JP2005241694A - Method of manufacturing optical fiber coupler - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To overcome the problem that an optical fiber coupler exists having a characteristic variation and a defective loss due to an improper heating process of an optical fiber. <P>SOLUTION: A method of manufacturing an optical fiber coupler comprises: aligning a plurality of optical fibers 10 and 20 of which the coating is removed; inputting a light beam from a light source to the input optical fiber 10; and fusing and extending each of the input optical fibers 10 and the other optical fiber 20 while detecting the light power outputted from optical fiber 10 and the optical fiber 20, wherein the light power outputted from the input optical fiber 10 once drops as time passes and recovers, drops again as secondary variation at the same time of the beginning of branching, and the primary drop is kept to be 0.3 dB or smaller, thus the obtained optical fiber has a low loss. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a method of manufacturing a fusion-stretching optical fiber coupler that branches or combines light of a single wavelength or a plurality of wavelengths, or combines or demultiplexes light of a plurality of wavelengths.

  With the advancement of optical communication systems and optical transmission technologies, a variety of optical communication systems, optical fiber demultiplexers, etc., which are easy to manufacture and have high mass productivity, are available. It is used in measurement systems.

  Here, an optical fiber coupler is for branching or coupling light of a single wavelength or multiple wavelengths, and an optical multiplexer / demultiplexer is for multiplexing or demultiplexing light of multiple wavelengths. In the invention, both are collectively referred to as an optical fiber coupler.

  The fusion-stretching optical fiber coupler has a structure as shown in FIG. 1A. First, the coating of the end portions of the optical fibers 10 and 20 is removed, and the core portions 1 and 2 are washed with alcohol or the like. . Next, the core portions 1 and 2 are aligned, while the center portion is heated with a burner such as oxygen-propane or hydrogen, the optical fibers 10 and 20 are stretched left and right in the longitudinal direction. The fusion extending part 3 is formed in the center part. The coupler body 4 composed of the optical fibers 10 and 20 and the fusion stretched part 3 thus manufactured is placed on a substrate 5 such as quartz glass, and is cured by energy rays such as ultraviolet rays and visible rays. Join through. The joint portion is generally at both end portions A of the substrate 5, but the core wire portions 1 and 2 and the substrate 5 are also joined at both end portions B of the fusion stretched portion 3. Although not shown, an optical fiber is formed by covering the substrate 5 with a substrate such as quartz glass, which is the same material as the substrate 5, and covering the substrate 5 with a cylindrical body such as metal, ceramic, or glass. The coupler reinforcement structure is completed.

  However, when fusion / stretching is performed at a constant gas flow rate, the amount of heat is the same even if the optical fibers 10 and 20 are thinned and the cores 1 and 2 of the optical fibers 10 and 20 are too strong. There was a risk that microbending loss would occur due to deformation and excess loss would increase.

  Further, in order to prevent this problem, there is a method of stretching with a certain amount of heat from the beginning. However, in that case, the amount of heat becomes insufficient at the initial stage of stretching, and the processing takes a long time. There is a problem that it is not possible to prevent an increase in excess loss due to deformation of 2 or generation of microbend loss.

  Therefore, in recent years, as one of the causes, proposals have been made to prevent characteristic fluctuations, loss defects, and the like by a heating process when fusing and stretching the optical fiber coupler.

For example, in Patent Document 1, as shown in FIG. 3, the heater is heated to about 1000 ° C. in advance, and the optical fiber is heated to a moderate strain temperature at the start of production to remove the strain of the optical fiber. Then, the temperature is raised to about 1600 ° C., and the optical fibers are fused. When the fusion is completed, the heating temperature is lowered, the drawing stage is moved, and the optical fiber drawing process is started. It has been proposed that the stretching process is provided in four stages according to the stretching speed and the heating temperature during stretching, thereby reducing the characteristic fluctuations and loss defects by moderately changing the temperature. (See Patent Document 1).
JP 2001-324644 A

  However, in the proposal of Patent Document 1, it is unclear how the heating temperature is lowered in the situation of the optical fiber only by the relationship between the heating time and the heating temperature. For this reason, the heating temperature is lowered when the optical fiber diameter is still thick, resulting in a shortage of heat, resulting in microcracks in the optical fiber due to the distortion, increasing excess loss, and conversely the heating temperature when the optical fiber diameter is thin. However, there is a risk that the heat loss becomes too high, the core of the optical fiber is deformed, microbend loss occurs, and the excess loss increases.

  Also, in evanescent coupling, where branching proceeds during the process of fusing and stretching the two optical fibers 10 and 20, there is light that leaks into the cladding as the distance between the cores gradually approaches and approaches a certain distance. However, it is known that the optical power is reduced and recovered before the start of branching due to the leakage. Even if the temperature is lowered stepwise as in Patent Document 1, the light before the start of branching is known. Since the temperature does not change in accordance with the power reduction, the above-mentioned evanescent coupling state is poor and microbends and microcracks of the core are generated. This primary reduction is as large as about 0.5 dB, and the completed optical fiber coupler. There was a problem that the excess loss of P exceeded about 0.1 dB.

  The method of manufacturing an optical fiber coupler of the present invention aligns a plurality of optical fibers from which the coating has been removed, inputs light from a light source to the input optical fiber, and reduces the optical power output from the input optical fiber and other optical fibers. A method of manufacturing an optical fiber coupler in which each optical fiber is fused and stretched while detecting, and the optical power output from the input optical fiber is first reduced over time, and then recovered. It is characterized in that there is a change that secondarily decreases simultaneously with the start of branching, and that the first decrease is 0.3 dB or less.

  In addition, the optical fiber is fused and stretched while being heated at a constant gas flow rate in advance, and after measuring the change of the optical power output from each optical fiber until the branching is completed, the input light is measured. The gas flow rate is reduced by 5% or more from the initial gas flow rate before the optical power of the fiber is linearly reduced.

  Further, the gas flow rate is further reduced at the time of recovery of the optical power and immediately before the secondary reduction.

  According to the method of manufacturing an optical fiber coupler of the present invention, a plurality of optical fibers from which the coating has been removed are aligned, light from a light source is input to the input optical fiber, and light output from the input optical fiber and other optical fibers is output. A method of manufacturing an optical fiber coupler in which each optical fiber is fused and stretched while detecting power, and the optical power output from the input optical fiber is first reduced over time and then recovered. In addition, since the secondary reduction changes at the same time as the start of branching and the primary reduction is 0.3 dB or less, the heating conditions are appropriate for the change in diameter due to the stretching of the optical fiber. The excess loss of the optical fiber coupler can be suppressed to 0.1 dB or less.

  In addition, the optical fiber is fused and stretched while being heated at a constant gas flow rate in advance, and after measuring the change of the optical power output from each optical fiber until the branching is completed, the input light is measured. Since the gas flow rate is decreased by 5% or more from the initial gas flow rate before the optical power of the fiber is reduced to the primary level, the amount of the primary reduction can be made 0.3 dB or lower, so that the loss of the completed optical fiber coupler is lower. Can be.

  Furthermore, since the gas flow rate is further reduced at the time of the recovery of the optical power and immediately before the secondary reduction, since the heating conditions are appropriate due to the change of the diameter due to the drawing of the optical fiber, the light loss can be further reduced. A fiber coupler can be obtained.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  As shown in FIG. 1 (a), the optical fiber coupler is formed by fusing and stretching a part of the core portions 1 and 2 from which the coated portions of the optical fibers 10 and 20 coated on the substrate 5 are removed. The extending portion 3 is formed and bonded to the concave groove 5a of the substrate 5 such as quartz glass via an adhesive 6.

  The optical fiber coupler of the present invention is for branching or combining light of a single wavelength or multiple wavelengths, and the optical multiplexer / demultiplexer is for multiplexing or demultiplexing light of multiple wavelengths. Here, both are collectively referred to as an optical fiber coupler.

  The present invention relates to a method for manufacturing such an optical fiber coupler. When the optical fiber coupler is formed from, for example, two optical fibers 10 and 20 as shown in FIG. The covered portions of the optical fibers 10 and 20 are removed, and the exposed core wire portions 1 and 2 are washed with alcohol or the like. After that, as shown in the second step, the core parts 1 and 2 are closely aligned by paralleling or crossing with an alignment jig or the like, and heated by a burner such as oxygen-propane or hydrogen as shown in the third step. Meanwhile, the optical fibers 10 and 20 are stretched left and right in the longitudinal direction to form a tapered fusion stretched portion 3. The coupler body 4 thus obtained is placed on a concave groove 5a provided in a substrate 5 such as quartz glass as shown in FIG. 1 (a), and is irradiated with energy rays such as ultraviolet rays and visible rays. It joins via the adhesive agent 6 which hardens | cures.

  Here, the fusion / stretching process shown in the third step will be described in detail by taking as an example the case where the dispersion-shifted optical fibers 10 and 20 of 1550 nm are used.

  When heating the core portions 1 and 2 of the optical fibers 10 and 20, light from an LD light source of 1550 nm is incident on one optical fiber 10, for example, about 100 μW. This optical power is stored in the computer as an initial value. Here, an optical fiber serving as a through port in an optical fiber coupler is shown with an optical fiber 10 that receives light from a light source as an input optical fiber 10.

  A power meter is connected to the optical fibers 10 and 20 on the output side, and the loss state is grasped until branching is completed while always detecting the output optical power.

For heating, oxygen-propane mixed gas having a flow rate of oxygen of 16 × 10 ⁻⁶ m ³ / sec and propane of 8 × 10 ⁻⁶ m ³ / sec is ejected from a 0.5 mm diameter outlet. The burner was slid at a speed of 2.4 mm / sec in the longitudinal direction of the optical fibers 10 and 20 within a range of 4 mm on the left and right and heated for 40 seconds.

  The distance between the burner and the core wire parts 1 and 2 is preferably such that when the flame of the burner is applied from above the optical fibers 10 and 20, the leading edge of the flame slightly emerges from the lowest part of the core wire parts 1 and 2.

  Then, after fusion for 40 seconds, the optical fibers 10 and 20 are stretched at a speed of 0.08 mm / sec.

  Under the above conditions, the stretching was stopped when the branching ratio of the optical powers of the light 10 and 20 became 50% and when the stretching length reached about 12.5 mm.

  The time course of the optical power of the optical fibers 10 and 20 at this time is shown in FIG.

  The input optical fiber 10 has a constant value up to the A region with the passage of time, recovers after first decreasing in the B region, and further changes in the C region to secondarily decrease at the start of branching. This is because evanescent coupling occurs in the process of fusing and stretching the two optical fibers 10 and 20, the distance between the cores gradually approaches, and there is light that leaks into the cladding in the process until it reaches a fixed interval. The leakage causes the optical power to drop once before the start of branching and then recover.

  The time until the optical power of the input optical fiber 10 starts to decrease is defined as the boundary between the A region and the B region. The primary decrease in the B region is the time until the optical power decreases and starts increasing again. It shows the time from when the power increases until the optical fibers 10 and 20 start to branch, and the C region is from the start to the end of branching.

  Here, in the manufacturing method of the optical fiber coupler of the present invention, it is important that the primary reduction of the input optical fiber 10 is 0.3 dB or less.

  Thereby, the excess loss of the completed optical fiber coupler can be 0.1 dB or less. The reason for this is that in the evanescent coupling in which branching proceeds in the process of fusing and stretching the two optical fibers 10 and 20, the distance between the cores gradually approaches and leaks into the clad until it approaches a certain interval. There is a phenomenon in which light is present and the optical power decreases and recovers before the start of branching due to leakage, but by reducing the leakage amount to 0.3 dB or less, there is no stress such as microbending or microcracking. Therefore, light can be smoothly refracted and moved between the cores, and excess loss can be reduced.

  If the amount of primary reduction exceeds 0.3 dB, the optical power does not recover until the end of branching, and the excess loss of the completed optical fiber coupler tends to be 0.1 dB or more. The reason for this is that branching loss due to evanescent coupling occurs in the B region and C region in FIG. 2 as the branching progresses. Therefore, it is preferable that the loss is as small as possible in the A region and B region. As a result, the core microbends and microcracks occur, and losses other than the leakage of light from the core to the cladding occur.

  For example, when the heating condition is strong, the core is slightly deformed, light leaks from the deformed portion, and the leak angle is irregular, so that the light becomes clad mode light and does not return to the core again.

  When the heating conditions are weak, microcracks are generated in the core, and light leaks from the portion. Since the leak angle is irregular, the light becomes clad mode light and does not return to the core again.

  In the case of a dispersion-shifted fiber for 1550 nm, the first-order loss is 0.3 dB or less, more preferably 0.2 dB or less, and in the case of a single-mode fiber for 1310 nm, 0.2 dB or less, preferably 0.1 dB or less. More preferably.

To measure optical power, use a power meter to constantly monitor the power until the branching is completed, and to calculate the amount of loss from this optical power, for example, If the initial value is 100 μW, and the optical power drops to 98 μW during the stretching process, the loss is 0.088 dB from the calculation formula: loss (dB) = − ₁₀ Log ₁₀ 98/100.

  Note that the amount of the primary reduction is 0.3 dB or less, preferably 0.2 dB or less in the case of a dispersion-shifted fiber for 1550 nm, and is preferably 0.2 dB or less in the case of a single mode fiber for 1310 nm, preferably 0.2 dB or less. It is preferable to be 0.1 dB or less.

  In order to make this primary decrease 0.3 dB or less, it is important that the core wire parts 1 and 2 are cleaned cleanly before the start of fusion. If impurities such as contaminants are attached, light is refracted and leaks from that portion, resulting in an increase in loss.

  In addition, it is important that the core wire parts 1 and 2 are clamped with the same tension and are tightly clamped. If the tension is non-uniform or the adhesion is poor, stress is generated at the different boundary points during stretching, resulting in microbending and loss.

  Furthermore, in order to reduce the primary drop, finely adjust the stretching speed, burner sliding speed, the distance from the core wire parts 1 and 2 to the burner to obtain the optimum conditions, and further, to stabilize the flame, no wind. It is important to maintain the environment, to reduce the flow rate change error of the oxygen-propane mixed gas by keeping the environment temperature constant.

  As described above, it is necessary to optimize various flame conditions, etc., but the most influential among them is the optimization of the amount of heat when heating the core portions 1 and 2 of the stretched optical fibers 10 and 20. It is. If the amount of heat is too weak, microbends such as microcracks are generated in the optical fibers 10 and 20 that are being stretched, and light leaks from the portions, resulting in loss. On the other hand, if the amount of heat is too strong, the stretch length becomes short, the taper shape becomes steep, and light easily leaks from the core, increasing the loss.

  In order to further reduce the loss, the light of the input optical fiber 10 until the branching is completed by melting and stretching the core wire parts 1 and 2 while heating them at a constant gas flow rate as shown in FIG. It is preferable to reduce the gas flow rate by 5% or more of the initial gas flow rate before measuring the change of power with time and before the primary reduction of the optical power. Further preferable conditions are 5% or more and less than 10%.

  Thereby, even if the optical fibers 10 and 20 are thinned by stretching, optimization of the flame conditions and the like can be obtained as described above, and the primary degradation can be minimized. On the other hand, when the gas flow rate is reduced within a range of less than 5% from the initial gas flow rate, the amount of heat applied to the optical fibers 10 and 20 is poor, the stretch length becomes short, the taper shape becomes steep, and light tends to leak from the core. Therefore, the loss increases or the core is deformed due to an excessive amount of heat and becomes a microbend, so that light easily leaks from the core and the loss increases. In addition, when the gas flow rate is decreased by 10% or more from the initial gas flow rate, the amount of heat applied to the optical fibers 10 and 20 is too much, the heat amount is too weak, and the optical fibers 10 and 20 that are being stretched have microcracks or the like. Microbending occurs, light leaks from that portion, and loss increases.

Further, the timing of reducing the gas flow rate is preferably 5 seconds or more before the primary reduction starts, more preferably 10 to 20 seconds before, for example, the initial gas flow rate is 16 × 10 ⁻⁶ m ³ / sec, When propane is 8 × 10 ⁻⁶ m ³ / sec, it is preferable to reduce oxygen to 15 × 10 ⁻⁶ m ³ / sec and propane to about 7.5 × 10 ⁻⁶ m ³ / sec.

  This is because the diameter of the optical fiber core portions 1 and 2 becomes smaller as the drawing progresses, so the amount of heat must be reduced. If it is 20 seconds or more before that time, the diameters of the core portions 1 and 2 of the optical fibers 10 and 20 are not yet reduced, and the amount of heating is reduced, resulting in microbending loss. When the time is shorter than 5 seconds, the diameter is already narrow, and the appropriate distance between the core and the clad is destroyed, resulting in microbend loss.

  Furthermore, it is preferable to further reduce the gas flow rate at the time of recovery of the optical power shown in FIG. 2 and immediately before the secondary reduction.

  If the stretching is continued under the above-described proper conditions, light starts to move from the core wire portion 1 to the core wire portion 2 after the optical power is recovered as in the region C in FIG. Loss generation can be further suppressed by heating the optical fibers 10 and 20 by further reducing the gas flow rate before the start of the branching. This is because the diameters of the core portions 1 and 2 of the optical fibers 10 and 20 become thinner as the drawing progresses, so that the amount of heat is also reduced, so that the microbend loss generated in the core can be effectively prevented.

  The gas flow rate at this time is preferably further reduced by about 2 to 6% with respect to the gas flow rate reduced before the primary reduction.

For example, when oxygen is 15 × 10 ⁻⁶ m ³ / sec and propane is 7.5 × 10 ⁻⁶ m ³ / sec before the primary decrease, oxygen is 14.5 × 10 ⁻ immediately before the secondary decrease. It is preferable that the pressure is ⁶ m ³ / sec and propane is about 7.2 × 10 ⁻⁶ m ³ / sec.

  Thus, by adjusting the heating conditions in accordance with changes in the diameters of the optical fibers 10 and 20, the amount of primary reduction is made as small as 0.3 dB or less, and the loss of the obtained optical fiber coupler is set to 0. It can be as small as 1 dB or less.

  In addition, the manufacturing method of the optical fiber coupler of this invention is not limited to the above-mentioned embodiment, A various change is possible if it is in the range which does not deviate from the summary of this invention.

  Next, examples of the present invention will be described.

  Here, ten optical fiber coupler samples as shown in FIG. 1A were prepared using a dispersion-shifted fiber for 1550 nm.

  The target values were set such that the branching ratio was 50 ± 3% and the excess loss was 0.1 dB or less.

As an example of the present invention, first, the change with time of optical power was measured at a constant gas flow rate, oxygen at 16 × 10 ⁻⁶ m ³ / sec, and propane at 8 × 10 ⁻⁶ m ³ / sec.

In this case, the primary decrease in optical power occurred 110 seconds after the start of fusion. Therefore, oxygen was reduced to 15 × 10 ⁻⁶ m ³ / sec and propane to 7.5 × 10 ⁻⁶ m ³ / sec 100 seconds after the start of fusion 10 seconds before. Further, oxygen was reduced to 14.5 × 10 ⁻⁶ m ³ / sec and propane was reduced to 7.2 × 10 ⁻⁶ m ³ / sec before the optical power was recovered and branching started.

  In addition, various heating conditions were set so that the primary decrease in optical power was 0.2 dB or less.

  A burner ejected from a 0.5 mm diameter outlet was slid at a speed of 2.4 mm / sec in the longitudinal direction of the optical fibers 10 and 20 within a range of 4 mm to the left and right and heated for 40 seconds.

  The distance between the burner and the core wire parts 1 and 2 was set so that the flame of the burner was applied from above the optical fibers 10 and 20, and the leading edge of the flame protruded 1.5 mm from the lowest part of the core wire parts 1 and 2 at the initial gas flow rate. .

  Then, after fusion for 40 seconds, stretching of the optical fibers 10 and 20 was started at a speed of 0.085 mm / sec.

  In this case, the primary reduction in optical power could be suppressed to 0.1 dB to 0.2 dB.

  Under the above conditions, the stretching was stopped when the branching ratio of the optical powers of the lights 10 and 20 reached 50%.

  As a comparative example, ten optical fiber coupler samples as shown in FIG. 1 were prepared using a dispersion shifted fiber for 1550 nm.

  The target values were set such that the branching ratio was 50 ± 3% and the excess loss was 0.1 dB or less.

  The primary decrease in optical power before branching was not considered.

Here, the initial gas flow rate was fused and stretched to the end while oxygen was 16 × 10 ⁻⁶ m ³ / sec and propane was 8 × 10 ⁻⁶ m ³ / sec. Other conditions were the same as above.

  In this case, the primary decrease in the optical power before branching exceeded 0.3 dB in most cases, and was about 0.3 dB to 0.5 dB.

  The branching ratio and excess loss after completion of each fusion / stretching were measured and compared.

  Here, the branching ratio and excess loss were measured as follows.

  An LD light source was connected to the input side optical fiber 10 and light of about 100 μW was input. The output side optical fibers 10 and 20 were connected to a power meter, and monitoring was performed while controlling the computer until the branching was completed while monitoring the optical power of the output side optical fibers 10 and 20 at all times.

  The branching ratio is calculated using the formula: branching ratio (%) = P2 / (P1 + P2) × 100 (%, where P1 is the optical power of the output-side optical fiber 10 after stretching and P2 is the optical power of the output-side optical fiber 20. ).

The excess loss is calculated by assuming that the optical power of the input-side optical fiber 10 before starting stretching is P0, the optical power of the output-side optical fiber 10 after stretching is P1, and the optical power of the output-side optical fiber 20 is P2. Excess loss (dB) = − ₁₀ Log ₁₀ (P1 + P2) / P0.

The results were as shown in Table 1.

  From the results shown in Table 1, all of the samples of the present invention were able to reduce the branching ratio to within 50% ± 2% and all the excess loss to 0.1 dB or less.

  On the other hand, the sample of the conventional example had large variations in branching ratio and excess loss.

(A) is a top view which shows the structure of an optical fiber coupler, (b) is a top view for demonstrating the manufacturing method of the optical fiber coupler of this invention. FIG. 4 is a relationship diagram between the passage of time of an optical fiber coupler and the transition of optical power. It is a related figure of the heating time and heating temperature in the manufacturing method of the conventional optical fiber coupler.

Explanation of symbols

1, 2: Core wire part 3: Fusion stretch part 4: Coupler body 5: Substrate 5a: Groove 6: Adhesive 7: Adhesive 8: Other adhesive 10, 20: Optical fiber A, B: Adhesive position

Claims (3)

Align multiple optical fibers from which the coating has been removed, input light from the light source to the input optical fiber, and fuse each optical fiber while detecting the optical power output from the input optical fiber and other optical fibers, A method of manufacturing a stretched optical fiber coupler, wherein the optical power output from the input optical fiber recovers after a primary decrease over time, and further changes to decrease secondary at the same time as branching starts. And the said primary reduction is 0.3 dB or less, The manufacturing method of the optical fiber coupler characterized by the above-mentioned. The optical fiber is fused and stretched in advance while heating at a constant gas flow rate, and after measuring the change with time of the optical power output from each optical fiber until the branching is completed, the input optical fiber 2. The method of manufacturing an optical fiber coupler according to claim 1, wherein the gas flow rate is reduced by 5% or more from the initial gas flow rate before the optical power is reduced by the primary reduction. 3. The method of manufacturing an optical fiber coupler according to claim 2, wherein the gas flow rate is further reduced at the time of recovery of the optical power and immediately before the secondary reduction.

Images