JPS62280812A - Optical branching and joining device - Google Patents

Abstract

PURPOSE:To easily add and remove the titled device, even after an optical fiber system has been constructed, by making two pieces of optical fibers contact each other, and bringing both the optical fibers to a snake motion plural times in a place in which axial cores of two optical fibers are contained. CONSTITUTION:In an optical branching and joining device 1, two pieces of optical fibers A, B contact each other closely in a state of flapping, curved alternately and contact each other. Since such a snake motion curve is given, two holding fixtures 3, 4 hold the side faces of the optical fibers A, B. Holding surfaces become uneven surfaces 5, 6, and matching oil 7 is injected so as to fill an air gap between the uneven surfaces 5, 6 of the holding fixtures and the optical fibers A, B. When the optical fiber through which light beams being propagated are main and the optical fiber for making the light beams branch thereafter is subordinate, if they are projected and recessed in the direction of the main fiber and in the direction of the sub-fiber, respectively, by curving large a snake motion line, a larger radiated light beams are emitted from the projecting part of the main fiber, and the recessed part of the corresponding sub-fiber can make a larger radiated light beams incident.

Description

Detailed Description of the Invention 3. Detailed Description of the Invention (1) Technical Field The present invention relates to a branching/combining device that can input and output light from the middle of an optical fiber, which is an optical transmission line.

When performing optical communication using an optical fiber, there is a strong demand for an optical branch that is provided in the middle of the optical fiber and can directly extract signals.

(a) Prior Art The following types of optical branching/merging devices for branching and merging light propagating through optical fibers have conventionally been available.

FIG. 4 shows an example of a conventionally used optical branching/merging device. This uses a beam splitter. The light emitted from the optical fiber 20 is made into parallel light by a lens 23, and separated into reflected light and transmitted light by a beam splitter 26. This is condensed by lenses 24 and 25, and connected to the other two optical fibers 2L.
22.

This emits light once from an optical fiber into free space,
After splitting the light, the light is re-injected into the optical fiber.

This can divide the light into approximately equal amounts, and since it uses a lens to focus the light, there is relatively little loss. the current,
Most of the optical branching methods in practical use are of this type.

However, this branch has a drawback in that it is not easy to add or remove branching or merging sections. This is because the optical fiber will be cut. Another drawback was that it required alignment of the lens.

The second optical branching/merging device uses a fiber coupler as shown in FIG. This is two optical fibers fused together.

Light entering from the input end 32 of one optical fiber comes to exit at two output ends 33 and 34.

However, this method uses a special optical component called an optical fiber coupler. It is not easy to add optical branches after the system is completed and then remove them later.

Light branching and merging sections that can be added or removed are strongly desired.

Although it is not a fiber-to-fiber coupling, an experiment has been attempted in which a portion of the propagating light is extracted from an optical fiber and the light intensity is detected using a light receiving element.

For example, C, Stewart & W, J, Stewart
1Directional couplerf'or
single mul, timode optical
fiber EC0C267 (1976) is a method in which an optical fiber is placed on a transparent body having periodic irregularities, and by pressing the optical fiber, the optical fiber is periodically bent in a direction perpendicular to the plane. As the optical fiber meanderes, light is emitted from the protrusion. The emitted light propagates diagonally through the transparent body, but is condensed by a concave mirror formed at the end of the transparent body and enters the light receiving element. The light receiving element detects the intensity of these lights. This is proportional to the intensity of the light propagating through the optical fiber.

Stuart calls the unevenness of a transparent body a grating. When an optical fiber is pressed against this, it will meander with the same period as the grating. In this way,
Conversion from propagation mode to radiation mode occurs on the outside of the meander, that is, at the convex portion, and radiation mode light exits the optical fiber and enters the transparent body.

Let A be the pitch, or period, of the grating. It is said that the light passing through the grating undergoes mode conversion and is converted into light with a larger mode, and that the difference between the phase constants β1 and β2 of the two modes satisfies the condition for the period of the grating. .

Then, a grating with A=1.7WIR was made from a transparent body, and light leaked into the transparent body.
Light intensity can be detected by a light receiving element fixed to a transparent body. It is recommended that a soft rubber plate be used to hold down the optical fiber from above.

A grating can be defined by a period A and an amplitude.

It is stated that in an example where the amplitude is 3.5 μI, A=1, and 7 m, 5.6% of the light incident on the optical fiber can be extracted and captured by the light receiving element.

Furthermore, Stewart et al.
1Directional Couplersfor
optical f'1bre communica
tion systems “EC0C170 (1977
) also proposed the same optical branching coupler. The same technique is used to make a fiber meander and allow light to leak into a transparent body. Stewart also says that by bending such a fiber to a larger radius of curvature, light is emitted only in the convex portion, which increases the efficiency of the split.

Figure 6 shows Stuart's optical branching coupler.

Such a coupler can be attached directly to an optical fiber. There is no need to cut the optical fiber. -! It can also be easily removed. In other words, they can be added and removed later from the main optical fiber. Stuart does not provide a splitter that can split the light directly. It is necessary to receive the light once with a light receiving element.

If it is necessary to further transmit the branched light, it is necessary to amplify the signal of the light receiving element, drive a light emitting diode with the amplified signal, and pass this light to another optical fiber.

Therefore, a power source, an electronic circuit, a light emitting diode, etc. are required at the branch.

Stuart emphasizes periodic curvature. Because it is periodic, the part of his coupler that deforms the optical fiber is called a grating.

Bruce D, Campball 1JAWS-A
coupler for computer rin
g “, 5PIE 479e86 (1984)
proposed an optical branching coupler that extracts a radiation mode by a single bend. Blues has a radius of 5ffll
11 curves are made, and a light-receiving element is provided here to collect the light in the radiation mode among the light propagating through the optical fiber.

This coupler also has the disadvantage that a light receiving element, a power source for driving the light receiving element, an electronic circuit, etc. must be provided at each branch.

sec) Problems with conventional technology So far, we have already pointed out the problems of conventional optical splitters.

Since the present invention has common features with the optical splitter coupler of Stewart et al., Stuart's coupler will be described again.

Stewart emphasizes periodicity, and therefore uses gratings to periodically bend the fiber.

Periodicity is necessary because the selection side of equation (1) is imposed. The waves up to the μth mode are waveguide modes, and μ+
Assume that the first mode is the radiation mode. The choice is that the difference between the phase constants β1 and β2 must be 2π/A.

However, the selection side of equation (1) is incorrect. The inventor thinks as follows.

Assume that the μth mode is converted into the μ+1th mode and this is radiated from the fiber. Then, since the μth mode disappears, a conversion from the (μm1)th mode to the μth mode must occur. That is, the (j-1)-5th mode conversion must occur in sequence. For this to be possible, a relationship of phase constants is required. In other words, the phase constant is required to be linear with respect to the mode order.

1 end, I,) lead? , Togi ha・
100% water is served on Saturday, Sunday and Taiwan! Kanade can also be said to be the group velocity in the propagation direction, that is, the Z direction, and although it can be approximated by a quadratic function of the mode order, it is not a linear function. The interval between modes of the phase constant β is not constant, and increases as the mode order increases.

Another problem with equation (1) is that the curvature of the fiber, or the radius of curvature R, is supposed to be an important parameter;
This means that it does not appear in equation (1) at all. Which lower-order mode is converted to which higher-order mode? This means that it must depend on the radius of curvature H rather than the period of the grating.

For this reason, the inventor considers Stewart's theory questionable.

The inventor believes that the condition "periodic" is meaningless. It is easy to create something that is periodic, but it can also be aperiodic.

00 Purpose Easily add to the fiber optic system even after building it.
A first object of the present invention is to provide an optical branching/combining device that can be removed.

Enables direct fiber-to-fiber coupling,
A second object of the present invention is to provide an optical branching/combining device that does not require a light receiving element, a power source, a light emitting element, an electronic circuit, etc. at the branching part.

A third object of the present invention is to provide an optical branching/merging device that does not use optical elements such as lenses or mirrors, and therefore does not require delicate adjustments such as axis alignment.

(3) Configuration The optical branching/merging device of the present invention brings two optical fibers into contact with each other on the sides, and meanderes both optical fibers multiple times within a plane that includes the axes of the two optical fibers. It is something. That is, two optical fibers are bent multiple times in parallel, and the emitted light from one optical fiber is made to enter the other optical fiber.

Since the present invention couples optical fibers using conversion into radiation mode by microbending, the optical fibers must be multimode fibers.

Each optical fiber includes a core, a cladding, and a primary coating.

For example, a quartz glass fiber with a core diameter of 50 μm1
A typical example is one with a cladding diameter of 125 μm.

- Silicone, urethane, epoxy resin, etc. are used for the second coating. There is also an optical fiber with a core diameter of 60 μm, 180 μm, 1100 μm and a cladding diameter of 150 μm. These are graded index optical fibers.

The present invention can also be applied to graded index optical fibers.

Although the diameter of the primary coating is 300 μm to 400 μm, since the two optical fibers are brought into contact in parallel, the distance between the core centers is about this length.

FIG. 1 is a longitudinal sectional view of the optical branching/merging device of the present invention. Second
The figure is a sectional view taken along line II in FIG.

There are two optical fibers A1B. These have both free ends, but in the optical branching/merging device 1 of the present invention,
They are in close contact while undulating.

They touch each other while undulating, that is, curving alternately. This wave may or may not have periodicity.

In order to give such a meandering curve, two presser metal fittings 3 are used.
．． 4 presses the side surface of the optical fiber A1B.

The holding fitting 3.4 holds the optical fibers A,
This is to give B a micropend.

Therefore, the holding surface of the holding fitting 3.4 has an uneven surface 5.6.

Matching oil 7 is injected to fill the gaps between the uneven surface 5.6 of the presser fitting and the optical fibers A and B. If this part is air, it will be difficult for it to escape from the optical fiber and enter the optical fiber, so it is filled with matching oil 7 whose refractive index is close to, but slightly larger than, the refractive index of the optical fiber cladding. be done.

The thickness of the holding fitting 3.4 may be approximately equal to the outer diameter of the optical fibers A and B. Further, a parallel holding plate 8.9 holds the holding plate 3.4 and the fiber A1B, and the holding plate 3.4 may be provided so as to be movable relative to the holding plate 8.9. In this case, the coupling depth of fiber A1B can be adjusted as desired.

The presser fitting 3.4 may be fixed after being positioned once.

The traveling directions of the optical fibers A and B inside the optical branching/merging device are two directions, the direction perpendicular to the plane is the X direction, and the direction in which the optical fibers come into contact and perpendicular to the Z axis is the X direction.

In this example, the meandering directions of the optical fibers A and B are straight lines about the two axes, but this does not mean that they have to be straight lines.

The optical fiber through which light is propagating is called the main fiber.
From now on, the optical fiber to which light is to be branched will be referred to as a slave fiber. In this case, the meandering line may be largely curved so that it is concave in the direction of the main fiber and convex in the direction of the secondary fiber.

By doing so, a larger amount of radiated light can be emitted from the convex portion of the main fiber, and a larger amount of radiated light can be input into the corresponding concave portion of the secondary fiber.

There is no requirement that two optical fibers be brought into contact. It is also possible to separate the two optical fibers. FIG. 3 shows such an example. In this way, by separating the two optical fibers A1B, the curvature of both can be set freely.

The curve here refers to the curve of the meandering line. The main optical fiber A can be made convex toward the restoration fiber B, and the restoration fiber B can be made convex toward the main optical fiber A. In other words, a relationship similar to that of a concave lens can be created.

Since the two optical fibers are not in contact, a transparent guide plate 10 is provided in the middle to fill the gap. An example of this would be an acrylic board.

Since the light emitted from the optical fiber A exits obliquely, the transparent guide plate 10 faces in an oblique direction.

Also on the sides of the transparent guide plate 10 that correspond to the optical fibers A and B,
A textured surface 11.12 is provided. These uneven surfaces 1
1.12 and the uneven surface 5.6 of the presser fitting 5.6, the optical fibers A and B are meandered independently.

Filling the gap between the optical fiber A1B, the uneven surface, and the holding plate with the matching oil 7 is the same as in the previous example.

a) Production Light in a propagation mode is propagating in the main optical fiber.

Since it is a multimode optical fiber, many modes are excited.

The mode within the fiber is defined by how many nodes are present in the lateral direction. Higher order modes have small group velocities. This corresponds to light rays that are frequently reflected at the core-cladding boundary. This will propagate if the optical fiber is straight, but if the optical fiber is bent, the total internal reflection condition will no longer be satisfied and the light will protrude outside the core.

Such a mode is called a radiation mode. This can be explained in terms of wave optics by saying that the higher-order propagation mode has become a radiation mode.

However, since it has a large number of modes, it is more suitable for handling in terms of geometric optics than in wave optics.

This means that the light rays have exited from the outwardly convex and curved portion. This happens because the curvature η is not O'''C.

The radius of curvature R is the reciprocal of the curvature η.

Normally, the sign of curvature is if it is convex downward in the coordinate system.
This is considered positive, and if it is concave downward, it is considered negative.

However, since this does not distinguish the unevenness of the meandering fiber, here, the normal line is set from the core to the cladding, and if the normal line is convex in the normal direction, η and R are assumed to be positive. If it is concave in the normal direction, η and R are negative.

Defining this way, it can be said that a portion where the curve is convex outward has a positive curvature η. At this point, light leaks from the optical fiber and exits to the outside.

The fact that it leaks to the outside means that the guided mode (propagation morton has become a radiation mode) Z TJ
X, l -a V L +, two beaks, one oil A L Ik 0111 J= IL is difficult to leak, so here it is said to go out, emit, or leak.

As long as η〉O, light will definitely leak from this curve. The point where the largest amount of light leaks is the point where the curvature η is maximum.

In which direction does the light come out? However, first of all, it is the tangential direction. If the curvature is small, light will only emit light at a slight divergence angle from the tangential direction.

FIG. 7 is an explanatory diagram showing such a thing. C11C3
has a small curvature, so there is little spread. The amount of light is also low.

If the curvature is large, light will be emitted from the tangential direction over a fairly wide spread angle. 02 points of wide emitted light are shown.

Next, consider the conditions under which light enters the optical fiber from the outside.

It would seem that light would efficiently enter the parts where the curvature η is negative, that is, the four parts, but this is not the case. When you shine light on the four parts diagonally, once it enters the core, you can see the ten major steps;
srb = 14 inches at the end of the path - In order for the light to enter the optical fiber, the curvature η must be positive.

In other words, at the location of the convex portion, both the output and the input are performed efficiently.

FIG. 8 illustrates this. Light enters efficiently where the curvature is greatest.

What is important is not only that it is incident, but also that it does not come out of the core again.

The present invention focuses on the input into the optical fiber rather than the output from the optical fiber.

I will discuss this in detail later.

There is a principle like this.

In FIG. 1, when light is passed through optical fiber A, the light exits in a substantially tangential direction at the convex portion. A portion of this light enters the optical fiber B at the convex portion of the optical fiber B. This happens at each point of the curve. Therefore, light energy gradually transfers from optical fiber A to optical fiber B.

In the transition from optical fiber A to optical fiber B, the length of the optical path is not constant. However, this is not a problem since it is multi-mode.

B) Calculation example Let's calculate the number of modes for one common optical fiber. Luto with quartz glass fiber. Core diameter is 5
0μm1 cladding diameter, 125μm1 core refractive index 1.
45, the numerical aperture NA is 0.2. λ=0.83μm
shall be.

The normalized frequency V is v-2πn1ζψ (3). a is the core radius, n is the core refractive index, and Δ is the refractive index difference ratio (nl-n2)/"1. Δ = 0.048, 'I = 85 (4)
The number of modes can be approximated by V/2, which is approximately 360
0 mode.

Since such a large number of modes exist in the optical fiber, geometrical optics approximation is well established. This has an NA of 0.2, which is quite large, but the NA is 0.2.
Even if it is 1, the number of modes is about 1800.

(ri) Output from optical fiber FIG. 9 is an explanatory diagram showing a curved portion of an optical fiber.

EAC indicates the core-cladding boundary. The FBD is also a core-cladding boundary. The part surrounded by EAC and FBD is the core. Straight line, AC in EAlFB.

Assume that the BD is curved.

According to the previous definition, AC has a positive curvature and BD has a negative curvature.

Let the center of bending be O.

Let G be the intersection of the extension of the extension line FB and the outer arc AC.

The problem is the opposite angle of the light beam at the curve AC.

Let θ be the total reflection angle in the core and cladding.

This is determined by the core refractive index n1 and the clad refractive index n2. To~Ω　-tri1 It can be defined by

Suppose that a certain ray is reflected at a point P between AEs at an angle θ with respect to the boundary line, travels through the core, and is reflected or refracted at Q in AC. The problem is the angle that ray PQ makes with tangent QS at point Q. If this is smaller than θ, total reflection will occur. If this angle is larger than e, most of the core will come out.

The angle ψ formed between the tangent QS and the straight line FA is naturally equal to the central angle AOQ. When this is determined, the angle (ψ-θ) that the ray PQ makes with the tangent Sq can be found.

With A as the origin, take the X coordinate toward the left and the Y coordinate toward the bottom. Let AP=x. The linear equation representing straight line PQ is y=-10(x-x) (6) Arc A
The formula expressing Q, GC is X +
(Y-R) = R” (7).If we combine (6) and (7) and solve this, we get
The x1Y coordinates of point Q are determined. However, X is negative. This value and ψ have the relationship R51nψ=-X (8). Therefore. (9) is a strict expression.

This is difficult to handle as it is, and assuming that θ is a sufficiently small value, the reflection angle/5QP=(ψ−Q) can be approximated as follows. The condition that it exits the fiber core at point Q is
This is larger than θ.

The emission condition is given by θ × ψ−0 (11). The critical angle θm is defined as: If it is 0m, this ray will be reflected and remain in the core. If 0 m x θ x θ, this light is emitted.

A similar calculation is performed for the light reflected at angle 0 from point U on the inner side FB. With a similar approximation, the critical angle On is defined as . D is the core diameter. If θ<On, it is reflected and stays in the core (at point W), and Onku0kue
If so, go outside the core.

It is clear that it is easier to emit light from point U on the inner side FB.

FIG. 10 is a graph showing the conditions under which light reflected at an angle θ at a point P on the outer edge AE is emitted at a curve AC. Xθ
=D means that straight line PQ does not hit BP' but curve AC,
This is the condition. Since it is originally a propagating light, θ is always smaller than the total reflection angle e. Further, θ must be larger than 0m defined by equation (12).

Even though it is said that the radiation is emitted here, it does not mean that all of it is emitted, but only a part of it is emitted. The shaded portion (X, θ) provides the conditions for emission.

Even in this region, the ratio of output power is small as long as θ is close to Om2, and as θ becomes larger, the ratio of output power increases.

FIG. 11 shows the conditions under which light reflected at an angle θ at a point U on the inner side BF' hits a curve AC and is emitted. The part surrounded by θ, On, and D/x indicates the emission condition.

i) Injection into optical fiber The most important feature of the present invention is that light is input obliquely into the optical fiber from the outside.

In Stuart's experiment, the light was emitted from an optical fiber.
This is what is being done. There are also some questionable points,
Emission condition (1) and the like have already been proposed.

However, the emphasis of the invention is not on the output, but on the input.

It is well known that light leaks from an optical fiber due to microbending. Other measurements using all of these methods are also in practical use.

In the present invention, light is made to enter an optical fiber by microbending. Injecting it is much more difficult than emitting it.

Even if light is shined into the optical fiber from an angle, almost no light will enter the inside. Even if you enter temporarily, you will leave immediately.

Light is input obliquely into a straight fiber and is denoted by k.

Unless there is a defect in the core that scatters this light,
It completely leaves the core. The fact that the light entered from an oblique direction means that it has an inclination angle with respect to the core surface that is larger than the total reflection angle θ. Such light is reflected at the core-cladding boundary several times, but quickly disappears. This is because if it is a step-index fiber and there are no defects, the reflection angle will be conserved and will not become smaller than e.

In reality, there are defects in the optical fiber that cause scattering, so a small portion of the light incident obliquely becomes propagated light.

However, in the present invention, the use of microbending to allow light to enter does not rely on such defects.

Even if the optical fiber is defect-free, propagating light can be incident obliquely into the optical fiber. The novelty of the present invention lies in this idea.

This does not mean that the present invention can only be applied to defect-free optical fibers, nor does it assume that defect-free optical fibers can actually be manufactured.

FIG. 12 is a diagram for considering the incidence of light into the optical fiber core. HLJ is the outer edge of the core, and KIM is the inner edge of the core. Core refractive index is nl, cladding refractive index is n2
It is.

Assume that light enters the core from the cladding at point H. Let the angle between the incident light and the core surface be φ. Actually, it is the angle between the tangent and the ray, but for the sake of simplicity, the illustration of the tangent is omitted.

The terms incidence, refraction, and reflection angles will be used below, but unlike ordinary optics, these are angles formed not with respect to the center line (normal) but with respect to a surface.

Let the refraction angle be φ'. This is larger than the total reflection angle θ.

Also, n2cosφ W n1cosφ'
(14). Suppose that the light travels straight through the core and is reflected at one point.

At this time, let the angle between it and the boundary surface be ψ. In order to be totally reflected here, this must be smaller than e.

If KI is parallel to HL, φ'=ψ, so it is impossible for φ' to be larger than θ and ψ smaller than e.

e 〉 φ' (15) ψ 〉 θ
(16), KI,
It must be concavely curved downward between HL.

In other words, the condition for incidence is that the curvature η is positive, and the reason why it is shown in FIG. 8 is because of the conditions (15) and (16).

The optical fiber is required to have a positive curvature because it enters from the outside and is totally reflected at one point inside.

But that's not enough.

There is a problem as to what happens to the total reflection at one point forming an angle ψ at a point J on the outer edge of the core-cladding boundary.

Determine the angle that the ray makes with the boundary at five points.

If ζ=φ', most of this light will be light in the v direction passing through to the cladding. The angle is equal to the initial angle of incidence φ.

Of course, some of it is reflected at one point and goes in the N direction, but this is small.

Since it entered from point H at an angle φ, it is natural that it exits from point 5 at an angle φ.

In order to prevent this from happening, ζ must be smaller than φ'.

ζ ku φ' (17
) More preferably, ζ is smaller than the total reflection angle θ.

ζ ku θ (18
) For this to be the case, the radius of curvature Z between IM must be longer than the radius of curvature R between KI.

In other words, the center 0' of the arc of IM and LJ is KI.

It must be farther away than the center 0 of the arc of HL.

Simply put, this means that the curvature η must decrease in order for the light entering from the incident point H to be confined within the core.

The initial curvature R is limited by the angle of incidence φ.

Conversely, it can also be said that φ depends on the curvature R. Furthermore, 2
The curvature Z up to the second reflection point is limited by φ.

OK,,,01=R,0f(=OL=R+D.

D is the core diameter.

From the law of sine, in ΔOHI, this, fish θ-n2/n1, and equation (16) (14
) Formula % Formula % (20) This can also be considered as a restriction on the incident angle φ.

φ is an angle close to O, but the smaller the radius of curvature R,
In other words, the larger the curvature, the wider the allowable range of φ.

In other words, it is easy to enter. The inequality (21) is a condition that total reflection occurs at one point.

Furthermore, consider the condition ζ × θ that total reflection occurs at three points. 80' From the sine theorem for IJ, O'I=0'
Let M=Z, O'L-0'J=Z+D, and from these, obtain. From (23), it follows that Z)R.

Such curvature changes occur between KIs or IMs.

This length can be approximately evaluated by D/φ'.

Curvature can be written as (1/R) and (1/Z).

The derivative of the curvature along the optical fiber is given by.

The curvature must be reduced. Its speed is determined by the angle of incidence φ
The larger the fiber diameter, the faster it must be, and the smaller the fiber diameter, the faster it must be.

In FIG. 8, the condition of equation (24) is satisfied only in the vicinity of points D2 and D3. Since the curvature increases between D1 and D2, the condition (24) cannot be satisfied.

For example, suppose that an optical fiber is deformed to have an amplitude of A and a pitch of A.

2π f'(x) = As1n(-x) (
25). The curvature η is obtained by second-order differentiation. If the slope f'(x) can be ignored relative to 1, then 2πx/A is assumed to be in the fourth quadrant. In other words, it is assumed that it is between 3π/2 and 2π. This is D in Figure 8.
It corresponds to 2 to D3.

The incident angle φ can be found from equation (21). Since φ is a small value, it can be approximated as follows.

Next, consider condition (24). φ' is (14)
Given that φ is sufficiently small, φI
= (mouth (28)nl-n・
(4) When approximated to nl, it is given by (27) and (80) limit the value of φ.

FIG. 13 illustrates the range of φ given by (27) and (30).

As shown in FIG. 8, the branching optical fiber can be said to have one entrance window in the fourth quadrant.

In the end, if we stack optical fiber A, which passes propagated light, on the bottom and optical fiber B, which serves as a branch from which signal light is taken out, on top, and enter light into optical fiber A from the left, Figures 7 and 8 As shown, the light emitted from the optical fiber A is active in the first and second quadrants of C1 to C3. The light incident on the optical fiber B is concentrated in the fourth quadrant.

Since the locations are different, it seems difficult for light exchange to occur between the two.

However, in reality, the cores are not in contact with each other;
Light exchange occurs because the cladding and primary coat are present in between.

The light emitted from point C1 of optical fiber A is at point D of optical fiber B.
Achieving a score of 3 is possible depending on the thickness of the cladding and coat.

However, in the present invention, periodicity is not required for the curve of the curved meandering fiber. There is no problem even if the curve is non-periodic.

In this respect, it differs from Stuart's optical bifurcation mentioned above.

(b) Matching oil Matching oil is filled to fill the gap between the two optical fibers. ) The refractive index is the core of the optical fiber.
The one that is close to the cladding refractive index is chosen. This is to reduce reflection at the boundary surface.

In addition to this, when considering light-gathering properties, it is better for the refractive index of the matching oil to be lower than the refractive index of the cladding.

←) Example: Racks with pitch A of 1.5 were placed opposite each other and used as presser plates. The thickness of the holding plate is 400 μm. One side of the rack is movable, and in order to measure the amount of push, press the rack with a micrometer and find that the number of teeth on the 60 rack is 5.10.2
Three types of 0 were used.

A laser beam of λ=0.83 μm was input into the optical fiber A. The incident power P1 is 200 μW. The fiber length of the optical fiber A from the laser beam input end to the optical branching/merging device was set to 3 to 4 m, and this portion was wound 10 turns to a diameter of 12 mm to form a coil.

As mating oil, an ultraviolet curing adhesive (Norland 65) with a refractive index of 1.65 was filled.

I lowered the micrometer if + - Yuki no Ratsu by 1 so that the output light from the other end of the single-kicker fiber A was 160 μW (
(equivalent to 1 dB reduction).

At this time, the coupled light amount P3 of the optical fiber B was measured. Measurements are given by.

There are three types of optical fibers: I, II, and H.

I T/L/Timode NA=0.2 Core diameter 5
0μm Clad diameter 125μm Silicon coating Multimode NA=0.2 Core diameter 50μm
Cladding diameter: 125 μm UV acrylate coating ■ Single mode However, ■ is not covered by the present invention, but was experimented for comparison. The amount of combined light was 0.

Table 1 shows the results of this experiment.

Table 1 Types of optical fibers and amount of depression of rack gear In this way, light of -40 to -7 QdB can be obtained from optical fiber B.

LEDs and LDs can be used as light sources, but LE
The light emitting power of D has been increasing more and more in recent years.

High brightness LEDs are available with powers of 1QdBm. (ldBm = 1 mW) Furthermore, the sensitivity of the light receiving element also increases significantly.

Since the performance of the light emitting element and the light receiving element has improved, it is now possible to use this as signal light if the attenuation between the input power P1 and the output power P3 is -4QdB.

In this example, a matching oil with a higher refractive index than the cladding is used, but if a matching oil with a lower refractive index than the cladding is used, the amount of light transfer will further increase.

(b) Effect: An optical branch/combiner can be easily constructed without destroying the optical fiber. Therefore, they can be easily added and removed even after the optical fiber system has been constructed.

Direct fiber-to-fiber coupling. In optical branching, a light receiving element, a light emitting element, an amplifier circuit, a power supply, etc. are not required.

[Brief explanation of drawings]

FIG. 1 is a cross-sectional view of the optical branching/merging device of the present invention taken along a cross section including an optical fiber. FIG. 2 is a sectional view taken along line I-1 in FIG. FIG. 3 is a sectional view showing another embodiment of the present invention. FIG. 4 is a configuration diagram of a conventional optical branch using a beam splitter and a lens. FIG. 5 is a block diagram of a conventional optical branch using an optical fiber coupler. FIG. 6 is a cross-sectional view of Stewart's meandering optical fiber branch. FIG. 7 is an explanatory diagram showing that light is emitted from a curved optical fiber. FIG. 8 is an explanatory diagram showing that light enters a curved optical fiber from an oblique direction. FIG. 9 is a diagonal diagram explaining that light is emitted at a portion of the optical fiber that changes from a straight section to a curved section. FIG. 10 is a graph showing the range of reflection angles O for emitting light reflected at points on the outer side AE of the straight line portion in FIG. FIG. 11 is a graph showing the range of reflection angle 0 for the emission of light reflected at a point on the inner side BF of the straight line section in FIG. FIG. 12 is a diagram illustrating the conditions for light to enter a curved optical fiber obliquely. FIG. 13 is a diagram showing that when a cluster fiber is superimposed on the main optical fiber and the meandering of the cluster fiber is expressed as sin (2πx/A), an incidence window is generated in the fourth quadrant. 1...Light splitter/combiner 3.4...Press fitting 5.6...Irregular surface 7...Matching oil 8.9...Holding plate July 3, 1985 Commissioner of the Japan Patent Office U Tono Kadobe 1. Indication of the case Patent application 1982-125429 2. Name of the invention ``Light splitter/combiner 3'' Relationship with the person making the amendment Patent applicant's residence Address: 5-15 Kitahama, Higashi-ku, Osaka City Name (213) Sumitomo Electric Industries Co., Ltd. Representative President Satoshi Kawakami Department 4, Agent Address '3-15-16 Nakamichi, Higashinari-ku, Osaka City Complete all drawings that were forgotten to be attached at the time of application: 1

Claims (5)

[Claims] (1) Two multimode optical fibers that are repeatedly bent in a meandering manner are placed close together so that the direction of the bending deformation is on the same plane, and a part of the light propagating in one optical fiber is An optical branching/combining device characterized in that the light is emitted to the outside at a curved part, and a part of this light is made to enter the curved part of another optical fiber to become light that propagates within this optical fiber. . (2) The optical branching/merging device according to claim (1), wherein the optical fiber comprises a core, a cladding, and a primary coat. (3) The optical branching/merging device according to claim (2), wherein the space surrounding the optical fiber is filled with matching oil. (4) Two holding metal fittings 3 and 4 having uneven surfaces 5 and 6 are supported between two holding plates 8 and 9, and two optical fibers are passed between the holding metal fittings 3 and 4. , an optical branching/merging device according to claim (3), characterized in that the light branching/merging device is repeatedly bent in a serpentine shape by being pressed by uneven surfaces 5 and 6. (5) A patent characterized in that there is a transparent guide plate between two optical fibers, and the two meandering optical fibers are deformed with a large radius of curvature so as to be concave outward. An optical branching/merging device according to claim (3).