Fan-in / fan-out devices

Abstract

To suppress increase in optical coupling loss even if a beam waist diameter changes and variations in the beam waist diameter occur.SOLUTION: An FIFO device 10 includes an MCF 20, a first lens 30 having a first optical axis parallel to a central axis of the MCF, a second lens group 40 having a plurality of second lenses 41 to 44 having second optical axes parallel to the first optical axis, and a single core optical fiber group 50 having the same number of single optical fibers 51 to 54 as that of the second lenses. When a maximum value of a beam waist distance sum in a case where an inter-lens distance between the first lens and the respective second lenses is equal to a beam waist distance sum is regulated as a distance sum maximum value, in a case where the inter-lens distance is set at the distance sum maximum value, the MCF 20, the first lens 30, the second lens group 40 and the single core optical fiber group 50 are arranged so that the beam waist distance sum becomes 91.5% or more of the distance sum maximum value.SELECTED DRAWING: Figure 11C

Description

[Technical Field]

[0001] The present invention relates to a fan-in / fan-out device. In particular, it relates to a spatially coupled fan-in / fan-out device comprising a multi-core optical fiber and a plurality of single-core optical fibers, which optically couple the two. [Background technology]

[0002] Internet traffic demand is increasing year by year, and there is a growing need for even faster and higher-capacity optical communications. To meet this demand, technologies such as Wavelength Division Multiplexing (WDM) and digital coherent technology have been used to increase transmission capacity.

[0003] In recent years, Space Division Multiplexing (SDM) technology using multicore optical fibers has attracted attention as a new multiplexing technology. SDM technology is said to enable even higher speeds and larger capacities. With the progress of research and development of SDM technology, the demand for fan-in / fan-out (FIFO) devices is increasing. A FIFO device is an optical device that comprises a multicore optical fiber and multiple single-core optical fibers, and optically couples the two.

[0004] Examples of FIFO devices include spatially coupled, fiber bundled, and melt-stretched devices. Spatially coupled FIFO devices are characterized by optically coupling a multicore optical fiber and a single-core optical fiber using a lens (including a glass block, etc.) (see Patent Document 1). Hereinafter, spatially coupled FIFO devices will simply be referred to as "FIFO devices". [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Patent No. 6554891 [Overview of the project]

[0006] The FIFO device comprises a first lens provided on the multicore optical fiber side and a plurality of second lenses provided on the plurality of single-core optical fibers. In the FIFO device, a light ray emitted from one of the multicore optical fibers or single-core optical fibers passes through the first and second lenses and is incident on the other of the multicore optical fiber or single-core optical fiber. Optical coupling loss becomes zero when the incident end-face beam diameter, which is the beam diameter of the light ray at the end face of the other optical fiber, matches the mode field diameter of that optical fiber.

[0007] If an error occurs in the first distance between the multicore optical fiber and the first lens, or in the second distance between the single-core optical fiber and the corresponding second lens, the beam waist diameter (and beam waist position) of the light ray will change. As a result, the incident end-face beam diameter will no longer match the mode field diameter, leading to an increase in optical coupling loss. Therefore, when manufacturing FIFO devices, it is desirable to minimize errors in the first and second distances.

[0008] However, since minute errors in the first and second distances are unavoidable during the manufacturing process, there is a need to develop a FIFO device that can suppress the increase in optical coupling loss (i.e., one that is less prone to an increase in optical coupling loss) even when the beam waist diameter changes due to these errors.

[0009] Furthermore, in FIFO devices, the end faces of multicore optical fibers are sometimes beveled to reduce reflected light. In this case, variations occur in the first distance depending on the number of cores or core arrangement of the multicore optical fiber, resulting in variations in the beam waist diameter. For example, if two variations occur in the first distance due to beveling of the end face of the multicore optical fiber, the beam waist diameter will take on two different values, resulting in variation. Variations in beam waist diameter lead to an increase in optical coupling loss. Therefore, it is desirable to suppress the increase in optical coupling loss not only when the beam waist diameter changes due to manufacturing errors, but also when variations occur in the beam waist diameter due to beveling of the end face of the multicore optical fiber.

[0010] This invention was made to address the problems described above. Specifically, one of the objectives of this invention is to provide a technique that can suppress the increase in optical coupling loss even when the beam waist diameter changes or when there are variations in the beam waist diameter.

[0011] The fan-in / fan-out device (10) according to the present invention is A multicore optical fiber (20) comprising a plurality of columnar first cores (C1 to C4) extending along the axial direction, and a common cladding (CL) surrounding the plurality of first cores, A first lens (30) is provided corresponding to the multicore optical fiber, having a first optical axis parallel to the central axis of the multicore optical fiber (20), A second lens group (40) having a plurality of second lenses (41 to 44) having a second optical axis parallel to the first optical axis, A single-core optical fiber group (50) comprising the same number of single-core optical fibers (51 to 54) as the second lenses (41 to 44), each having a columnar second core (C) extending along a central axis parallel to the second optical axis, and cladding (CLs) surrounding the second core (C), The multicore optical fiber (20) is configured such that light rays propagate in either a first propagation direction, which is the direction in which light rays are emitted from each of the first cores (C1 to C4) of the multicore optical fiber (20), pass through the first lens (30) and the second lenses (41 to 44) corresponding to each of the first cores, and converge on the second core (C) of the single-core optical fiber (51 to 54) corresponding to the second lens, or a second propagation direction, which is the direction in which light rays are emitted from each of the second cores, pass through the corresponding second lens and the first lens, and converge on each of the first cores corresponding to the second lens. Assuming that the light rays propagate in the first direction of propagation, the beam waist diameter of each light ray emitted from the first lens (30) is defined as the first beam waist diameter (2Ω1), and the distance from the first lens to the beam waist position in the direction of propagation of the principal light ray of each light ray is defined as the first beam waist distance (D1). Assuming that the light ray propagates in the second direction of propagation, the beam waist diameter of the light ray emitted from each of the second lenses (41 to 44) is defined as the second beam waist diameter (2Ω2), and the distance from the second lens to the beam waist position in the direction of propagation of the principal ray of the light ray is defined as the second beam waist distance (D2). If the inter-lens distance (Z), which is the distance between the first lens (30) and each of the second lenses (41 to 44) in the direction of propagation of the principal ray of the light, is equal to the sum of beam waist distances (D1+D2), which is the sum of the first beam waist distance (D1) and the second beam waist distance (D2) when the first beam waist diameter (2Ω1) and the second beam waist diameter (2Ω2) coincide, then the maximum value of the sum of beam waist distances is defined as the maximum value of the sum of beam waist distances (D1+D2_max), The multicore optical fiber (20), the first lens (30), the second lens group (40), and the single-core optical fiber group (50) are arranged such that the inter-lens distance (Z) is substantially equal to the maximum value of the distance sum (D1 + D2_max), and the beam waist distance sum (D1 + D2) is 91.5% or more of the maximum value of the distance sum.

[0012] Furthermore, another fan-in / fan-out device according to the present invention is A multicore optical fiber (20p) comprising a plurality of columnar first cores (C1 to C4) extending along the axial direction, and a common cladding (CL) surrounding the plurality of first cores, A first lens (30) is provided corresponding to the multicore optical fiber, having a first optical axis parallel to the central axis of the multicore optical fiber (20p), A second lens group (40) having a plurality of second lenses (41 to 44) having a second optical axis parallel to the first optical axis, A single-core optical fiber group (50) comprising the same number of single-core optical fibers (51 to 54) as the second lenses (41 to 44), each having a columnar second core (C) extending along a central axis parallel to the second optical axis, and cladding (CLs) surrounding the second core (C), The multicore optical fiber (20p) is configured such that light rays propagate in either a first propagation direction, which is the direction in which light rays are emitted from each of the first cores (C1 to C4) of the multicore optical fiber (20p), pass through the first lens (30) and the second lenses (41 to 44) corresponding to each of the first cores, and converge on the second core (C) of the single-core optical fiber (51 to 54) corresponding to the second lens, or a second propagation direction, which is the direction in which light rays are emitted from each of the second cores, pass through the corresponding second lens and the first lens, and converge on each of the first cores corresponding to the second lens. Assuming that the light rays propagate in the first direction of propagation, the beam waist diameter of each light ray emitted from the first lens (30) is defined as the first beam waist diameter (2Ω1), and the distance from the first lens to the beam waist position in the direction of propagation of the principal light ray of each light ray is defined as the first beam waist distance (D1). Assuming that the light ray propagates in the second direction of propagation, the beam waist diameter of the light ray emitted from each of the second lenses (41 to 44) is defined as the second beam waist diameter (2Ω2), and the distance from the second lens to the beam waist position in the direction of propagation of the principal ray of the light ray is defined as the second beam waist distance (D2). If the inter-lens distance (Z), which is the distance between the first lens (30) and each of the second lenses (41 to 44) in the direction of propagation of the principal ray of the light, is equal to the sum of beam waist distances (D1+D2), which is the sum of the first beam waist distance (D1) and the second beam waist distance (D2) when the first beam waist diameter (2Ω1) and the second beam waist diameter (2Ω2) coincide, then the maximum value of the sum of beam waist distances is defined as the maximum value of the sum of beam waist distances (D1+D2_max), The end face (20ap) of the multicore optical fiber (20p) is obliquely polished so as to be inclined by a predetermined polishing angle in a predetermined inclination direction with respect to a plane perpendicular to its central axis, thereby creating n variations in the distance (d1(1) to d1(4)) between each of the first cores (C1 to C4) and the first lens (30) in the direction of light propagation, and also creating n variations in the first beam waist diameter (2Ω1) of the light rays from each of the first cores. Each of the single-core optical fibers (51 to 54) is positioned relative to the corresponding second lens (41 to 44) such that the second beam waist diameter (2Ω2) of the light ray corresponding to each of the single-core optical fibers coincides with the corresponding first beam waist diameter (2Ω1). Furthermore, if we define the first and second beam waist diameters (2Ω1=2Ω2) when the sum of beam waist distances (D1+D2) is equal to the maximum value of the sum of distances (D1+D2_max), and define the maximum and minimum values ​​of the n possible first and second beam waist diameters as the maximum beam waist diameter (2Ωmax) and minimum beam waist diameter (2Ωmin), respectively, The multi-core optical fiber (20p) is arranged with respect to the first lens (30) such that the maximum beam waist diameter (2Ωmax) is larger than the beam waist diameter at the maximum distance (Ω_Dmax), and the minimum beam waist diameter (2Ωmin) is smaller than the beam waist diameter at the maximum distance.

[0013] According to the present invention, even when the beam waist diameter changes or there is variation in the beam waist diameter, it is possible to suppress an increase in optical coupling loss.

Brief Description of the Drawings

[0014] [Figure 1] It is a perspective view showing a FIFO device according to a first embodiment of the present invention. [Figure 2] It is a side view of the FIFO device. [Figure 3] It is a view showing an end face of a multi-core optical fiber included in the FIFO device. [Figure 4] It is a view showing a state where the principal ray of the light emitted from a certain core of the multi-core optical fiber passes through the corresponding second lens. [Figure 5] It is a view showing an end face of a single-core optical fiber included in the FIFO device. [Figure 6] It is a view for explaining the first beam waist diameter 2Ω1 and the first beam waist distance D1. [Figure 7] It is a view for explaining the second beam waist diameter 2Ω2 and the second beam waist distance D2. [Figure 8A] It is a graph defining the relationship between the distance d2 from the single-core optical fiber to the corresponding second lens and the second beam waist distance D2. [Figure 8B] It is a graph defining the relationship between the distance d2 and the second beam waist radius Ω2. [Figure 9] It is a view showing the positional relationship of the components when the coupling loss of the FIFO device is zero. [Figure 10]This graph defines the relationship between the beam waist radius Ω1=Ω2 and the beam waist distance sum D1+D2 when 2Ω1=2Ω2 holds true. [Figure 11A] This graph illustrates the coupling loss when the distance between lenses Z = 60 mm. [Figure 11B] This graph illustrates the coupling loss when Z = 70 mm. [Figure 11C] This graph illustrates the coupling loss when Z = 80.89 mm. [Figure 12A] This graph corresponds to Figure 11C, showing the case when the second lens is changed to a different lens with a different focal length. [Figure 12B] This graph corresponds to Figure 11C, showing the case when the second lens is changed to yet another lens with a different focal length. [Figure 13] This is a normalized graph used to examine the beam waist distance sum D1+D2 that can suppress coupling loss to 0.15 dB or less. [Figure 14] This is a side view showing only the obliquely polished multicore optical fiber and the first lens of a modified FIFO device according to the present invention. [Figure 15] This graph defines the relationship between the distance d1 from the center of a multicore optical fiber to the first lens and Ω1 for each ray from the core. [Figure 16] This graph defines the relationship between two variations of Ω1. [Figure 17] This graph illustrates the beam waist distance sum D1+D2 that can suppress coupling loss to 0.15 dB or less. [Figure 18] This graph illustrates the conditions that two different combinations of Ω1 satisfy for a FIFO device according to a second embodiment of the present invention. [Figure 19] This graph defines the relationship between two variations of Ω1. [Figure 20] This is a graph of a comparative example of a FIFO device according to the third embodiment of the present invention. [Figure 21A] This graph illustrates the coupling loss when Z=80mm, and explains the conditions that the two variations of Ω1 combinations satisfy. [Figure 21B] This graph illustrates the coupling loss when Z = 79 mm. [Figure 21C] This graph illustrates the coupling loss when Z = 78 mm. [Modes for carrying out the invention]

[0015] (First Embodiment) A FIFO device 10 according to the first embodiment of the present invention will be described with reference to the drawings. Hereinafter, "FIFO device" will also be simply referred to as "device".

[0016] Figure 1 is a perspective view of device 10, and Figure 2 is a side view of device 10. As shown in Figures 1 and 2, device 10 comprises a multicore optical fiber 20, a first lens 30, a second lens group 40, and a single-core optical fiber group 50. These components are arranged in the above order along axis A1. A Cartesian coordinate system is set up for device 10. The z-axis extends parallel to axis A1 such that the direction from the multicore optical fiber 20 toward the first lens 30 is the positive direction. The y-axis is perpendicular to the z-axis and extends such that the plane of the paper is the positive direction. The x-axis is perpendicular to the z-axis and y-axis. Hereinafter, the multicore optical fiber and single-core optical fiber will also be referred to as "MCF" and "SCF," respectively. In this specification, for the sake of clarity, the dimensions of certain components (e.g., MCF 20 and SCF group 50) and the angle of light rays have been changed in the illustrations.

[0017] The MCF20 is cylindrical, and its central axis, at least at its +z-axis end, coincides with axis A1. The end face 20a of the MCF20 (see Figure 2) is parallel to the plane perpendicular to axis A1 (the xy-plane). Figure 3 shows the end face 20a of the MCF20 viewed along its central axis. As shown in Figure 3, the MCF20 comprises four cores C1 to C4 and a common cladding CL surrounding these cores C1 to C4. The cores C1 to C4 are located at the vertices of a square centered on the end face 20a and extend along the axial direction. The distance between adjacent cores (core pitch p1) is 50 μm. Both the cores C1 to C4 and the cladding CL are formed from glass with quartz as the main component. The refractive index of the cores C1 to C4 is greater than that of the cladding CL. The MCF20 is an optical fiber through which single-mode light propagates. Furthermore, the materials of the cores C1 to C4 and the cladding CL are not limited to glass primarily composed of quartz, but may be formed from other materials. In addition, in this specification, cylinders with curved axes are also included.

[0018] As shown in Figures 1 and 2, the +z-axis end of the MCF20 is inserted into and held by a cylindrical ferrule 22. The end face 22a of the ferrule 22 is coplanar with the end face 20a of the MCF20. This is because the end face 20a of the MCF20 is polished together with its end face 22a while it is inserted into the ferrule 22. In Figure 2, the MCF20 inside the ferrule 22 is shown by a dashed line, but the cores C1 to C4 are not shown.

[0019] Light rays propagating through each core C1 to C4 of the MCF20 are emitted from the end face 20a toward the first lens 30. That is, the MCF20 functions as an emission member. In Figure 1, only the principal rays B1 to B4 of the light rays emitted from each core C1 to C4 (see Figure 3) are shown, and in Figure 2, only the principal rays B2 and B3 of the light rays emitted from cores C2 and C3 are shown. The principal rays of the light emitted from each core C1 to C4 are parallel to each other, but each emitted light is divergent light that diverges as it propagates (described later). In this embodiment, a light ray with a wavelength of 1.55 μm is used, but the wavelength value is not limited to this.

[0020] The first lens 30 is a collimating lens with a focal length df1 of 1.3 mm, and more specifically, an aspherical lens having a rotationally symmetric curved surface. The first lens 30 collimates (parallels) the light rays emitted and diverging from each core C1 to C4. The optical axis of the first lens 30 is located on the central axis of the MCF 20 (i.e., on axis A1). The first lens 30 deflects and emits light rays whose principal rays B1 to B4 are parallel to each other, emitted from each core C1 to C4. In other words, the first lens 30 focuses the light rays from each core C1 to C4 at a focal point f1. That is, the first lens 30 is a lens provided in correspondence with a multicore optical fiber. Note that the curved surface of the first lens 30 may be rotationally symmetric as long as it can deflect and emit light rays from each core C1 to C4. Furthermore, the first lens 30 may be a spherical lens or a GRIN lens, or it may be a lens with one flat surface.

[0021] The second lens group 40 has the same number of second lenses 41 to 44 as the number of cores in the MCF20 (four in this example) (see Figure 1). The second lenses 41 to 44 are all collimating lenses with a focal length df2 of 2.5 mm, and more specifically, they are aspherical lenses with rotationally symmetric curved surfaces. Hereafter, the second lenses 41 to 44 will be simply referred to as "lenses 41 to 44". The optical axes of each lens 41 to 44 are parallel to the optical axis of the first lens 30. Also, the principal points of lenses 41 to 44 are located on the same plane, and this plane is perpendicular to axis A1. In Figure 2, only lenses 42 and 43, into which the principal rays B2 and B3 are incident, are shown among lenses 41 to 44.

[0022] Figure 4 shows how the principal ray B3 of the light emitted from core C3 passes through the corresponding lens 43. As shown in Figures 1, 2, and 4, the light ray from core C3 of the MCF20 emitted from the first lens 30 passes through focal point f1. The principal ray B3 of the light ray that has passed through focal point f1 travels in a straight line and passes through focal point f2 of lens 43 (see Figure 4), and is incident on position Ps3 of lens 43 at a predetermined incident angle (approximately 1.6° in this example). The principal ray B3 that has been incident on position Ps3 is emitted from lens 43 as a ray parallel to the optical axis As3 of lens 43 (see Figure 4).

[0023] Similarly, as shown in Figure 1, the principal rays B1, B2, and B4 of the light rays that have passed through focal point f1 travel in a straight line and pass through focal points f2 (not shown) of lenses 41, 42, and 44, respectively, and are incident on positions Ps1, Ps2, and Ps4 of lenses 41, 42, and 44 at a predetermined angle of incidence (approximately 1.6° in this example). The principal rays B1, B2, and B4 that have been incident on positions Ps1, Ps2, and Ps4 are emitted from lenses 41, 42, and 44, respectively, as rays parallel to the optical axes of lenses 41, 42, and 44.

[0024] As shown in Figures 1 and 2, the second lens group 40 focuses the light rays emitted from each core C1 to C4 from the first lens 30 using the corresponding lenses 41 to 44 (only the principal rays are shown in Figures 1 and 2). The curved surfaces of each lens 41 to 44 may be non-rotationally symmetric as long as they can emit light rays that deflect the corresponding cores C1 to C4. Furthermore, each lens 41 to 44 may be a spherical lens or a GRIN lens, or a lens with one flat surface.

[0025] The SCF group 50 has the same number of SCFs 51 to 54 as lenses 41 to 44 (four in this example) (see Figure 1). SCFs 51 to 54 are all optical fibers through which single-mode light rays propagate. Since SCFs 51 to 54 have the same configuration as each other, the configuration of SCF 53 will be described below. SCF 53 is cylindrical, and its central axis at least at its -z-axis end is parallel to the optical axis As3 of the corresponding lens 43 (see Figure 4). Also, SCF 53 is located spaced apart from the optical axis of the first lens 30. The end face 53a of SCF 53 is parallel to the xy plane. Figure 5 is a view of the end face 53a of SCF 53 along its central axis. As shown in Figure 5, SCF 53 comprises one core C extending along its central axis and cladding CLs surrounding the core C. Each core C of SCFs 51 to 54 is located at the vertex of a square centered on axis A1. In this embodiment, the device 10 is designed such that the pitch p2 between adjacent cores is approximately 3.1 mm. Both the core C and the cladding CLs are formed from glass primarily composed of quartz. The refractive index of the core C is greater than that of the cladding CLs. Note that the materials of the core C and cladding CLs are not limited to glass primarily composed of quartz, but may be formed from other materials.

[0026] As shown in Figures 1 and 2, the -z-axis end of the SCF53 is inserted into and held by a cylindrical ferrule 63. The end face 53a of the SCF53 is polished together with the ferrule 63 while it is inserted into it. As a result, the end face 53a of the SCF53 and the end face 63a of the ferrule 63 are located on the same plane (xy plane). In Figure 2, the SCF53 inside the ferrule 63 is shown by a dashed line.

[0027] The end face 53a of the SCF53 is positioned so that the light rays emitted from the core C3, which are emitted from the lens 43, converge on the core C (more precisely, on the center of the core C). In other words, the SCF53 is positioned so that the principal ray B3 is incident on the center of the core C.

[0028] Similarly, SCF51, 52, and 54 are positioned spaced apart from the optical axis of the first lens 30, and their end faces 51a, 52a, and 54a (see Figure 1) are positioned so that the light rays emitted from cores C1, C2, and C4 from lenses 41, 42, and 44 converge on core C (more precisely, on the center of core C). That is, SCF51, 52, and 54 are positioned so that the principal rays B1, B2, and B4 are incident on the center of core C, respectively. As a result, the light emitted from cores C1, C2, and C4 is incident on core C of SCF51, 52, and 54 with low loss.

[0029] In this way, the first lens 30 and the second lens group 40 optically couple the MCF 20 and the SCF group 50. Hereafter, the direction of propagation of the light ray described above will also be referred to as the "first direction of propagation". When the light ray propagates in the first direction of propagation, device 10 functions as a "fan-out (FO) device". Here, since the path of light is reversible, the light ray can propagate in the device 10 in the direction opposite to the first direction of propagation. That is, device 10 can also be configured to propagate in a direction in which the light ray is emitted from each core C of SCF 51 to 54, passes through lenses 41 to 44 and the first lens 30, and converges on the cores C1 to C4 of MCF 20. Hereafter, this direction of propagation (opposite to the first direction of propagation) will also be referred to as the "second direction of propagation". When the light ray propagates in the second direction of propagation, device 10 functions as a "fan-in (FI) device".

[0030] Next, the relationship between the positional relationships of the components 20, 30, 40, and 50 of device 10 and the optical coupling loss of the light ray will be explained. In this specification, the propagation of the light ray is assumed to follow the Gaussian beam model. A Gaussian beam is a light ray in which the light intensity distribution in a cross-section perpendicular to the direction of propagation follows a Gaussian function. Note that optical coupling loss is synonymous with insertion loss, and in the following, it may simply be referred to as "coupling loss".

[0031] Figure 6 shows only the light rays emitted from the core C2 of the MCF20 and deflected by the first lens 30 (i.e., light rays traveling in the first direction of propagation). B2 represents the principal ray of these light rays. As described above, the light rays emitted from the core C2 are collimated by the first lens 30, but more precisely, they converge at the first beam waist position, which is a predetermined distance from the first lens 30, and then diverge from the first beam waist position toward the lens 42. The diameter of the light rays (hereinafter also referred to as "beam diameter") is smallest at the first beam waist position. The beam diameter and beam radius at the first beam waist position are referred to as "first beam waist diameter 2Ω1" and "first beam waist radius Ω1," respectively.

[0032] The distance from the first lens 30 to the first beam waist position in the direction of propagation of the principal ray B2 (in other words, the length of the principal ray B2 from the first lens 30 to the first beam waist position) is referred to as the "first beam waist distance D1". If the direction in which the optical axis of the first lens 30 extends is defined as the "optical axis direction," then the first beam waist distance D1 and the first beam waist diameter 2Ω1 depend on the beam diameter 2ω1 at the end face 20a of the MCF 20, the focal length df1 (=1.3 mm) of the first lens 30, and the distance d1 between the end face 20a of the MCF 20 and the first lens 30 in the optical axis direction. If the MCF 20 and the first lens 30 are not changed, the first beam waist distance D1 and the first beam waist diameter 2Ω1 substantially depend on the distance d1 (described later). Note that the beam diameter 2ω1 is equal to the mode field diameter of the MCF 20, and in this embodiment it is set to 8 μm. The above explanation also applies to the light rays emitted from cores C1, C3 and C4.

[0033] Figure 7 shows a light ray emitted from the core C of the SCF52 and deflected by the lens 42 (i.e., a light ray traveling in the second direction of propagation). The SCF52 in Figure 7 is the same component as the SCF52 used in device 10, but it is shown reversed along the x-axis, y-axis, and z-axis to explain the case where the light ray travels in the second direction of propagation. The light ray emitted from the core C is collimated by the lens 42, but more precisely, it converges at the second beam waist position, which is a predetermined distance from the lens 42, and then diverges from the second beam waist position toward the first lens 30. The beam diameter of the light ray is smallest at the second beam waist position. The beam diameter and beam radius at the second beam waist position are referred to as "second beam waist diameter 2Ω2" and "second beam waist radius Ω2," respectively.

[0034] The distance from lens 42 to the second beam waist position in the direction of propagation of the principal ray emitted from lens 42 (in other words, the length of the principal ray from lens 42 to the second beam waist position) is referred to as the "second beam waist distance D2". The second beam waist distance D2 and the second beam waist diameter 2Ω2 depend on the beam diameter 2ω2 at the end face 52a of the SCF52, the focal length df2 (=2.5 mm) of the lens 42, and the distance d2 between the end face 52a of the SCF52 and the lens 42 in the optical axis direction. If the SCF52 and lens 42 are not changed, the second beam waist distance D2 and the second beam waist diameter 2Ω2 substantially depend on the distance d2. The beam diameter 2ω2 is equal to the mode field diameter of the SCF52, and in this embodiment it is set to 10.4 μm. The above description also applies to the light rays emitted from SCF51, 53, and 54.

[0035] Figure 8A is a graph defining the relationship between distance d2 and the second beam waist distance D2 in the configuration of Figure 7, and Figure 8B is a graph defining the relationship between distance d2 and the second beam waist radius Ω2 in the configuration of Figure 7. According to Figure 8A, the second beam waist distance D2 is 2.5 mm when distance d2 is equal to the focal length df2 (=2.5 mm) of lens 42, and is maximum when distance d2 is slightly greater than the focal length df2. According to Figure 8B, the second beam waist radius Ω2 is maximum when distance d2 is equal to the focal length df2, and decreases as distance d2 increases. The above trend also applies to the relationship between distance d1 and the first beam waist distance D1 and the first beam waist radius Ω1 in the configuration of Figure 6. In this embodiment, the MCF20 is arranged relative to the first lens 30 such that distance d1 satisfies d1 > df1, and the SCF group 50 is arranged relative to the second lens group 40 such that distance d2 satisfies d2 > df2.

[0036] In Figure 7, an example is shown in which a light ray emitted from the core C of SCF52 travels in a second direction of propagation via lens 42. However, as mentioned above, the path of light is reversible. Therefore, when a light ray emitted from the core C2 of MCF20 travels in a first direction of propagation towards the core C of SCF52 via the first lens 30 and lens 42, the path of the light ray coincides with the path shown in Figure 7. Accordingly, each parameter 2Ω2(Ω2), D2, 2ω2, and d2 can be used even when the light ray travels in a first direction of propagation.

[0037] Figure 9 shows the positional relationships of each component 20, 30, 40, and 50 when the coupling loss of device 10 is zero. In Figure 9, only the light ray emitted from the core C2 of MCF20 and the corresponding lenses 42 and SCF52 are shown. As shown in Figure 9, the theoretical coupling loss is zero when the following two conditions are met. (Condition 1) The waist diameter of the first beam 2Ω1 and the waist diameter of the second beam 2Ω2 are equal (2Ω1 = 2Ω2). (Condition 2) The inter-lens distance Z, which is the distance between the first lens 30 and the corresponding lenses 41 to 44 in the direction of propagation of the principal ray of the light emitted from the first lens 30, is equal to the sum of the first beam waist distance D1 and the second beam waist distance D2 (Z = D1 + D2).

[0038] In other words, if the beam diameter 2ω2 of the incident light ray at the end faces 51a to 54a of SCF51 to 54 matches the mode field diameter of SCF51 to 54, the coupling loss is zero. If the beam diameter 2ω2 does not match the mode field diameter, a coupling loss occurs. For example, in the example in Figure 9, if the lens 42 is shifted in the +z axis direction, the beam diameter 2ω2 no longer matches the mode field diameter of SCF51 to 54, and a coupling loss occurs. Hereinafter, when condition 1 is met, the first beam waist diameter 2Ω1 (or radius Ω1) and the second beam waist diameter 2Ω2 (or radius Ω2) will be referred to as "beam waist diameter 2Ω1 = 2Ω2" or "beam waist radius Ω1 = Ω2". Also, the sum of the first beam waist distance D1 and the second beam waist distance D2 when condition 1 is met will be referred to as the "beam waist distance sum D1 + D2".

[0039] Figure 10 is a graph that defines the relationship between the beam waist radius Ω1=Ω2 and the beam waist distance sum D1+D2 when condition 1 is met. According to curve L1, D1+D2 increases as Ω1=Ω2 increases, then decreases after reaching a maximum value. Figure 10 shows that by matching the inter-lens distance Z to the beam waist distance sum D1+D2 on curve L1, the coupling loss can be reduced to zero.

[0040] Incidentally, errors may occur in the manufacturing process of device 10, such as in distance d1 and distance d2. If an error occurs in distance d1, the first beam waist diameter 2Ω1 and the first beam waist distance D1 will change. If an error occurs in distance d2, the second beam waist diameter 2Ω2 and the second beam waist distance D2 will change. In this case, conditions 1 and 2 will no longer be met, and the coupling loss may increase. For this reason, there is a need for the development of a device in which the coupling loss is less likely to increase even if manufacturing errors (variations) occur in distance d1 and / or distance d2 (in other words, a device in which the coupling loss is less likely to fluctuate even if the first and second beam waist diameters 2Ω1 and 2Ω2 change).

[0041] To investigate this point, the inventors of the present invention investigated the behavior of coupling loss by changing the inter-lens distance Z through simulation. Figures 11A to 11C are graphs that define the relationship between the beam waist radius Ω1=Ω2 and the sum of beam waist distances D1+D2 (see curve L1), and the relationship between the beam waist radius Ω1=Ω2 and coupling loss (see curves L2, L3, and L4). In Figure 11A, the inter-lens distance Z is set to 60 mm, in Figure 11B, the inter-lens distance Z is set to 70 mm, and in Figure 11C, the inter-lens distance Z is set to 80.89 mm.

[0042] In the example in Figure 11A, the beam waist distance sum D1+D2 satisfying condition 2 is 60 mm, and the beam waist radius Ω1=Ω2 at this time is 76 μm. In the example in Figure 11B, the beam waist distance sum D1+D2 satisfying condition 2 is 70 mm, and the beam waist radius Ω1=Ω2 at this time is 94 μm. In the example in Figure 11C, the beam waist distance sum D1+D2 satisfying condition 2 is 80.89 mm, and the beam waist radius Ω1=Ω2 at this time is 133 μm. Figures 11A to 11C show that as the inter-lens distance Z increases, the coupling loss becomes less prone to fluctuation even when the beam waist radius Ω1=Ω2 changes. Therefore, when manufacturing device 10, it is desirable to control Z so that the inter-lens distance Z is substantially equal to the maximum value of the beam waist distance sum. Hereinafter, the maximum value of D1+D2 that satisfies Z=D1+D2 will also be referred to as the "maximum distance sum D1+D2_max". In this embodiment, the maximum value of the distance sum D1+D2_max is 80.89 mm.

[0043] By setting the inter-lens distance Z to the maximum distance sum D1+D2_max, a device can be realized in which coupling loss is less likely to increase even if manufacturing errors occur in distance d1 and / or distance d2. Here, the minimum coupling loss of current FIFO devices (strictly speaking, FI devices or FO devices) is 0.15 dB. For this reason, the inventors of this application have investigated, based on simulations, the range of beam waist distance sum D1+D2 that can suppress coupling loss to 0.15 dB or less. The coupling loss referred to here means the coupling loss (insertion loss) for each core C1 to C4 (i.e., the corresponding core C of each SCF51 to 54). For this reason, investigations were conducted to suppress the maximum coupling loss for each core C1 to C4 to 0.15 dB or less.

[0044] Figure 12A is a graph corresponding to Figure 11C (df2=2.5mm) when lenses 41 to 44 are changed to lenses with a focal length df2=1.8mm. Curve L5 shows the relationship between Ω1=Ω2 and D1+D2, and curve L6 shows the relationship between Ω1=Ω2 and coupling loss when the inter-lens distance Z is the maximum distance sum D1+D2_max. Figure 12B is a graph corresponding to Figure 11C when lenses 41 to 44 are changed to lenses with a focal length df2=3.5mm. Curve L7 shows the relationship between Ω1=Ω2 and D1+D2, and curve L8 shows the relationship between Ω1=Ω2 and coupling loss when the inter-lens distance Z is the maximum distance sum D1+D2_max. According to Figures 12A, 11C, and 12B, the maximum distance sum D1+D2_max increases as the focal length df2 increases. On the other hand, as the focal length df2 decreases, the coupling loss becomes less prone to fluctuation even when the beam waist radius Ω1=Ω2 changes.

[0045] Figure 13 is a graph that defines the relationship between the normalized beam waist radius Ω1 = Ω2_norm and the normalized beam waist distance sum D1 + D2_norm, and the relationship between the normalized beam waist radius Ω1 = Ω2_norm and the coupling loss. Curves Ln5, Ln1, and Ln7 correspond to the normalized curves of curve L5 (see Figure 12A), curve L1 (see Figure 11C), and curve L7 (see Figure 12B), respectively. Curves Ln6, Ln4, and Ln8 correspond to the normalized curves of curve L6 (see Figure 12A), curve L4 (see Figure 11C), and curve L8 (see Figure 12B), respectively. Points P6, P4, and P8 are the intersections of the line representing the coupling loss = 0.15 dB with curves Ln6, Ln4, and Ln8, respectively. There is one more intersection point for each point, but these intersections do not affect the analysis and are therefore not considered.

[0046] The values ​​of the normalized beam waist radius Ω1=Ω2_norm were 0.752 at point P6, 0.759 at point P4, and 0.792 at point P8. When Ω1=Ω2_norm=0.752, the D1+D2_norm of curve Ln5 was 0.906, when Ω1=Ω2_norm=0.759, the D1+D2_norm of curve Ln1 was 0.905, and when Ω1=Ω2_norm=0.792, the D1+D2_norm of curve Ln7 was 0.913. This means that when the inter-lens distance Z is set to the maximum distance sum D1+D2_max, and a lens with a focal length df2 of 3.5 mm or less is used in the second lens group 40, the coupling loss can be suppressed to 0.15 dB or less if D1+D2_norm is 0.913 or greater. Based on this simulation, the inventors of the present invention found that if the inter-lens distance Z is substantially equal to the maximum distance sum D1+D2_max, the coupling loss can be suppressed to 0.15 dB or less by arranging the components 20, 30, 40, and 50 of the device 10 such that the beam waist distance sum D1+D2 is 91.5% or more of the maximum distance sum D1+D2_max.

[0047] Device 10 is designed based on the knowledge obtained in this way. With this configuration, even if manufacturing errors occur in distances d1 and / or d2 when Z ≈ D1 + D2_max is satisfied, the coupling loss can be suppressed to 0.15 dB or less if the beam waist distance sum D1 + D2 is 91.5% or more of the maximum distance sum D1 + D2_max. In addition, the behavior of the coupling loss when the beam waist diameter 2Ω1=2Ω2 changes becomes more gradual as Z increases (see Figures 11A to 11C). Therefore, by setting Z to be substantially equal to D1+D2_max, a device 10 with robust coupling loss to changes in beam waist diameter 2Ω1=2Ω2 can be realized compared to a configuration where Z is significantly shorter than D1+D2_max (see Figures 11A and 11B).

[0048] In particular, in this embodiment, when the focal length df2 = 2.5 mm, the distance between lenses Z is 80.89 mm, and the pitch p2 of SCF51 to 54 (see Figure 1) is approximately 3.1 mm. Pitch p2 is a major parameter that determines the external shape of the device 10, and it is desirable to miniaturize the device 10 by reducing the pitch p2. On the other hand, if the pitch p2 is made too small, problems such as interference between the ferrules 61 to 64 (or lenses 41 to 44) arise. For this reason, the target minimum value of pitch p2 is generally set to about 3 mm. Pitch p2 depends on the focal lengths df1, df2 and the distance between lenses Z. According to this embodiment, when df1 = 1.3 mm, df2 = 2.5 mm, and Z = 80.89 mm, the pitch p2 is approximately equal to the target minimum value. Therefore, in addition to the effect that coupling loss is less likely to fluctuate due to manufacturing errors, the effect of miniaturizing the device 10 (radially) can also be achieved.

[0049] The number of cores in the MCF is not limited to four; for example, it may be five or seven. If the MCF has five cores, the cores may be located at the vertices and center of a square. If the MCF has seven cores, the cores may be located at the vertices and center of a regular hexagon. This also applies to the modified examples, second embodiment, and third embodiment described later.

[0050] (modified version) Next, a modified FIFO device according to the present invention will be described. Figure 14 shows only the MCF20p and the first lens 30 of the FIFO device. This FIFO device differs from device 10 in that the MCF20 is replaced by the MCF20p. As shown in Figure 14, the MCF20p is obliquely polished so that the end face 20ap is inclined by a predetermined polishing angle α (8° in this example) in a predetermined inclination direction with respect to a plane (xy plane) perpendicular to its central axis. More specifically, the direction of oblique polishing of the MCF20p is the +y axis direction. Here, the direction of oblique polishing is the direction when viewed along the central axis of the MCF20p, from the distal end (farther away from the first lens 30) to the proximal end (closer to the first lens 30) of the end face 20ap which exhibits an elliptical shape due to oblique polishing.

[0051] The end face 20ap of the MCF20p is bevel-ground together with the end face 22ap of the ferrule 22p. By bevel-grinding the MCF20p, the reflected light caused by reflected light at the end face 20ap of the MCF20p is reduced. The number of cores and core arrangement of the MCF20p are the same as those of the MCF20.

[0052] When MCF20p is obliquely polished in the +y-axis direction, the principal rays of the light rays from each of the cores C1 to C4 (not shown) emitted from its end face 20ap (only the principal rays B2 and B3 are shown in this modified example) are inclined by a predetermined angle θ with respect to the axis in the yz plane. When the distances dt (in other words, the lengths of the principal rays B1 to B4 between each of the cores C1 to C4 and the first lens 30) in the traveling direction of the principal rays are defined as d1(1), d1(2), d1(3), and d1(4) respectively, the relationship d1(1) = d1(2) < d1(3) = d1(4) holds among these distances. That is, there are two variations (d1(1) = d1(2) or d1(3) = d1(4)) in the distance dt. Note that MCF20p is shifted in the -y-axis direction by a predetermined distance with respect to the optical axis (not shown) of the first lens 30. Thereby, the ray angles of the principal rays B1 to B4 of the light rays from each of the cores C1 to C4 emitted from the first lens 30 are made equal to each other.

[0053] The distance in the optical axis direction between the center of the end face 20ap of MCF20p and the first lens 30 is defined as the distance d1. In this case, the length d1(0) of the virtual line B0 extending from the center of the end face 20ap to the first lens 30 along the traveling direction of the principal ray is expressed as d1(0) = d1 / cosθ. Also, although the principal rays B2 and B3 and the virtual line B0 are not located on the same plane, since all of them exist on the yz plane, they can be regarded as being located on the same plane in the side view of MCF20p. Therefore, the distances d1(1) to d1(4) can be expressed as follows. d1(1) = d1(2) = d1(0) - Δda + Δdb d1(3) = d1(4) = d1(0) + Δda - Δdb Here, Δda is the distance in the direction of propagation of the principal ray from, for example, "core C3" to "the foot of the perpendicular drawn from the center of end face 20ap to the principal ray B3" in a side view of MCF20p, and can be expressed as Δda = p1sin(α+θ) / (2cosα). Δdb is the distance in the direction of propagation of the principal ray from, for example, "the intersection of virtual line B0 and the first lens 30" to "the foot of the perpendicular drawn from the intersection of principal ray B3 and the first lens 30 to virtual line B0" in a side view of MCF20p, and can be expressed as Δdb = (p1cos(α+θ)tanθ) / (2cosα).

[0054] Figure 15 is a graph that defines the relationship between distance d1 and the first beam waist radius Ω1 for each ray from cores C1 to C4 in the configuration shown in Figure 14. Curve L9 shows the behavior of rays from cores C1 and C2, and curve L10 shows the behavior of rays from cores C3 and C4. As shown in Figure 15, the value of the first beam waist radius Ω1 corresponding to an arbitrary distance d1 differs between rays from cores C1 and C2 (curve L9) and rays from cores C3 and C4 (curve L10). That is, two variations arise in the first beam waist radius Ω1 due to two variations in distance dt (although not shown in the figure, two variations also arise in the first beam waist distance D1). In this modified example, the MCF20p is positioned relative to the first lens 30 such that the length d1(0) (=d1 / cosθ) satisfies d1(0)>df1 (=1.3mm), and the SCF group 50 is positioned relative to the second lens group 40 such that the distance d2 satisfies d2>df2 (the same applies to the second and third embodiments). For this reason, the region d1≦1.3cosθmm is not considered in the graph of Figure 15. According to Figure 15, the Ω1 of the rays from cores C1 and C2 is greater than the Ω1 of the rays from cores C3 and C4. This is because the distance d1(1)=d1(2) is shorter than the distance d1(3)=d1(4).

[0055] Figure 16 is a graph that defines the relationship between the first beam waist radius Ω1 of rays from cores C1 and C2 (hereinafter also referred to as "Ω1(C1,C2)") and the first beam waist radius Ω1 of rays from cores C3 and C4 (hereinafter also referred to as "Ω1(C3,C4)") based on curves L9 and L10 in Figure 15. According to Figure 16, it can be seen that if one of the two variations of the first beam waist radius Ω1 is determined, the other radius Ω1 is uniquely determined.

[0056] In this modified example, SCF51 to 54 are arranged such that the second beam waist diameter 2Ω2 matches the corresponding first beam waist diameter 2Ω1 with respect to the corresponding lenses 41 to 44 (the same applies to the second and third embodiments). Therefore, condition 1 (see the first embodiment) is satisfied. For example, when Ω1(C1,C2) = 146 μm and Ω1(C3,C4) = 133 μm, the distance d2 between SCF51 and 52 and lenses 41 and 42 (see Figure 9) is adjusted so that the second beam waist radius Ω2 corresponding to SCF51 and 52 is Ω2 = 146 μm, and the distance d2 between SCF53 and 54 and lenses 43 and 44 is adjusted so that the second beam waist radius Ω2 corresponding to SCF53 and 54 is Ω2 = 133 μm. In other words, there are two variations in the second beam waist radius Ω2 due to the two variations in the distance dt.

[0057] Thus, when there are two variations in the beam waist radius Ω1=Ω2, there are also two variations in the beam waist distance sum D1+D2, and consequently, two variations in the coupling loss. To achieve low coupling loss, it is desirable to suppress the coupling loss to 0.15 dB or less in all cases. The inventors of this application have found that when the inter-lens distance Z is substantially equal to the maximum distance sum D1+D2_max, the coupling loss can be suppressed to 0.15 dB or less in both cases if the two beam waist distance sums D1+D2 are 91.5% or more of the maximum distance sum D1+D2_max.

[0058] Figure 17 plots points P11, P12, P11L, and P12L on a graph of coupling loss scaled from Figure 11C (i.e., Z=D1+D2_max=80.89mm). Point P11 shows the beam waist distance sum D1+D2 when Ω1(C1,C2)=146μm, and its value is 91.5% or more of the maximum distance sum D1+D2_max. Point P11L shows the coupling loss when Ω1(C1,C2)=146μm, and its value is 2.7×10⁻⁶ -3 The value is dB (≤0.15dB). On the other hand, point P12 shows the beam waist distance sum D1+D2 when Ω1(C3,C4)=133μm, and its value is equal to the maximum distance sum D1+D2_max. Point P12L shows the coupling loss when Ω1(C3,C4)=133μm, and its value is zero (≤0.15dB).

[0059] As shown in the example in Figure 17, in the case of Z = D1 + D2_max, it was confirmed that the coupling loss can be suppressed to 0.15 dB or less in both cases by adjusting the distances d1 and d2 so that both of the two beam waist distance sums D1 + D2 are 91.5% or more of the maximum distance sum D1 + D2_max. Based on the above, with the modified configuration, even if two variations occur in the beam waist radius Ω1=Ω2 due to oblique polishing of the MCF20p, by setting Z to be substantially equal to D1+D2_max and adjusting distances d1 and d2 so that D1+D2 is 91.5% or more of D1+D2_max, the coupling loss can be suppressed to 0.15 dB or less in both cases. Note that as long as D1+D2 is 91.5% or more of D1+D2_max, the combination of Ω1(C1,C2) and Ω1(C3,C4) is not limited to the above combination, and other combinations can be adopted based on Figure 16. By setting Z to be substantially equal to D1+D2_max, a device with robust coupling loss to changes in beam waist diameter 2Ω1=2Ω2 can be realized.

[0060] Furthermore, the above findings can also be applied when there are n variations in distance dt due to an increase in the number of cores or a change in core arrangement. That is, if the inter-lens distance Z is substantially equal to the maximum distance sum D1+D2_max, then if all n beam waist distance sums D1+D2 are 91.5% or more of the maximum distance sum D1+D2_max, the coupling loss can be suppressed to 0.15 dB or less in all cases.

[0061] (Second Embodiment) Next, a FIFO device according to a second embodiment of the present invention will be described. The FIFO device of this embodiment is identical to the modified FIFO device. However, the method for selecting the two combinations of beam waist radii Ω1=Ω2 differs from that of the modified embodiment. Hereinafter, the beam waist radius when the sum of beam waist distances D1+D2 is at its maximum value D1+D2_max will be defined as the "beam waist radius at maximum distance Ω_Dmax".

[0062] In this embodiment, when Z = D1 + D2_max, the distances d1 and d2 are adjusted so that Ω1(C1,C2) and Ω1(C3,C4) satisfy Ω1(C3,C4) < Ω_Dmax < Ω1(C1,C2). Figure 18 is a graph plotted on the graph of Figure 17 (i.e., Z = D1 + D2_max = 80.89 mm) with points P21, P22, P21L, and P22L plotted instead of points P11, P12, P11L, and P12L. Point P21 shows the beam waist distance sum D1 + D2 when Ω1(C1,C2) = 139 μm (see Figure 16). Point P21L shows the coupling loss when Ω1(C1,C2) = 139 μm, and its value is 1.2 × 10⁻⁶ -4 The value is dB. On the other hand, point P22 shows the beam waist distance sum D1 + D2 when Ω1(C3,C4) = 126 μm (see Figure 16). Point P22L shows the coupling loss when Ω1(C3,C4) = 126 μm, and its value is 2.0 × 10⁻⁶. -4 It is dB.

[0063] As shown in the example in Figure 18, in the case of Z = D1 + D2_max, it was confirmed that the coupling loss can be further reduced compared to the example in Figure 17 by adjusting the distances d1 and d2 such that the two first beam waist radii Ω1, Ω1(C1,C2) and Ω1(C3,C4), satisfy Ω1(C3,C4) < Ω_Dmax < Ω1(C1,C2). Based on the above, according to the configuration of the second embodiment, even if two variations occur in the beam waist radius Ω1=Ω2 due to oblique polishing of MCF20p, coupling loss can be further reduced by setting Z to be substantially equal to D1+D2_max and adjusting the distances d1 and d2 to satisfy Ω1(C3,C4)<Ω_Dmax<Ω1(C1,C2). In this embodiment, D1+D2 when Ω1(C1,C2)=139μm and D1+D2 when Ω1(C3,C4)=126μm are both 91.5% or more of the maximum distance sum D1+D2_max. However, as long as Ω1(C3,C4)<Ω_Dmax<Ω1(C1,C2) holds, D1+D2 corresponding to Ω1(C1,C2) and D1+D2 corresponding to Ω1(C3,C4) do not necessarily need to be 91.5% or more of the maximum distance sum D1+D2_max. Furthermore, as long as the above relationship holds, the combinations of Ω1(C1,C2) and Ω1(C3,C4) are not limited to the above combinations, and other combinations can be adopted based on Figure 16. By setting Z to be substantially equal to D1+D2_max, a device with robust coupling loss to changes in beam waist diameter 2Ω1=2Ω2 can be realized.

[0064] Furthermore, the above findings can also be applied when there are n variations in distance dt due to an increase in the number of cores or a change in the core arrangement. That is, if the maximum and minimum values ​​of the n first beam waist radii Ω1 are defined as "maximum beam waist radius Ωmax" and "minimum beam waist radius Ωmin," then when the inter-lens distance Z is substantially equal to the maximum sum of distances D1+D2_max, all n coupling losses can be suppressed by adjusting distances d1 and d2 such that Ωmin<Ω_Dmax<Ωmax holds. This is because, in the case of Z=D1+D2_max, when Ωmin<Ω_Dmax<Ωmax holds, the coupling loss at any Ω1 satisfying Ωmin<Ω1<Ωmax will always be smaller than the coupling loss at Ωmin and the coupling loss at Ωmax.

[0065] (Third embodiment) Next, a FIFO device according to the second embodiment of the present invention will be described. This embodiment differs from the second embodiment in that an MCFp (not shown) is used instead of an MCF20p, and there are two variations. The MCFp has four cores C11 to C14 arranged at the vertices of a square, and the core pitch p1 is 80 μm. That is, the core pitch p1 of the MCFp is larger than the core pitch p1 of the MCF20p (= 50 μm).

[0066] When the first beam waist radius Ω1 has two variations, the separation amount of these variations depends on the core pitch p1. For this reason, the separation amounts of the two variations of Ω1 in this embodiment are greater than the separation amounts of the two variations of Ω1 in the second embodiment. Figure 19 is a graph that defines the relationship between the first beam waist radius Ω1 (Ω1(C11,C12)) of the rays from cores C11 and C12 and the first beam waist radius Ω1 (Ω1(C13,C14)) of the rays from cores C13 and C14. According to Figure 19, the separation amount of Ω1 (ΔΩ1=143-122=21μm) is larger than the separation amount of Ω1 in Figure 16 (ΔΩ1=146-133=13μm).

[0067] In the second embodiment of the FIFO device, when the inter-lens distance Z is substantially equal to the maximum sum of distances D1+D2_max, the distances d1 and d2 are adjusted so that the maximum beam waist radius Ωmax and the minimum beam waist radius Ωmin satisfy the relationship Ωmin < Ω_Dmax < Ωmax (or, in the case of two variations, the relationship Ω1(C3,C4) < Ω_Dmax < Ω1(C1,C2)). This configuration is useful when the separation amount between Ωmax and Ωmin is relatively small, but as the separation amount increases, the coupling loss at Ωmax and the coupling loss at Ωmin increase, respectively, so it may not be possible to adequately suppress the coupling loss.

[0068] Figure 20 is a graph of a comparative example in which distances d1 and d2 were adjusted using the FIFO device of this embodiment (i.e., p1 = 80 μm) so that the relationship Ω1(C13, C14) < Ω_Dmax < Ω1(C11, C12) holds, and corresponds to Figure 18 (i.e., Z = D1 + D2_max). Point P31 shows the beam waist distance sum D1 + D2 when Ω1(C11, C12) = 143 μm (see Figure 19). Point P31L shows the coupling loss when Ω1(C11, C12) = 143 μm, and its value is 9.3 × 10⁻⁶. -4 The value is dB. On the other hand, point P32 shows the beam waist distance sum D1 + D2 when Ω1(C13,C14) = 122 μm (see Figure 19). Point P32L shows the coupling loss when Ω1(C13,C14) = 122 μm, and its value is 1.3 × 10⁻⁶. -3 The value is dB. As shown in Figure 20, the coupling loss has increased due to the increased separation amount of Ω1 compared to Figure 18. Therefore, there is a need to develop a technology that can appropriately suppress the coupling loss even when the separation amount of Ω1 increases due to an increase in core pitch p1.

[0069] Therefore, the inventors of the present application have found that the above problems can be solved by intentionally reducing the distance Z between the lenses from the maximum distance sum D1 + D2_max. This will be specifically described with reference to FIGS. 21A to 21C. FIGS. 21A to 21C are all graphs defining the relationship between the beam waist radius Ω1 = Ω2 and the beam waist distance sum D1 + D2 (see curve L1), and the relationship between the beam waist radius Ω1 = Ω2 and the coupling loss (see curves L11, L12, and L13). In FIG. 21A, the distance Z between the lenses is set to 80 mm, in FIG. 21B, the distance Z between the lenses is set to 79 mm, and in FIG. 21C, the distance Z between the lenses is set to 78 mm.

[0070] According to FIGS. 21A to 21C, it can be seen that the behavior of the coupling loss depends on the distance Z between the lenses. More specifically, when Z < D1 + D2_max (= 80.89 mm), curves L11 to L13 all have two minimum points where the coupling loss becomes zero and one maximum point located between these two minimum points. If the beam waist radii Ω1 = Ω2 of the two minimum points are defined as "first minimum point beam waist radius Ωmin1" and "second minimum point beam waist radius Ωmin2" in order from the larger one, the difference between Ωmin1 and Ωmin2 increases as Z decreases. Also, the maximum value increases as Z decreases.

[0071] As described above, since the separation amount ΔΩ1 (= Ω1(C11,C12) - Ω1(C13,C14)) of the first beam waist radius Ω1 having two variations depends on the core pitch p1, when the core pitch p1 is unchanged, the separation amount ΔΩ1 is also substantially constant. In the present embodiment, the distance Z between the lenses is set such that the difference between Ωmin1 and Ωmin2 is substantially equal to the separation amount ΔΩ1. In addition, the distance d1 is adjusted so that Ω1(C_{11},C_{12}) and Ω1(C_{13},C_{14}) (in other words, the maximum beam waist radius Ωmax and the minimum beam waist radius Ωmin when the number of variations is two) satisfy Ω1(C_{11},C_{12}) ≈ Ωmin1 and Ω1(C_{13},C_{14}) ≈ Ωmin2.

[0072] In the example of FIG. 21A, Z is set such that the difference between Ωmin1 and Ωmin2 is equal to the separation amount ΔΩ1. In addition, the distance d1 is adjusted so as to satisfy Ω1(C11, C12) ≈ Ωmin1 and Ω1(C13, C14) ≈ Ωmin2. The point P131L indicates the coupling loss when Ω1(C11, C12) = 143 μm, and the value is 6.6×10 -5 dB. On the other hand, the point P132L indicates the coupling loss when Ω1(C13, C14) = 122 μm, and the value is 3.4×10 -5 dB.

[0073] According to the example of FIG. 21A, it was confirmed that by setting (adjusting) Z and d1 as described above, the coupling loss can be significantly reduced as compared with the example of FIG. 20. In this embodiment as well, as long as the relational expressions of ΔΩ1 ≈ Ωmin1 - Ωmin2, Ω1(C11, C12) ≈ Ωmin1, and Ω1(C13, C14) ≈ Ωmin2 hold, D1 + D2 corresponding to Ω1(C11, C12) and D1 + D2 corresponding to Ω1(C13, C14) do not necessarily have to be 91.5% or more of the maximum distance sum D1 + D2_max.

[0074] On the other hand, in the example of FIG. 21B (Z = 79 mm), the coupling loss when Ω1(C11, C12) = 143 μm is 2.9×10 -4 dB (see point P231L), and the coupling loss when Ω1(C13, C14) = 122 μm was 8.2×10 -4 dB (see point P232L). Also, in the example of FIG. 21C (Z = 78 mm), the coupling loss when Ω1(C11, C12) = 143 μm is 1.8×10 -3 dB (see point P331L), and the coupling loss when Ω1(C13, C14) = 122 μm was 4.0×10 -3 dB (see point P332L). Thus, the coupling loss increases when Z decreases too much.

[0075] As described above, according to this embodiment, even when the separation amount of Ω1 increases due to an increase in core pitch p1, coupling loss can be appropriately suppressed by intentionally reducing the inter-lens distance Z and adjusting the distance d1. Furthermore, the behavior of coupling loss near the minimum point becomes gradual as Z increases (see Figures 11A to 11C and 21A to 21C). Therefore, a device with robust coupling loss to changes in beam waist diameter 2Ω1 = 2Ω2 can be realized.

[0076] The inventors of this application also performed similar simulations for a FIFO device equipped with MCF20p (i.e., p1 = 50 μm). For Ω1(C1,C2) and Ω1(C3,C4), combinations of Ω1(C1,C2) = 141 μm and Ω1(C3,C4) = 128 μm were adopted. First, as a comparative example, the coupling loss in the case of Z = D1 + D2_max (= 80.89 mm) was 3.8 × 10 when Ω1(C1,C2) = 141 μm. -4 The result is dB, and when Ω1(C3,C4) = 128 μm, it is 5.0 × 10⁻⁶. -5 The result is dB, and the average value is 2.2 × 10⁻⁶. -4 The result was dB. In contrast, the coupling loss when Z = 80.45 mm was 6.7 × 10 when Ω1(C1,C2) = 141 μm. -5 The result is dB, and when Ω1(C3,C4) = 128 μm, it is 4.6 × 10⁻⁶. -5 The result is dB, and the average value is 5.6 × 10⁻⁶. -5 The result was dB. The coupling loss when Z = 80.00 mm is 1.2 × 10 when Ω1(C1,C2) = 141 μm. -5 The result is dB, and when Ω1(C3,C4) = 128 μm, it is 4.4 × 10⁻⁶. -4 The result is dB, and the average value is 2.2 × 10⁻⁶. -4 The result was dB. Comparing the average values, it was found that the coupling loss could be reduced most effectively when Z = 80.45 mm.

[0077] Although the FIFO device according to the embodiments and modified versions has been described above, the present invention is not limited to the embodiments and modified versions described above, and various modifications are possible as long as they do not depart from the purpose of the present invention.

[0078] For example, instead of a group of single-mode single-core optical fibers, a group of single-core optical fibers comprising multiple single-core optical fibers that support multimode may be used. If the FIFO device is designed so that conditions 1 and 2 above are satisfied for the 0th-order mode ray, coupling loss can be reduced even when multimode rays propagate.

[0079] Furthermore, the core arrangement of the MCF does not need to be symmetrical. Even if the core arrangement is asymmetrical, the first lens 30 and the second lens group 40 can function appropriately as a FIFO device by positioning the second lens group 40 at positions corresponding to the light rays emitted from each core from the first lens 30.

[0080] Furthermore, SCF and MCF are not limited to a cylindrical shape, but may also be columnar in shape with a cross-section perpendicular to the axis of any shape (for example, an ellipse or a polygon).

[0081] Furthermore, the end faces 51a to 54a of SCF51 to 54 may be bevel-polished. In this case, in order to reduce optical coupling loss, the principal rays of the light incident on the end faces 51a to 54a must be tilted by a predetermined angle with respect to the optical axis of lenses 41 to 44. For this reason, unlike in Figure 4, the principal rays of the light incident on lenses 41 to 44 are designed to pass through a position shifted by a predetermined distance from the focal point f2.

[0082] Furthermore, the FIFO devices of the above embodiments and modified examples can also function as devices through which light rays propagate in a second direction. [Explanation of Symbols]

[0083] 10: FIFO device, 20: multicore optical fiber, 20a: end face, 30: first lens, 40: second lens group, 41, 42, 43, 44: second lens, 50: single-core optical fiber group, 51, 52, 53, 54: single-core optical fiber

Claims

1. A multicore optical fiber (20) comprising a plurality of columnar first cores (C1 to C4) extending along the axial direction, and a common cladding (CL) surrounding the plurality of first cores, A first lens (30) is provided corresponding to the multicore optical fiber, having a first optical axis parallel to the central axis of the multicore optical fiber (20), A second lens group (40) having a second optical axis parallel to the first optical axis and a plurality of second lenses (41 to 44) having a focal length different from that of the first lens, A single-core optical fiber group (50) comprising the same number of single-core optical fibers (51 to 54) as the second lenses (41 to 44), each having a columnar second core (C) extending along a central axis parallel to the second optical axis, and a cladding (CLs) surrounding the second core (C), In a fan-in / fan-out device (10) comprising a multicore optical fiber (20), the light rays are configured to propagate in either a first propagation direction, which is the direction in which light rays are emitted from each of the first cores (C1 to C4) of the multicore optical fiber (20), pass through the first lens (30) and the second lenses (41 to 44) corresponding to each of the first cores, and converge on the second core (C) of the single-core optical fiber (51 to 54) corresponding to the second lens, and a second propagation direction, which is the direction in which light rays are emitted from each of the second cores, pass through the corresponding second lens and the first lens, and converge on each of the first cores corresponding to the second lens, Assuming that the light rays propagate in the first direction of propagation, the beam waist diameter of each light ray emitted from the first lens (30) is defined as the first beam waist diameter (2Ω1), and the distance from the first lens to the beam waist position in the direction of propagation of the principal light ray of each light ray is defined as the first beam waist distance (D1). Assuming that the light ray propagates in the second direction of propagation, the beam waist diameter of the light ray emitted from each of the second lenses (41 to 44) is defined as the second beam waist diameter (2Ω2), and the distance from the second lens to the beam waist position in the direction of propagation of the principal ray of the light ray is defined as the second beam waist distance (D2). If the inter-lens distance (Z), which is the distance between the first lens (30) and each of the second lenses (41 to 44) in the direction of propagation of the principal ray of the light, is equal to the sum of beam waist distances (D1 + D2), which is the sum of the first beam waist distance (D1) and the second beam waist distance (D2) when the first beam waist diameter (2Ω1) and the second beam waist diameter (2Ω2) coincide, then the maximum value of the sum of beam waist distances is defined as the maximum value of the sum of beam waist distances (D1 + D2_max), The multicore optical fiber (20), the first lens (30), the second lens group (40), and the single-core optical fiber group (50) are arranged such that the inter-lens distance (Z) is substantially equal to the maximum value of the distance sum (D1 + D2_max), and the beam waist distance sum (D1 + D2) is 91.5% or more of the maximum value of the distance sum. Fan-in / fan-out device.

2. In the fan-in / fan-out device according to claim 1, The end face (20ap) of the multicore optical fiber (20p) is obliquely polished so as to be inclined by a predetermined polishing angle in a predetermined inclination direction with respect to a plane perpendicular to its central axis, thereby creating n variations in the distance (d1(1) to d1(4)) between each of the first cores (C1 to C4) and the first lens (30) in the direction of propagation of the principal ray of the light, and also creating n variations in the first beam waist diameter (2Ω1) of the light ray from each of the first cores. Each of the single-core optical fibers (51 to 54) is positioned relative to the corresponding second lens (41 to 44) such that the second beam waist diameter (2Ω2) of the light ray corresponding to each of the single-core optical fibers coincides with the corresponding first beam waist diameter (2Ω1). The distance between the lenses (Z) is substantially equal to the maximum value of the distance sum (D1 + D2_max), and the n beam waist distance sums (D1 + D2) corresponding to the n first and second beam waist diameters (2Ω1 = 2Ω2) are all 91.5% or more of the maximum value of the distance sum (D1 + D2_max). Fan-in / fan-out device.

3. A multicore optical fiber (20p) comprising a plurality of columnar first cores (C1 to C4) extending along the axial direction, and a common cladding (CL) surrounding the plurality of first cores, A first lens (30) is provided corresponding to the multicore optical fiber, having a first optical axis parallel to the central axis of the multicore optical fiber (20p), A second lens group (40) having a second optical axis parallel to the first optical axis and a plurality of second lenses (41 to 44) having a focal length different from the focal length of the first lens, A single-core optical fiber group (50) comprising the same number of single-core optical fibers (51 to 54) as the second lenses (41 to 44), each having a columnar second core (C) extending along a central axis parallel to the second optical axis, and a cladding (CLs) surrounding the second core (C), In a fan-in / fan-out device (10) comprising a multicore optical fiber (20p), the light rays are configured to propagate in either a first propagation direction, which is the direction in which light rays are emitted from each of the first cores (C1 to C4) of the multicore optical fiber (20p), pass through the first lens (30) and the second lenses (41 to 44) corresponding to each of the first cores, and converge on the second core (C) of the single-core optical fiber (51 to 54) corresponding to the second lens, and a second propagation direction, which is the direction in which light rays are emitted from each of the second cores, pass through the corresponding second lens and the first lens, and converge on each of the first cores corresponding to the second lens, Assuming that the light rays propagate in the first direction of propagation, the beam waist diameter of each light ray emitted from the first lens (30) is defined as the first beam waist diameter (2Ω1), and the distance from the first lens to the beam waist position in the direction of propagation of the principal light ray of each light ray is defined as the first beam waist distance (D1). Assuming that the light ray propagates in the second direction of propagation, the beam waist diameter of the light ray emitted from each of the second lenses (41 to 44) is defined as the second beam waist diameter (2Ω2), and the distance from the second lens to the beam waist position in the direction of propagation of the principal ray of the light ray is defined as the second beam waist distance (D2). If the inter-lens distance (Z), which is the distance between the first lens (30) and each of the second lenses (41 to 44) in the direction of propagation of the principal ray of the light ray, is equal to the sum of beam waist distances (D1 + D2), which is the sum of the first beam waist distance (D1) and the second beam waist distance (D2) when the first beam waist diameter (2Ω1) and the second beam waist diameter (2Ω2) coincide, then the maximum value of the sum of beam waist distances is defined as the maximum value of the sum of beam waist distances (D1 + D2_max), The end face (20ap) of the multicore optical fiber (20p) is obliquely polished so as to be inclined by a predetermined polishing angle in a predetermined inclination direction with respect to a plane perpendicular to its central axis, thereby creating n variations in the distance (d1(1) to d1(4)) between each of the first cores (C1 to C4) and the first lens (30) in the direction of light propagation, and also creating n variations in the first beam waist diameter (2Ω1) of the light beam from each of the first cores. Each of the single-core optical fibers (51 to 54) is positioned relative to the corresponding second lens (41 to 44) such that the second beam waist diameter (2Ω2) of the light ray corresponding to each of the single-core optical fibers coincides with the corresponding first beam waist diameter (2Ω1). Furthermore, if we define the first and second beam waist diameters (2Ω1 = 2Ω2) when the sum of beam waist distances (D1 + D2) is the maximum value of the sum of distances (D1 + D2_max) as the beam waist diameter at maximum distance (Ω_Dmax), and define the maximum and minimum values ​​of the n possible first and second beam waist diameters as the maximum beam waist diameter (2Ωmax) and the minimum beam waist diameter (2Ωmin), respectively, The multicore optical fiber (20p) is positioned relative to the first lens (30) such that the maximum beam waist diameter (2Ωmax) is greater than the beam waist diameter at maximum distance (Ω_Dmax), and the minimum beam waist diameter (2Ωmin) is smaller than the beam waist diameter at maximum distance. Fan-in / fan-out device.

4. In the fan-in / fan-out device according to claim 3, The distance between the lenses (Z) is substantially equal to the maximum value of the distance sum (D1 + D2_max). Fan-in / fan-out device.

5. In the fan-in / fan-out device according to claim 3, The curve defining the relationship between the first and second beam waist diameters (2Ω1 = 2Ω2) and the optical coupling loss has two local minimums where the optical coupling loss is zero when the inter-lens distance (Z) is less than the maximum value of the distance sum (D1 + D2_max), and one local maximum located between the two local minimums. If the first and second beam waist diameters of the two minimum points are defined as the first minimum beam waist diameter (2Ωmin1) and the second minimum beam waist diameter (2Ωmin2) in descending order, then the difference between the first minimum beam waist diameter and the second minimum beam waist diameter increases as the inter-lens distance (Z) decreases. If n = 2, The inter-lens distance (Z) is set such that the difference is substantially equal to the separation amount between the maximum beam waist diameter (2Ωmax) and the minimum beam waist diameter (2Ωmin), and The multicore optical fiber (20p) is positioned relative to the first lens (30) such that the maximum beam waist diameter (2Ωmax) is approximately equal to the minimum point first beam waist diameter (2Ωmin1), and the minimum beam waist diameter (2Ωmin) is approximately equal to the minimum point second beam waist diameter (2Ωmin2). Fan-in / fan-out device.

6. In the fan-in / fan-out device according to Claim 1, The distance (d1) between the end face (20a) of the multicore optical fiber (20) and the first lens (30) in the optical axis direction of the first lens (30) is greater than the focal length (df1) of the first lens (30). The distance (d2) between the end faces (51a to 54a) of the single-core optical fibers (51 to 54) in the optical axis direction of the second lens (41 to 44) and the second lens (41 to 44) is greater than the focal length (fd2) of the second lens (41 to 44). Fan-in / fan-out device.

Images