1 x N Optical Switch

Abstract

An optical switch is provided which is capable of driving control by the same FPGA and the same driving circuit configuration, and hence is capable of driving at a high speed and a low consumption power. The optical switch of the present disclosure is a 1×N optical switch having a structure in which with respect to an optical switch, a driving circuit of the optical switch is integrated in the vicinity of a control electrode of the optical switch. The optical switch includes a plurality of 2×2 optical switches and N optical gates. Different bias voltages (V.sub.b) are set between the optical switches and the optical gates, and a driver for the 2×2 optical switch of the driving circuit and a driver for the optical gate are of the same circuit form

Claims

1. A 1×N optical switch having a structure in which with respect to an optical switch, a driving circuit of the optical switch is integrated in the vicinity of a control electrode of the optical switch, the optical switch comprising a plurality of 2×2 optical switches and N optical gates, wherein different bias voltages (V.sub.b) are set between the 2×2 optical switches and the optical gates, and a driver for the 2×2 optical switch of the driving circuit and a driver for the optical gate are of the same circuit form.

2. A 1×N optical switch having a structure in which with respect to an optical switch, a driving circuit of the optical switch is integrated in the vicinity of a control electrode of the optical switch, the optical switch comprising a plurality of 2×2 optical switches and N optical gates, wherein a power supply voltage (V.sub.d−V.sub.s) of a driver for the 2×2 optical switch of the driving circuit and a power supply voltage of a driver for the optical gate are equal.

3. The optical switch according to claim 1, wherein for the 2×2 optical switch, a Mach-Zehnder interferometer (MZI) is used, the optical switch has a structure in which a plurality of 2×2 MZI's including an electrode for applying a voltage or a current to each of two waveguides of the MZI are arranged in a multistage, and is characterized by a structure in which one of two input ports of the 2×2 MZI at a latter stage is connected to each of two output ports of the 2×2 MZI at a previous stage in a tree shape.

4. The optical switch according to claim 1, wherein the optical gate is an electric field absorbing type optical gate.

5. The optical switch according to claim 1, wherein the optical switch and the driving circuit are integrated on the same substrate.

6. The optical switch according to claim 1, wherein the optical switch and the driving circuit are integrated by flipchip mounting.

7. The optical switch according to claim 1, further comprising: a distributed parameter line to be connected with the driving circuit; and a FPGA to be connected with the distributed parameter line, wherein as a control signal of the optical switch to be input to the driving circuit, a LVDS signal from the FPGA is used.

8. The optical switch according to claim 1, wherein a terminating circuit for terminating a LVDS signal from the FPGA, and the driving circuit using a transistor having a deep ridge waveguide structure drive the 2×2 optical switch and the optical gate.

9. The optical switch according to claim 2, wherein for the 2×2 optical switch, a Mach-Zehnder interferometer (MZI) is used, the optical switch has a structure in which a plurality of 2×2 MZI's including an electrode for applying a voltage or a current to each of two waveguides of the MZI are arranged in a multistage, and is characterized by a structure in which one of two input ports of the 2×2 MZI at a latter stage is connected to each of two output ports of the 2×2 MZI at a previous stage in a tree shape.

10. The optical switch according to claim 2, wherein the optical gate is an electric field absorbing type optical gate.

11. The optical switch according to claim 2, wherein the optical switch and the driving circuit are integrated on the same substrate.

12. The optical switch according to claim 2, wherein the optical switch and the driving circuit are integrated by flipchip mounting.

13. The optical switch according to claim 2, further comprising: a distributed parameter line to be connected with the driving circuit; and a FPGA to be connected with the distributed parameter line, wherein as a control signal of the optical switch to be input to the driving circuit, a LVDS signal from the FPGA is used.

14. The optical switch according to claim 2, wherein a terminating circuit for terminating a LVDS signal from the FPGA, and the driving circuit using a transistor having a deep ridge waveguide structure drive the 2×2 optical switch and the optical gate.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0015] FIG. 1 is a view showing a configuration of an N×N optical switch.

[0016] FIG. 2 is a view showing a configuration in which the 2×2 optical switches are arranged in a tree shape, and an optical gate is provided at each of N optical outputs.

[0017] FIG. 3 is a view showing a structure of a conventional optical switch, and a driving circuit of the optical switch.

[0018] FIG. 4(a) is a view for illustrating the operation of the driving circuit at the time of signal light blocking, and FIG. 4(b) is a view for illustrating the operation of the driving circuit at the time of signal light passing.

[0019] FIG. 5 is a view showing the structure of the optical switch in an embodiment of the present invention and the driving circuit of the optical switch.

[0020] FIG. 6 is a cross sectional view showing the layer structure of a HEMT in accordance with an embodiment of the present invention.

[0021] FIG. 7 is a graph showing the transmittance with respect to the injection current of a 2×2 MZI for use in an embodiment of the present invention.

[0022] FIG. 8 is a graph showing the transmittance with respect to the injection voltage of an optical absorption gate for use in an embodiment of the present invention.

[0023] FIG. 9 is a cross sectional view showing an optical waveguide forming an element of the optical switch in accordance with an embodiment of the present invention.

[0024] FIG. 10 is a view showing the structure of an optical switch in accordance with Example of the present invention, and the driving circuit of the optical switch.

DESCRIPTION OF EMBODIMENTS

[0025] Below, an optical switch in an embodiment of the present invention will be described by reference to the accompanying drawings.

[0026] FIG. 5 shows the optical switch driving structure in the present embodiment. As the optical switch driving driver, a driving circuit using a High Electron Mobility Transistor (HEMT) capable of implementing a high-speed voltage amplification circuit is used. As the form of a driver-integrated type optical switch 502, the structure is adopted in which for the optical switch, the driving circuits thereof are integrated in a high density in the vicinity of the control electrode of the optical switch. The integrated type may be implemented by integration onto the same InP substrate, or flipchip mounting or the like of individually manufactured optical switch chips and driver chips. In the present embodiment, using a LVDS transmission line 503 of a differential signal system using two transmission paths as an example of the distributed parameter line, the driver-integrated type optical switch 502 is operated. This operation enables high-speed signal transmission up to several GHz, low power consumption (3.5 mA driving, signal amplitude 350 mV), and high noise resistance (the same phase noise is cancelled by a differential signal).

[0027] The layer structure of the HEMT for use in the present embodiment is shown in FIG. 6. On a semi-insulating (semi insulator: SI) substrate 601, as a buffer layer 602, for example, an i-InAlAs-containing layer, and an i-InGaAs-containing layer on the i-InAlAs-containing layer are formed. As a channel layer 603, for example, from the lower layer, an i-InAlAs-containing layer, a Si δ-doped layer, an i-InAlAs-containing layer, and an i-InP-containing layer are sequentially formed. A structure is manufactured in which on the channel layer 603, a gate electrode 605g is formed, and a source electrode 605s and a drain electrode 605d are formed via a cap layer 604 including an n-InAlAs-containing layer, and an n-InGaAs-containing layer on the n-InAlAs-containing layer stacked therein. With the HEMT in accordance with the present embodiment, the gate length is set at 0.1 μm, and the gate width is set at 25 μm. The design values become important parameters for determining the characteristics of the HEMT. The gate length determines the response speed of the HEMT. However, it is known as follows: with a gate length of 0.1 μm as in the present embodiment, the operation in response to a high speed signal up to GHz is implemented. Further, it is possible to determine the value of the current flowing between the source and the drain according to the size of the gate width.

[0028] The optical switch structure is assumed to be the same as that shown in FIG. 2. The structure is adopted in which the 2×2 optical switches 201.sub.N using a MZI or the like are connected in a multistage, and an optical gate 202.sub.N is provided at the final stage.

[0029] The operation of the optical switch of the MZI will be described. Using a Multi-Mode Interference (MMI) type optical coupler with respect to an input light, two optical waveguides are multiplexed, and then branched into two optical waveguides. The two branched input lights undergo two waveguide phase difference, and then are coupled using the MMI optical coupler again. Then, due to the interference effect, when the phase difference between the two optical waveguides is ±nπ, the light is output from the optical waveguide opposite to the light input, and when the phase difference is ±(2n+1) π/2, light is output from the optical waveguide on the same side as that of the light input. Therefore, when a phase modulation area is arranged in one optical waveguide, and is controlled, the 2×2 switching operation can be obtained. In order to obtain phase modulation, it is essential only that the refractive index of the optical waveguide is changed. With an InP type optical waveguide, using the FK effect or the QCSE effect due to voltage application, or the plasma effect due to current injection, the refractive index of the optical waveguide is changed. With the LN type, using the Pockels effect due to voltage application, the refractive index of the optical waveguide is changed. As a result, a switching operation can be performed. Alternatively, for the MMI optical coupler for halving the optical strength, a directional coupler, or the like may be used. When the injection current to the two arm optical waveguides is 0 mA, the input optical signal is output to the cross port side of the 2×2 switch. As shown in FIG. 7, when a current is injected to any one control electrode (the control electrode set on the waveguide forming the MZI), the refractive index of the arm optical waveguide to which the current has been injected changes, so that the phase of the propagating light changes. When the injection current to the arm optical waveguide becomes about 5 mA, the output from the optical output port PO.sub.1 is minimized, and the optical output to the optical output port PO.sub.2 is maximized. At this step, a ratio of the optical output to the optical output port PO.sub.1 and the optical output to the optical output port PO.sub.2 of 20 dB or more can be obtained.

[0030] The operation of the optical gate 202.sub.N will be described. At the optical gate 202.sub.N provided at the final stage of the 2×2 optical switch 202.sub.N, blocking of the leaked light whose output is not desired is performed. With switching of the 2×2 optical switch, while most of the light is guided to a desirable port, a part of the light is leaked to different ports. The leaked light deteriorates the quality of the transmission signal, and hence is required to be minimized. In the case of the optical gate using the EAM of InP, a reverse bias voltage is applied between the p type electrode and the n type electrode. As a result, due to the Franz-Keldysh (FK) effect, the absorption end of the waveguide core is shifted, so that the absorption coefficient of the light propagating through the optical gate can be increased. As shown in FIG. 8, for example, when the application voltage is −3 V, an extinction ratio of 20 dB or more can be obtained. When the application voltage is −7 V, an extinction ratio of 40 dB or more can be obtained. As the optical gate 202.sub.N, a SOA which can be implemented by the same InP material, or the like may be used.

[0031] FIG. 9 shows the cross sectional structure of the waveguide. The waveguide forming the optical switch was manufactured by etching to the underlying portion of the InGaAsP-containing core layer 904, resulting in a deep ridge waveguide having a pin double hetero junction structure. The height of the waveguide was set at 4 μm, and the width thereof was set at 1.4 μm. A conventional optical switch is normally manufactured on an n substrate. In contrast, in the present embodiment, an optical switch is manufactured on a SI substrate, thereby isolating the substrate potentials of the MZI and the EAM. With the optical switch on an n substrate, the substrate back surface was the common electrode. However, by manufacturing an optical switch on a SI substrate with a high resistivity, insulation and isolation become possible.

[0032] A method for manufacturing an optical switch in accordance with the present embodiment will be described.

[0033] First, for example, on a SI-InP-containing semi-insulating substrate 901, five layers of an n.sup.+-InGaAsP-containing contact layer 902, an n-InP-containing lower cladding layer 903, a 1.4Ω composition 0.3 μm thick bulk i-InGaAsP-containing core layer 904, a P-InP-containing upper cladding layer 905, and a p′-InGaAsP-containing contact layer 906 are formed by growth by the Metal Organic Vapor Phase Epitaxy (MOVPE). Then, by photolithography and dry etching, a deep ridge waveguide structure is formed all together, thereby forming a trench reaching the SI substrate between the MZI area and the EAM area. Subsequently, Benzocyclobutene (BCB) of an organic material capable of being buried in a local area, and excellent in planarization is coated by spin coating, and is etched back until the substrate surface is exposed by RIE (Reactive Ion Etching) using an O.sub.2/C.sub.2F.sub.6 mixed gas, thereby planarizing the surface of the substrate. Then, by photolithography and dry etching, a well for forming an electrode for ground electrode is manufactured, and an n type electrode is formed. Finally, on the optical gate, a p type electrode is formed. As described up to this point, it becomes possible to perform MOVPE growth and the formation of the optical waveguide structure all together.

[0034] In the present embodiment, the 0.3 μm thick, 1.4 μm wide, and 1.4Ω composition InGaAsP-containing layer is used. The design values become the important parameters for determining the optical characteristics of the optical switch. In order to implement the operation at an input signal optical wavelength of, for example 1.53 μm to 1.57 μm, and a low loss, high speed, and low power consumption operation, the following conditions are preferably satisfied.

[0035] First, the thickness of the InGaAsP-containing core layer 904 preferably falls within the range of 0.1 μm to 0.4 μm, of the single mode guiding condition with respect to the input signal light, and the condition having sufficient light confinement to the InGaAsP-containing core layer 904.

[0036] Secondly, the width of the InGaAsP-containing core layer 904 desirably falls within the range of 0.8 μm to 3 μm, of the single mode guiding condition with respect to the input signal light.

[0037] Thirdly, from the viewpoint of reducing the power consumption of the driving circuit, desirably, the condition is such that the application voltage to the optical gate is 7 V or less, and the composition of the InGaAsP-containing core layer 904 is 1.3Ω to 1.5Ω, and each electrode length falls within the range of 100 μm to 2000 μm.

[0038] Although a description has been given that, for the optical switch in the present embodiment, as the InGaAsP-containing core layer of the optical gate 202.sub.N, the bulk layer is used, a multiple quantum well structure may be adopted. In that case, it becomes possible to perform quenching with high efficiency at the optical gate due to the Quantum Confined Stark Effect. Further, although the optical waveguide structure is set as a ridge waveguide structure, the optical waveguide structure may be manufactured as other structures, for example, a deep ridge type optical waveguide structure. Alternatively, a buried type optical waveguide structure in which both sides of the InGaAsP-containing core layer 904 are buried with a semiconductor, a rib type optical waveguide structure, or the like is also acceptable.

[0039] Although a description has been given using the InP type compound semiconductor in the present embodiment, a GaAs type compound semiconductor may be used. Alternatively, even using a material type capable of a change in refractive index and absorption coefficient of the nanosecond order such as a silicon wire waveguide, the present invention can be implemented similarly.

Example

[0040] FIG. 10 shows an example of an optical switch in accordance with Example of the present invention and the driving circuit thereof.

[0041] The configuration of the driving circuit will be described by reference to FIG. 10. A FPGA 501 is connected to a first LVDS transmission line 503a and a second LVDS transmission line 503b. The first LVDS transmission line 503a is connected to the gate of the transistor 1003 and one terminal of the terminating resistance. The drain of the transistor 1003 is connected to the gate of the transistor 1001 and one terminal of the resistance. Whereas, the second LVDS transmission line 503b is connected to the gate of the transistor 1004 and the other terminal of the terminating resistance.

[0042] The drain of the transistor 1004 is connected to the gate of the transistor 1002 and one terminal of the resistance. At the other terminal of the resistance, Vdd (2.2 v) is provided, and at the sources of the transistor 1003 and the transistor 1004, Vss (1.2 v) is provided. With the middle point between the transistor 1001 and the transistor 1002 as a totem-pole output, a MZI 201a and an EAM 202a are applied with a voltage.

[0043] The drain voltage V.sub.d, and the source voltage V.sub.s at the HEMT of the driving circuit can be set for every driving circuit. The signal from the FPGA 501 having an amplitude common to the MZI 201a and the EAM 202a is input to the driver-integrated type optical switch 502 by the LVDS transmission line 503. V.sub.b represents the bias potential of each n type electrode in the MZI and the EAM of the optical switch. From the structure of the optical switch shown in FIG. 9, the MZI and the EAM are individually manufactured, thereby enabling setting of different V.sub.b's between the MZI and the EAM. This enables driving at the common FPGA and driver circuit.

[0044] The operation of the driving circuit will be described by reference to FIG. 10.

[0045] A LVDS signal is a differential signal with an amplitude of 350 mV centered on 1.2 V. When the gate of the transistor 1003 is applied with 1.375 V, the gate of the transistor 1004 is applied with 1.025 V. At this step, the gate of the transistor 1001 is opened, and the gate of the transistor 1002 is closed. A current flows between the source and the drain of the transistor 1001, and the optical switch is applied with a voltage of V.sub.d. On the other hand, when the gate of the transistor 1003 is applied with 1.025 V, the gate of the transistor 1004 is applied with 1.375 V. At this step, the gate of the transistor 1001 is closed, and the gate of the transistor 1002 is opened. A current flow between the source and the drain of the transistor 1002, the optical switch is applied with a voltage of V.sub.s. Thus, by switching the LVDS differential signal, opening/closing of the transistor 1001 and the transistor 1002 is performed, thereby switching the voltage to be applied to the optical switch. As a result, the switching operation is implemented.

[0046] For example, in the driving circuit of a MZI 201a, it is set that V.sub.d=3 V, and V.sub.s=1 V. In the case where V.sub.b=1 V is applied, when the gate of the transistor 1001 is opened by the signal from the FPGA 501, the MZI 201a is applied with a potential difference of 2 V in the positive direction, allowing a current to flow therethrough. When the gate of the transistor 1002 is opened, a potential difference is not caused at the MZI 201a. For this reason, a current does not flow. Opening/closing of the gates of the transistor 1001 and the transistor 1002 adjusts the current of the MZI, thereby enabling switching of the output port.

[0047] On the other hand, in the driving circuit of the EAM 202a, it is set that V.sub.d=4.5 V, and V.sub.s=1 V. In the case where V.sub.b=4.5 V is applied, when the gate of the transistor 1001 is opened by the signal from the FPGA 502, a potential difference is not caused at the EAM 202a. For this reason, optical quenching is not performed. When the gate of the transistor 1002 is opened, a potential difference of 3.5 V is generated in the negative direction with respect to the EAM 202a. For this reason, quenching is performed. Opening/closing of the gates of the transistor 1001 and the transistor 1002 enables opening/closing of the gate of the EAM 202a.

[0048] The present embodiment provides a 1×N optical switch having a structure in which with respect to an optical switch, a driving circuit of the optical switch is integrated in the vicinity of a control electrode of the optical switch. The optical switch includes a plurality of 2×2 optical switches and N optical gates. Different bias voltages (V.sub.b) are set between the optical switch and the optical gates, and a driver for the 2×2 optical switch of the driving circuit and a driver for the optical gate are of the same circuit form.

[0049] Further, the present embodiment provides a 1×N optical switch having a structure in which with respect to an optical switch, a driving circuit of the optical switch is integrated in the vicinity of a control electrode of the optical switch. The optical switch includes a plurality of 2×2 optical switches and N optical gates. A power supply voltage (V.sub.d−V.sub.s) of a driver for the 2×2 optical switch of the driving circuit and a power supply voltage of a driver for the optical gate are equal.

[0050] In the 1×N optical switch, the MZI and the EAM are individually manufactured, thereby enabling setting of different V.sub.b's between the MZI and the EAM. This enables control using the configuration of the same FPGA and the same driving circuit. For this reason, high density integration of the driving circuits becomes possible, so that the high speed operation of the optical switch is implemented. Further, using a LVDS signal with a small amplitude and a low power consumption, the optical switch can be directly driven, and individual power supply, or the like is not required for the MZI and the EAM. For this reason, driving at a low power consumption is possible.

INDUSTRIAL APPLICABILITY

[0051] The present disclosure is applicable to a technical field of an optical switch element of an important optical component for supporting a large capacity optical communication network.