hand, as described in Patent Document 6, the value of the number 3 of electrode fingers 18 is preferably 0.24λ or less.

[0065] In the first embodiment, the electrode fingers 18 include a first layer 41 mainly made of aluminum and a second layer 42 mainly made of molybdenum, tungsten, platinum, ruthenium, or tantalum. This reduces the sound velocity of the surface acoustic waves excited by the electrode fingers, thereby miniaturizing the filter, and allows three or more resonance points 50 in the frequency characteristics of the real part of the admittance to exist within the passband width 52.

[0066] 11 and 12, in the first embodiment, the number of pitches D2 of the electrode fingers 18 in the end region 32 is preferably 5 or more. This allows three or more resonance points 50 in the frequency characteristics of the real part of the admittance to exist within the passband width 52. [Example]

[0067] 17(a) is a circuit diagram of a filter 200 according to a second embodiment. As shown in FIG. 17(a), a series resonator S1, a multimode filter 60, and a series resonator S2 are connected in series between an input terminal Tin and an output terminal Tout. A parallel resonator P1 and a parallel resonator P2 are connected in parallel between the input terminal Tin and the output terminal Tout. Inductors L1 to L3 may be connected between the parallel resonator P1 and ground, between the multimode filter 60 and ground, and between the parallel resonator P2 and ground. The multimode filter 60 can be the multimode filter of the first embodiment.

[0068] Fig. 17(b) is a plan view of a filter 210 according to a modified example of the second embodiment. As shown in Fig. 17(b), a multimode filter 62 and a multimode filter 64 are connected in series between an input terminal Tin and an output terminal Tout. At least one of the multimode filters 62 and 64 can be the multimode filter of the first embodiment. [Example]

[0069] FIG. 18 is a circuit diagram of a duplexer 300 according to a third embodiment. As shown in FIG. 18, a transmit filter 70 is connected between a common terminal Ant and a transmit terminal Tx. A receive filter 72 is connected between the common terminal Ant and a receive terminal Rx. The transmit filter 70 passes, to the common terminal Ant, signals in the transmit band among the high-frequency signals input from the transmit terminal Tx as transmit signals, and suppresses signals of other frequencies. The receive filter 72 passes, to the receive terminal Rx, signals in the receive band among the high-frequency signals input from the common terminal Ant as receive signals, and suppresses signals of other frequencies. At least one of the transmit filter 70 and the receive filter 72 can be the multimode filter of the first embodiment or the filter of the second embodiment. Although a duplexer has been described as an example of a multiplexer, a triplexer or a quadplexer may also be used.

[0070] Although the embodiments of the present invention have been described in detail above, the present invention is not limited to such specific embodiments, and various modifications and variations are possible within the scope of the gist of the present invention as described in the claims. [Explanation of symbols]

[0071] 10...support substrate, 11...first insulating layer, 12...second insulating layer, 14...piezoelectric layer, 16...metal film, 18...electrode fingers, 19...bus bar, 20...comb-shaped electrode, 21...input IDT, 22...output IDT, 23...reflector, 25...intersection region, 30...central region, 32...end region, 41...first layer, 42...second layer, 43...third layer, 50...resonance point, 52...passband width, 60...multimode filter, 62...multimode filter, 64...multimode filter, 70...transmit filter, 72...receive filter, 100...multimode filter, 200...filter, 210...filter, 300...duplexer

Claims

1. A support substrate; a piezoelectric layer having a thickness of 3 μm or less provided on the support substrate; one or more input IDTs provided on the piezoelectric layer, each having a plurality of electrode fingers, each of which is formed by stacking a first layer and a second layer having a density greater than that of the first layer; one or more output IDTs, which are arranged alternately with the one or more input IDTs on the piezoelectric layer, and which have the plurality of electrode fingers in which the first layer and the second layer are stacked, A multimode filter in which there are three or more resonance points in the frequency characteristics of the real part of admittance within the range of the passband width when the attenuation is -10 dB in the passband characteristics.

2. 2. The multimode filter according to claim 1, wherein the one or more input IDTs and the one or more output IDTs have a first pitch of the electrode fingers in end regions located at the ends of a central region located at the center in a direction in which the electrode fingers are arranged, the first pitch being smaller than a second pitch of the electrode fingers in the central region.

3. 3. The multimode filter according to claim 2, wherein the first pitch in the end regions decreases from the central region side toward the opposite side from the central region.

4. 4. The multimode filter of claim 3, wherein the first pitch decreases with each pitch.

5. The real part of the admittance of the three or more resonance points is 0.5×10 －１ 3. The multimode filter according to claim 1, wherein the .lambda.

6. 3. The multimode filter according to claim 1, wherein the thickness of said second layer is 20% or more of the thickness of said plurality of electrode fingers.

7. the plurality of electrode fingers are formed by stacking a plurality of layers including the first layer and the second layer, and when ρi is the density of each of the plurality of layers, hi is the film thickness of each of the plurality of layers, ρo is the density of molybdenum, and λ is twice the average pitch of the plurality of electrode fingers in a central region located at the center in a direction in which the plurality of electrode fingers are arranged in the one or more input IDTs and the one or more output IDTs, the value of the following formula is 0.08λ or more and 0.24λ or less:

3. The multimode filter according to claim 1 or 2.

8. the first layer is a layer containing aluminum as a main component, 3. The multimode filter according to claim 1, wherein the second layer is a layer containing molybdenum, tungsten, platinum, ruthenium, or tantalum as a main component.

9. A filter comprising the multimode filter according to claim 1 or 2.

10. A multiplexer comprising the multimode filter according to claim 1 or 2.

Images