JPH01222233A - Directional coupler optical switch - Google Patents

Abstract

PURPOSE:To reduce a necessary operating voltage by specifying the length of the coupling area of two directional couplers which are arranged in parallel. CONSTITUTION:The propagation constant difference generated when one of TE and TM modes is propagated in two waveguides while an electric field is applied is DELTAbeta1, and the propagation constant difference generated when the other mode which crosses said mode at right angles is propagated is DELTAbeta2. Then when alpha=DELTAbeta1/DELTAbeta2, 0.28<=alpha<=0.31. Further, the length of the coupling area of the directional couplers is denoted as l and the complete coupling lengths for the TE and TM mode waveguide light beams are denoted as LTE and LTM. In this case, either of the ratios l/LTE and l/LTM is denoted as x1 and the other is denoted as x2; and the l, LTE, LTM and alpha are so determined that expression I and II hold at the same time. Consequently, the arrangement of X or Y propagation is usable without using the arrangement of Z propagation and low-voltage operation is enabled.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a waveguide type optical switch used in an optical fiber communication system and an optical fiber sensor system.
In particular, it relates to a directional coupler optical switch that does not depend on the polarization state of guided light.

[Conventional technology]

Conventionally, a waveguide type optical switch using an electro-optic effect is described in, for example, US Pat. No. 4,012,113. This is an optical switch in which a directional coupler is provided with divided electrodes, but it usually has a problem of polarization dependence because the coupling length is different between the TE mode and the 7M mode.

-a, an optical switch used in an optical fiber communication system or an optical fiber sensor system is often inserted into an optical transmission line and used to switch the optical path of an optical signal propagating through the optical fiber.

Single-mode fibers are used as transmission paths in high-speed, large-capacity optical fiber communication systems, but since two orthogonal polarized lights exist as propagation modes and coupling can occur between them, the light propagating through the fiber is generally polarized. State is not saved.

Therefore, linearly polarized light may become elliptically polarized light, and the polarization state may also change over time.

Therefore, an optical switch inserted between optical fibers is required to have the characteristic of being able to switch the optical path without depending on the polarization state of input light.

As an optical switch with such polarization independence, the fifth
The one shown in the figure is known.

This is Appl by R C Alphaness.

Phys, Lett, 1979, Vol. 35, No. 10, 7
This is reported on pages 48-750.

This optical switch is provided with a pair of Ti diffusion waveguides 51 and 52 whose distance d changes with respect to the light propagation direction X on a LiNbO3 wafer 50, and with electrodes 53 that produce a propagation constant difference Δβ of the guided light.゜By providing 54A and 54B on the top surface of the wafer and optimizing the amount of Δβ, T
It is possible to switch to both E-7M modes.

The structure described above has the advantage that the extinction ratio can be improved compared to a conventional optical switch in which the spacing between the waveguides is constant.

[Problem that the invention seeks to solve]

However, in the directional coupler optical switch with the structure shown in Figure 5, in order to achieve polarization-independent characteristics, the spacing between the waveguides changes continuously, and there are regions with large spacing, so it is necessary to There was a serious problem in that the operating voltage was high. Furthermore, it is essential to make one of the waveguides curved according to a specific shape, and there is a problem in that it is very difficult to make a mask for forming the waveguide.

At the same time, there is a problem in that there are extremely limited methods for correcting when the waveguide deviates from its optimum state.

[Means for solving problems]

In order to solve the above conventional problems, the present invention provides LiNb
A directional coupler optical switch is provided with a directional coupler consisting of a pair of linear optical waveguides close to each other on a 0z wafer, and a control electrode is provided on the wafer divided along the waveguide. A secondary electrode is placed parallel to the electrode, and the interrelationships among the propagation constant difference, TE, the complete coupling length of the 7M mode, and the length l of the coupling region are selected as follows.

In other words, the difference in propagation constant caused by one mode propagating through the two waveguides under voltage application is Δβ1, and the difference in propagation constant caused by one mode propagating through the two waveguides under voltage application is Δβ1. Letting the resulting propagation constant difference be Δβ2, the ratio of both α=
Δβ1/Δβ2 is within the range of 0.28 to 0.31.

In addition, the complete bond length Lyi, L of TE mode and TM mode
The ratio of tH to the length l of the binding region 1/+, ? t1/Lt
If one of s is xl and the other is x2, the following two equations are simultaneously satisfied: xl+X! and the following two equations, ...(2a) where i=1. The values of l/Ltt - and l/LTH are selected so that they exist within the range of intersection with the range of xIaXt in which 2π and 2π are simultaneously satisfied.

[Effect]

According to the present invention, in two directional couplers arranged in parallel, the length of the coupling region is selected to satisfy a specific condition, so that when voltages of different polarities are applied to the divided electrodes, In this case, both the TE and TM modes are in a cross state, and when voltages of the same polarity are applied to the divided electrodes, both the TE and TM modes are in a bar state.

According to the invention, therefore, the directional coupler acts as a polarization-independent optical switch.

〔Example〕

FIG. 1 is a perspective view showing a directional coupler optical switch according to the present invention, in which two cuts of LiNbO3 wafer 3 are
A pair of optical waveguides 1.2 are provided by TI diffusion from the wafer surface, and both thermal waveguides 1.2 are spaced close to each other in the center of the wafer, excluding near both end surfaces, and the directivity is It constitutes a coupler 17.

The area of the directional coupler 17 is divided into two in the waveguide optical axis direction, and the waveguide 1.
2 on electrodes 4b, 4c and 5b.

5c, and electrodes 4a, 4b and 5a, 5d are provided outside the waveguide in parallel with the electrodes on the waveguide at intervals.
are provided for each.

Next, the operating principle will be explained.

First, the general basic operation of a directional coupler optical switch is −
, IEEE Journal of Quantum I
Electronics 1976 QE-12 Volume 7
, No. 396-401, Ecchi Koge/l/
It is described in detail in the paper by Nick et al.

In general, the bar state of optical path switching by a directional coupler optical switch for TE or 7M mode light means, for example, in FIG. is output to the output end 2b of the other waveguide 2, and is not output to the output end 2b of the other waveguide 2.
This is a state in which the input light to the input end 2a of the waveguide is output to the output end 2b of the same waveguide, and is not output to the output end 1b of the other waveguide 1 at all.

This state is achieved when the optical amplitude B1 in the uniform Δβ optical switch in the above paper is satisfied from the equation: B1 = 0, that is, the following equation: (4) The curve in the diagram with Δβl/π on the horizontal axis and 1/L on the vertical axis corresponds to this. In the optical switch shown in FIG. 1, if a positive potential is applied to the electrodes 4b and 5b of the directional coupler 17 and a negative potential is applied to the electrodes 4C and 50, the optical switch becomes a uniform Δβ optical switch. At this time, the electric field applied to the optical waveguide 1.2 is mainly an electric field component E2 parallel to the Z-axis of the crystal of the wafer 3, and changes in the refractive index of the Z-axis and X-axis due to the electro-optic effect, ■ Δη2−□ ηm' T 33EZ −(5a)■ Δη8=□η. 3rlJZ...(5b) is
The propagation constants of the 7M mode wave and TE mode wave propagating through the directional coupler 17 are changed.

If the amount of change is Δβπ and 627M, respectively, the ratio α−Δβ78/Δβ□ (6) becomes substantially 0.28 to 0.31.

In FIG. 2(A), the T waveguided to the directional coupler 17
The state of the output light of the E mode to 7M mode wave is α: when an electric field E2 is applied from points p and p2 in the figure, respectively.
Δβ, E and 627M increase at a ratio of 1 to 1:α,
It is possible to move to points P, , P, or their vicinity on the arc representing the bar states with radii of 2 and 6, respectively. At this time, the directional coupler 17 is substantially in a bar state for both the TE and 7M mode waves.

Next, a cross state means that the input light to the input end 1a of one waveguide 1 is output to the output end 2b of the opposing waveguide 2, and is not output to the output end 1b of the waveguide 1 at all. , and the input light to the input end 2a of the waveguide 20 is transmitted to the output end 2b of the waveguide 20.
There is no output from the output end 1b of the opposing waveguide 1.
This is the state output from. This is calculated from the following equation, which expresses the optical amplitude A2 when using the inversion Δβ in the above-mentioned paper, from A
2-0, that is, when the following equation holds, 1/L shown in Figure 2 (B) is achieved.
, Δβ1/π diagram.

In the directional coupler optical switch shown in Fig. 1, a positive potential is applied to the electrodes 4b and 50 at the cross position on the waveguide, and a negative potential is applied to the electrodes 4C and 5b at the symmetrical cross position. and +Δβ in the first half region 17a of the directional coupler 17, and -Δβ in the second half region 17b.
Since the propagation constant is given, the above directional coupler optical switch becomes an inverted Δβ optical switch.

At this time, the TE mode guided to the directional coupler 17 and the 7M
The output light state of each mode is changed from points P, , pg in FIG. 2(B) to Δβ7, . .

By increasing Δβ□, the point P5゜P on the curve,
Alternatively, it can be moved to a location very close to it.

At this time, the directional coupler 17 is in a substantially crossed state for both the TE and 7M mode waves.

Since the optical switch in the present invention operates based on the above principle, the points P+ and Pg in FIGS. 2(A) and (B)
If the settings of are not appropriate, the state of the optical output will not match even though when one shifts from P to pg or PS, the other shifts from P2 to near P4 or P6.

Furthermore, if the latter is defeated by points P4 to P, the former will no longer match points P to P5.

Therefore, the settings of P+ and Pg' are important. The curve 23 in FIG.
This is a condition for the transition to the bar state points P, ,P, at the same time. Further, the conditions for the cross state are shown in a curve 24. From this, it follows that at point P in the figure, j? / tyE
, l/L, that is, points Pt and Pg may be set.

However, due to constraints such as manufacturing errors, the point Pl+P2
Since it is difficult to completely set the point P7 in FIG. 4, the present invention sets the allowable range as follows.

Generally, in optical fiber communications, it is desired that the extinction ratio for optical path switching be -20 dB or less.

Therefore, in a substantial bar state or cross state, it is desirable that the optical output is 100:1 or less in equations (3) and (7). In other words, the following (9), Equation 01% Equation % (9) Moreover, in order to make it polarization independent, both Equation (9) and Equation 00 must hold for the TE mode and 7M mode at the same time. Must be.

The present invention stipulates the above conditions, and a region 28 surrounded by curves 26 and 27 in FIG. 4 corresponds to this.

Note that the curve 26 is a range where formula (1) is satisfied, and the curve 27 is a range where formula (2) is satisfied. For this reason,
If the directional coupler is made so that 17Ltt and Il/ltg are in the region 28, the extinction ratio is -20dB.
A polarization-independent optical switch having the above-mentioned properties can be realized.

LiNb0. which can be used in the present invention. The direction of + is Z
c) Electro-optic coefficients 712 and γ including Y propagation,
In addition to these, orientations such as X cut Y propagation, Y cut) X propagation, and 2 cut) X propagation are also effective.

Next, specific numerical examples of the present invention will be explained.

To fabricate the directional coupler optical switch shown in FIG. 1, a Ti film with a thickness of 470 mm and a diameter of 90 μm was formed on the surface of the Z-cut LiNb0:+ wafer 3 using the step-off method. Thermal diffusion was carried out at a temperature of toso° C. for 8 hours in an atmosphere containing water vapor to form an optical waveguide 1.2.

The optical waveguide 1.2 includes input ends 1a, 2a and output ends 1b, 2b which are formed of curved waveguides, and a directional coupler section 17 which is formed of straight waveguides that are close to each other in a parallel state. It will be done. The directional coupler 17 that switches the optical path of the guided light has a region length l of 19 mm and an interval d of 9 μm.

The optical waveguides 1 and 2 manufactured under the above conditions have a wavelength λ=1
．． It becomes a single mode waveguide for TE and 7M mode waves of 3 μm, and the characteristics of the directional coupler 17 are as follows: //Ltt regarding the ratio of the complete coupling length LfE+LTM to the length l of the coupling region for TE and 7M mode waves. =1.03, /
/Lyx=1. It showed a value of IO.

After forming a SiO2 film buffer layer with a thickness of 1500 mm on the directional coupler 17 as described above by sputtering, a total of 8 electrodes made of an An film made of steel having a length of 9 mm and a width of 10 μ are formed on this layer. Established. The electrode includes a primary electrode 4c directly above the optical waveguide 1.2 in the first half region 17a of the directional coupler 17.
, 5b are arranged, and secondary electrodes 4a, 4d are arranged on the outside thereof.
There is.

Also in the latter half region 17b, the optical waveguide l.

Primary electrodes 5b and 5c are provided directly above the electrodes 2, and secondary electrodes 5a and 5d are arranged outside of the primary electrodes 5b and 5c, respectively.

The measurement results of the optical switch manufactured through the above steps will be explained.

The bar state in which all the light input to the input end 1a of one waveguide 1!11 is output to the output end 1b of the same waveguide is the -order electrode 4.
This was achieved by applying a voltage of about 18 V to the secondary electrodes 4d and 5d and grounding the negative electrodes 4c and 5c and the secondary electrodes 4a and 5a.

At that time, the input end 1a of the waveguide 1 is connected at an angle of 45° to the Z axis.
When linearly polarized light with an inclined wavelength λ=1.3 μm was input, the ratio of optical power at the output end 1b of the same waveguide 1 to the output end 2b of the other waveguide was measured and found to be -23 dB.

In addition, the cross state in which all the input light to the input end 1a of one waveguide 1 is output to the output end 2b of the other waveguide 2 is -
Secondary electrodes 4b, 5c and secondary electrode 4d.

A voltage of about 5V is applied to 5a, and the remaining primary electrodes 4c, 5
This was achieved with the secondary electrodes 4a and 5d grounded.

At that time, the ratio of the input end 1a to the output end 2b is approximately -2
It was 0dB.

Next, another embodiment of the present invention is shown in FIG.

In this example, the optical waveguides 12 and 13 are X-cut LiNbO3
It is formed in the Y and two planes of the wafer 14, and the direction of the Z axis is rotated by 90° in the plane perpendicular to the propagation direction.

Therefore, the arrangement of the electrodes must be changed so that an electric field parallel to the Z axis can be applied.

In this case, in the first half region 18a of the directional coupler 18, primary electrodes 15 are provided on both sides and in the center of the optical waveguide 12.13.
b, 15d, and 15c are provided.

Note that the electrodes 15a and 15e arranged outside and parallel to the secondary electrodes 15b and 15d are secondary electrodes.

Further, in the rear half region 18b of the directional coupler 18, electrodes 168 to 16e similar to the above electrodes 158 to 15e are arranged, respectively. TE mode and 7M in this example
The operations on the mode 1/L1Δβl/π diagrams in FIGS. 2 and 3 correspond to the 7M mode and TE mode of the embodiment in FIG. 1, respectively, and similarly to the embodiment in FIG. We confirmed that it operates as a state-independent optical switch.

〔Effect of the invention〕

According to the present invention, in a directional coupler optical switch that does not depend on the polarization of input light, especially in an optical switch in which the waveguide spacing is continuously changed, the strict conditions for waveguide fabrication accuracy that were conventionally unavoidable have been significantly reduced. can be reduced.

Further, according to the present invention, since an X propagation or Y propagation arrangement can be used without using a Z propagation arrangement, there is an advantage that low voltage operation is possible, and the crosstalk level between output ports can be kept low. There is also.

Furthermore, the configuration is not complicated and can be made smaller.

Due to these advantages, according to the present invention, a practical polarization-independent directional coupler type optical switch suitable for mass production can be realized.

[Brief explanation of the drawing]

FIG. 1 is a perspective view showing an embodiment of the present invention, and FIG. 2 is a diagram showing a characteristic curve showing the principle of the present invention. (B) shows the case where voltages of different polarities are applied to the divided electrodes, and the third
The figure is a perspective view showing another embodiment of the present invention, Figure 4 is a diagram showing the effective coupling length area in a directional coupler, and Figure 5 is a conventional polarization-independent directional coupler type optical switch. FIG. 1.2.12.13... Optical waveguide, la, 2a--
Input end, lb, 2b...Output end, 3, 14...Li
Nb0+ wafer, 4 b, 4 c, 5 b+ 5
C515b. 15c, 15d, 16b, 16c, 16d...-Next electrode, 4a, 4d, 5a, 5d, 15a, 15e. 16a, 16e...Secondary electrode, 17.18...Directional coupler, 23...Characteristics for both TE and 7M modes to be in a bar state, 24...Characteristics for both TE and 7M modes to be in a cross state Characteristics to become, 26... Tolerance range of -20dB in bar state, 27... -20dB in cross state
Tolerance range of B: 1.28... Effective area of the optical switch according to the present invention. Figure 1 Figure 2 Figure 3 Figure 4 Multi (□ (X or Y cut) Figure 5 Procedural amendment document March 25, 1960)

Claims (1)

[Claims] A directional coupler consisting of two linear optical waveguides arranged closely at a constant interval formed on a substrate made of a material having an electro-optic effect, and directly above or above the directional coupler. placed parallel to the vicinity,
In a directional coupler optical switch equipped with a control electrode divided into at least two parts in the propagation direction, in order to realize a cross state in which the input light to one waveguide almost completely transfers to the other waveguide and is output, By using an inverted Δβ optical switch in which different electric fields are applied to one and the other of the control electrodes, and in order to realize a bar state in which the output light is almost completely extracted from the waveguide that guided the input light, both of the control electrodes are used. It utilizes the principle of a uniform Δβ optical switch in which an electric field of the same polarity is applied to the two waveguides, and the difference in propagation constant that occurs when one mode propagates through the two waveguides under the applied electric field is defined as Δβ_1 and the above. When the propagation constant difference that occurs when the other mode orthogonal to the mode propagates through the waveguide is Δβ_2, α=Δβ_1/Δβ_2, the range of α is 0.28
≦α≦0.31, and the length of the coupling region of the directional coupler is l, and the complete coupling length for TE and TM mode guided light is L_T_E, L_T_
When M is the ratio l/_L_T_E, l/_L_T_M, one of which is x_1 and the other is x_2, (i) In the range of 4≦Δβ_2l/π≦8, x＾2_i/x＾2_i(Δβ_il/π )^2・si
n^2π/√[x^2_i+(Δβ_il/π)]^2
≦1/101 (i=1, 2) (ii) In the range of 0≦Δβ_2l/π≦3, [1-2x
^2_i/x^2_i+(Δβ_il/π)^2・si
n^2π/4√[x^2_i+(Δβ_il/π)^2
]] The coupling region length l, the complete coupling lengths LTE and LTM, and the ratio α of the propagation constant difference are determined so that the following two equations are simultaneously satisfied: ^2≦1/101 (i=1, 2) A polarization-independent directional coupler optical switch.