JPH0727147B2 - Directional coupler type optical waveguide switch - Google Patents

Description

TECHNICAL FIELD The present invention relates to an optical control element for optical communication, and in particular, when one of two possible states does not require a bias voltage, Moreover, the present invention relates to a directional coupler type waveguide optical switch having excellent crosstalk characteristics.

[Conventional technology]

Optical communication systems have been put into practical use and are being developed into higher-level systems having large capacity and multiple functions. New functions such as a switching function of an optical transmission network and high-speed connection / switching between terminals in an optical data bus are required, and the need for an optical switching network that enables them is increasing.

Currently used optical switches are prisms, mirrors,
It is a means for mechanically moving a fiber or the like, and has drawbacks such as low speed and insufficient reliability, large shape, and unsuitable for matrix formation. What is being developed as a means to solve this is a waveguide type switch that uses an optical waveguide installed on a substrate, and has features such as high speed, multi-element integration, and high reliability. . In particular, those using ferroelectric materials such as lithium niobate (LiNbO ₃ ) crystals have the advantages of low light absorption and low loss, and high efficiency because they have a large electro-optical effect. .

A number of methods have been invented so far as the device type of the waveguide type switch, but the operating voltage is relatively low, and the radiation loss of light is small because it is composed of channel waveguides throughout the light transmission direction. Due to the advantage, a cobra type directional coupler type optical switch is often used.

The structure and operation principle of the cobra-type directional coupler-type optical switch is such that when no voltage is applied to the directional coupler, all the light in one waveguide is transferred to the other waveguide (this state is called a cross state). 2) that have the same phase constant to be placed in close proximity to each other.
When the coupling length of the optical waveguide of the book is set and an appropriate voltage is applied, the phase constants of the two waveguides differ, so that the coupling of light from one to the other does not occur (in this state, Switching is realized by selecting two states, that is, the x state and the = state, in which light is generally referred to as a state and is represented by the = symbol) It is a thing.

[Problems of conventional technology]

While this type of directional coupler type optical switch has the features described above, it has a sufficiently low optical output from the waveguide that should not be output in the x state when no voltage is applied, that is, high crosstalk. In order to secure the characteristics, it is necessary to set the joint length with extremely high accuracy. However, it is difficult to meet this with the actual manufacturing technique of the optical waveguide.

As a device method for relaxing the manufacturing accuracy of a cobra-type directional coupler-type waveguide optical switch, delta beta (Δ
β) The inversion method was invented. In this Δβ inversion method, an electrode for applying a voltage to a directional coupler is divided into two in the light transmitting direction and the electric field direction is inverted, whereby the coupling portion length is set to a certain length or more. For example, x and = can be realized by operating between two voltage values.

As a result, if the bias voltage is always applied, the x state with high crosstalk characteristics can be secured, but there is a drawback. That is, the typical one is not a single switch but a matrix switch in which a large number of switches are integrated, and the bias voltage of each element is not constant and varies. In such a case, it is practically difficult to adjust and give the value of the bias voltage to all the elements.

As described above, in order to construct a matrix switch in which a large number of switches are integrated, a device system that does not require bias is desired.

[Object of the Invention]

An object of the present invention is to eliminate the above-mentioned drawbacks of the conventional waveguide type optical switch and to eliminate the need for a bias voltage when in one of two possible states and to provide a directional coupler having excellent crosstalk characteristics. Type waveguide optical switch.

[Means for solving problems]

According to the present invention, two channel optical waveguides are closely and parallel to each other on a z-plate lithium niobate crystal, and an S-shape is drawn in the light transmitting direction by connecting curved lines of positive and negative circles. And an electric field applying electrode is provided on the two channel optical waveguides along the meandering of the two channel optical waveguides so that no bias voltage is required and the crosstalk characteristics are excellent. Is obtained.

〔Example〕

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

FIG. 1 is a top view showing the structure of a directional coupler type optical switch which is an embodiment of the present invention. 1 is a LiNbO ₃ crystal plate, and the substrate orientation is selected to be a z plate (that is, the normal line standing on the substrate is the z axis). Two optical waveguides 2 and 3 which are provided in proximity to each other by diffusing Ti or the like into the crystal substrate, and electrodes for applying a switch voltage provided on these waveguides along the meandering thereof. It is composed of 4 to 7.

The two optical waveguides have the same waveguide width and are parallel to each other to form a directional coupler. However, the shape of this directional coupler is not an ordinary linear shape but an S-shaped curve. The curvature of the convex portion and the curvature of the concave portion (that is, the waveguide 2a and the waveguide 3b) are made equal, and their lengths are also made equal. The waveguide 3a and the waveguide 2b have the same relationship.

The coupling characteristics of parallel directional couplers having a curvature (corresponding to the directional couplers of only the right half or the left half of the symmetry line AB in FIG. 1) are examined. (For example, the paper published in Applied Optics, Vol. 19, No. 3, pp. 398-403, 'Coupling bet between dielectric waveguides with curvature').
ween cur-ved dielectric wfveguides) '). According to the analysis in this paper, in a curved coupling waveguide having the same waveguide width and the same curvature center and parallel to each other, a normal linear directional coupling characteristic is exhibited when the radius of curvature is infinite. As the radius of curvature becomes smaller, the optical energy of one waveguide is not completely transferred to the other waveguide (complete coupling), and only incomplete coupling occurs. Below a certain radius of curvature (for example, in the calculation example of this paper, radius 30-40 mm
Hereinafter, this means that a phase constant difference is generated between the two waveguides, and just by applying an electric field with a linear directional coupler, the = state is generated. Understood to be equivalent.

In this embodiment, as shown in FIG. 1, curved portions are formed by combining concave and convex portions. This directional coupler has a left side (that is, waveguides 2a and 3a) with a line of symmetry AB as a boundary.
(A part consisting of), there is a positive Δβ between the two waveguides.
However, there is a negative Δβ on the right side (that is, the portion formed by the waveguides 2b and 3b). In other words, Δβ inversion is automatically hard-pressed. Therefore, in the state where no electric field is applied, the = state is automatically guaranteed. Next, a voltage is applied to the electrodes 4 to 7 provided on the waveguide so as to compensate for the bound Δβ. At a certain voltage value, this Δβ is canceled out, resulting in a × state.
According to the counting of the Δβ inversion method described above, this voltage value always exists.

Thus, the = state is ensured when no voltage is applied, and the x state is obtained when a voltage is applied. Especially,
Even when a large number of switch elements are integrated and formed, if the radius of curvature of the waveguide is set to a certain value or less, the variations generated in the waveguide creation process are absorbed,
All elements can be obtained in a state where no voltage is applied, and the bias voltage required in the conventional linear Δβ inversion method is not required.

〔The invention's effect〕

As described above, according to the present invention, a directional coupler type waveguide optical switch that does not require a bias voltage and has excellent crosstalk characteristics can be obtained.

[Brief description of drawings]

FIG. 1 is a top view for explaining the structure of one embodiment of the present invention. 1 is a LiNbO ₃ crystal substrate, 2a, b and 3a, b are optical waveguides, 4 to 7
Is an electrode.

Claims (1)

[Claims] 1. A z-plate lithium niobate crystal is provided with two channel optical waveguides in close proximity to each other and in parallel, and further, an S-shape is drawn in the light transmitting direction by connecting curves of positive and negative circles. A directional crystallizer type optical waveguide switch, which is characterized in that it is formed and an electric field applying electrode is installed on the two channel optical waveguides along the meandering thereof.