WO2003058286A2

WO2003058286A2 - Monolithic optical control components

Info

Publication number: WO2003058286A2
Application number: PCT/IL2003/000031
Authority: WO
Inventors: Dan Haronian; Ran Vered; Eitan Efron; Rahav Cohen
Original assignee: Galayor Networks Inc.
Priority date: 2002-01-10
Filing date: 2003-01-10
Publication date: 2003-07-17
Also published as: AU2003235773A1; WO2003058286A3; AU2003235773A8

Abstract

Novel optical waveguide components, operated by means of micro-actuators which move suspended section of waveguides, and especially variable optical attenuators, and optical couplers. Methods are also described for aligning and latching the micro-actuators in two or three dimensions, such that settings of variable attenuators can be maintained. Such micro-actuators are also used for ensuring good alignment between the various waveguide and fiber ports in integrated optical circuits. Additional components based on micro-actuators include multi-pole switches, digital variable optical attenuators, receiver input protectors, and multifunctional line protection chips. The components and systems described can be executed in monolithic form, thus engendering significant cost and space savings. Furthermore, methods of substrate etching to obtain accurately vertical faces are also described.

Description

MONOLITHIC OPTICAL CONTROL COMPONENTS

FIELD OF THE INVENTION

The present invention relates to the field of fiber optical communication system components, and especially to control components based on micro-mechanical actuators and executed in monolithic form.

BACKGROUND OF THE INVENTION

Variable optical attenuators are basic elements of many fiber optical communication networks. Because of the nature of such systems, it is of importance that its component parts be as miniaturized as possible, and in general, monolithic components are most desired where possible.

There are known in the prior art mechanical fiber optical variable optical attenuators (VOA's), such as are described in U.S. Patent No. 5,031,994, to D. Emmons for "Optical Switch Array" and in U.S. Patent No. 6,456,775 to W. Johnson et al., for "Optical Fiber Attenuator". In these and other similar Patents, two fibers are disposed with their cleaved ends in close proximity. The end of one of the fibers is moved laterally by means of a step motor, such that misalignment is created between the two fiber ends. This misalignment controls the attenuation of the light passing between the two fibers. The attenuation is determined by monitoring the intensities typically through 1:99 splices at the input and output of the device, by means of photodiodes. Such arrangements are generally executed only in hybrid or even hybrid/discrete format, and thus cannot be miniaturized by the use of all monolithic construction.

In US Patent No. 5,727,099, to M.R. Harman for "Positioning System for Controlling Optical Alignment of Optical Waveguides", there is described a similar system to the above described, using waveguides clamped in a cantilevered, spring-loaded assembly, adjusted by means of a thumb screw. This system too is unsuited for use in true integrated optics applications.

There also exist in the prior art, micro-mechanical fiber optical VOA's, such as are described in US Patent No. 6,173,105 and in U.S. Patent No. 6,075,239, in which a shutter is micro-mechanically moved in steps in a gap between the ends of two fibers positioned opposite each other and in close proximity, such that the light passing between them is attenuated according to the position of the shutter. These VOA's also have the disadvantage that they cannot be readily implemented in monolithic form as the power monitoring is performed in a manner similar to that of the mechanical VOA described above.

There also exists in the prior art, such as is described in U.S. Patent No. 6,275,320 to Dhuler et al., a shutter based VOA, with an electrostatic clamp to maintain the VOA setting. However, the clamping mechanism described therein requires to be powered all the time to maintain its action, such that in the event of a power removal, the VOA may change its setting.

Optical switches using bendable waveguides are described in the above-mentioned US Patent No. 5,727,099, to M.R. Harman, and in US Patent No. 5,078,514 to S. Valette et al., for "Switch and System for Switching Integrated Optical Multichannels and Switch Protection Method". In these patent documents, there are described waveguide switches which bend a waveguide such that its end can transfer optical energy to the end of one of multiple output waveguides disposed opposite the positions of bending of the input waveguide.

This prior art does not lend itself readily to fabrication in monolithic form on substrates, to produce highly miniaturized integrated optical circuits and systems, and there therefore exists a need for such components capable of such miniaturization by means of monolithic fabrication, preferably in the commonly used silicon system.

The disclosures of each of the publications mentioned in this section and in other sections of the specification, are hereby incorporated by reference, each in its entirety. SUMMARY OF THE INVENTION

The present invention seeks to provide new optical waveguide components, operated by means of micro-actuators, and especially variable optical attenuators, and optical couplers. The components and systems provided according to the present invention differ from much of the prior art in that they can be executed in monolithic form, thus engendering significant cost and space savings.

There is thus provided in accordance with a preferred embodiment of the present invention, a variable optical attenuator (VOA) comprising a suspended first waveguide section having a first end, a fixed second waveguide section having a second end disposed generally opposite the first end and in close proximity thereto, and a micro-actuator attached to the first waveguide section imparting lateral motion to the waveguide, such that change of alignment of the first end with the second end changes the attenuation of light traversing between the first and second waveguide sections.

In accordance with yet another preferred embodiment of the present invention, the VOA preferably also comprises a first coupler sampling light traversing the first waveguide, a second coupler sampling light traversing the second waveguide, and detectors detecting the sampled light. Eeven more preferably, it also comprises control circuitry receiving inputs from at least one of the detectors and providing control signals for the micro-actuator. At least the first and second waveguides, the micro-actuator and the couplers are preferably fabricated monolithically on a substrate, and either the control circuitry or the detectors or both can also be fabricated monolithically on the substrate.

Additionally, in the above described VOA, the first end and the second end are separated by a gap, and the variable optical attenuator may also comprise a second micro-actuator attached to the first waveguide section and imparting longitudinal motion thereto to change the gap, thus altering the attenuation of light traversing between the first and second waveguide sections. The micro-actuator may preferably be controlled to change the gap so as to reduce interference effects of light reflected between the ends of the waveguides or to reduce temperature change effects on the gap or other parts of the attenuator.

There is further provided in accordance with still another preferred embodiment of the present invention, an optical switch, comprising an input waveguide section having a first end, an output waveguide section, having a second end, a rigid suspended transfer waveguide section having a micro- actuator attached thereto, imparting lateral motion to the rigid suspended transfer waveguide section such that it switches between a bridging position between the first and second ends, such that light passes between the input and output waveguides, and a non-bridging position, such that light cannot pass between the input and output waveguide sections.

In the above-mentioned switch, the rigid suspended transfer waveguide section preferably has third and fourth ends, and in the bridging position, the third end is disposed opposite to and in close proximity to the first end, and the fourth end is disposed opposite to and in close proximity to the second end, so that light can pass between the input and output waveguide sections. Conversely, in the non-bridging position, the third end is disposed laterally distant from the first end, and the fourth end is disposed laterally distant from the second end, so that light cannot pass between the input and output waveguide sections. In the above described optical switch, the input waveguide section, the output waveguide section, the rigid suspended transfer waveguide section and the micro-actuator are preferably fabricated monolithically on a substrate.

In accordance with a further preferred embodiment of the present invention, there is also provided a multi-way optical switch, comprising an input waveguide section having a first end, a plurality of output waveguide sections, each having a second end, and a plurality of rigid suspended transfer waveguide sections, one for each of the plurality of output waveguide sections, the plurality of transfer waveguide sections having at least one micro-actuator attached, imparting lateral motion between predetermined positions to at least one of the transfer waveguide sections, such that at each of the predetermined positions, one of the transfer waveguide sections bridges between the first end of the input waveguide section, and the second end of one of the plurality of output waveguide sections, such that light passes between the input waveguide section and one of the plurality of output waveguide sections. The input waveguide section, the output waveguide sections, the rigid suspended transfer waveguide sections and the at least one micro-actuator are preferably fabricated monolithically on a substrate.

There is even further provided in accordance with a preferred embodiment of the present invention, a variable optical attenuator comprising an input waveguide section, an output waveguide section, a suspended waveguide section connecting the input and output waveguide sections, a micro-actuator attached to the suspended waveguide section imparting distortion to the suspended waveguide section from its initial shape, such that high order modes are excited in light traversing through the suspended waveguide section, and a serial component which attenuates the propagation of higher order modes, such that the high order modes excited in light traversing through the suspended waveguide section are attenuated. The distortion may be at least one of lateral bending distortion and torsional distortion. The serial component may be a single mode output fiber, and the suspended waveguide section may have at least one bend in its path. Furthermore, the high order modes may preferably be excited by geometrical distortion in the propagation path of light passing therethrough or by photoelastic effects on the material of the suspended waveguide section.

Furthermore, in accordance with yet another preferred embodiment of the present invention, there is provided an optical receiver protection device, comprising a receiver detector receiving an optical signal, an input waveguide supplying the optical signal, a coupler for extracting a sample of the optical signal, a monitor detector receiving the sample of the optical signal, and providing an electronic signal corresponding thereto, a variable optical attenuator controlled by a micro-actuator disposed between the input waveguide and the receiver detector, the micro-actuator being controlled by the electronic signal, such that when the electronic signal is indicative of an excessive optical signal in the input waveguide, the micro-actuator is operative to attenuate the input optical signal. The input waveguide, the coupler, and the variable optical attenuator are preferably fabricated monolithically on a substrate.

Furthermore, in the optical receiver protection device, the variable optical attenuator preferably comprises a suspended first waveguide section and a fixed second waveguide section, the ends of the sections being mutually laterally positioned by means of the micro-actuator, the micro-actuator being attached to the first waveguide section, such that the attenuation of the attenuator is varied. Additionally, the variable optical attenuator comprises a suspended waveguide section connected between the input waveguide and the receiver detector, the micro-actuator imparting distortion to the suspended waveguide section, such that high order modes are excited in the optical signal traversing through the suspended waveguide section, such that the attenuation of the attenuator is varied. In general, in such an optical receiver protection device, the receiver detector is significantly more costly than the other components, such that its protection with lower cost components is very cost effective.

There is also provided in accordance with a further preferred embodiment of the present invention, a monolithic in-line optical power monitor, comprising an input fiber conveying an optical signal to be monitored, an output fiber outputting the optical signal after monitoring, a waveguide section connecting the input fiber and the output fiber, a coupler disposed within the path of the waveguide, and sampling a part of the optical signal, and a detector measuring the sampled part of the optical signal, and providing an electrical signal in accordance with the intensity of the optical signal, wherein at least the waveguide section and the coupler are fabricated monolithically on a substrate. Alternatively and preferably, the detector is also fabricated monolithically on the substrate.

In accordance with yet another preferred embodiment of the present invention, there is provided a digitally controlled variable optical attenuator, comprising a fixed waveguide receiving an input optical signal, a plurality of suspended waveguide sections disposed serially along the fixed waveguide, and spaced therefrom, micro-actuators operative to move the suspended waveguide sections towards the fixed waveguide, such that the suspended waveguide sections attenuate the optical signal propagating in the fixed waveguide by evanescent field interaction, a coupler for extracting a sample of the optical signal after passing the plurality of suspended waveguide sections, and a monitor detector receiving the sample of the optical signal, and providing an electronic signal for controlling the micro-actuators, such that attenuation of the input optical signal is digitally determined. In this digitally controlled variable optical attenuator, the micro-actuators are preferably such as to move the suspended waveguide sections either into close proximity with the fixed waveguide, or into contact therewith.

Furthermore, the electronic signal for controlling the micro-actuators is generally processed to operate a combination of micro-actuators which generate a desired level of attenuation. In addition, in the above described digitally controlled variable optical attenuator, individual ones of the plurality of suspended waveguide sections generally have different attenuating effects.

There is further provided in accordance with yet another preferred embodiment of the present invention, a digitally controlled variable optical attenuator, comprising a fixed waveguide receiving an input optical signal, a plurality of suspended waveguide sections disposed serially along the fixed waveguide, and spaced therefrom, micro-actuators operative to move the suspended waveguide sections towards the fixed waveguide, such that the suspended waveguide sections attenuate the optical signal propagating in the fixed waveguide by evanescent field interaction, and wherein the micro-actuators generally have differing stiffnesses, a coupler for extracting a sample of the optical signal after passing the plurality of suspended waveguide sections, and a monitor detector receiving the sample of the optical signal, and providing an electronic signal corresponding to the optical signal for applying to the micro-actuators, such that attenuation of the input optical signal is digitally determined. Individual ones of the suspended waveguide sections generally have the same attenuating effect, and the electronic signal is generally applied to all of the micro-actuators, and is operative to actuate those micro-actuators having a stiffness up to that overcome by the electrical signal, such that a desired level of attenuation is generated.

In this digitally controlled variable optical attenuator, the suspended waveguide sections are preferably divided into groups, each group covering a different range of attenuation, and each group having a common applied voltage, each of the suspended waveguide sections in each group generally having a range of attenuating effects, and the suspended waveguide sections having stiffnesses corresponding to their attenuating effect, and the electronic signal is preferably processed into different signals for application separately to all of the micro-actuators in each of the groups, and is operative to actuate those micro-actuators in each group having a stiffness up to that overcome by the different signals, such that a desired level of attenuation is generated. Each of the groups preferably covers a decade of attenuation.

In accordance with still another preferred embodiment of the present invention, there is provided apparatus for the transfer of optical signals between different terminal ports of an optical system, comprising a connecting waveguide having ends for connecting between two of the different terminal ports, and at least one micro-actuator mechanism attached to at least one end of the connecting waveguide, the micro-actuator mechanism being adjustable in at least one dimension to manipulate the position of the end of the connecting waveguide relative to at least one of the terminal ports, wherein at least the connecting waveguide and the micro-actuator mechanism are constructed monolithically on a substrate. The micro-actuator mechanism may also be adjustable in two dimensions or three dimensions.

Furthermore, the at least one micro-actuator mechanism is preferably lockable, such that the position of the end of the connecting waveguide can be fixed in its optimum position. The micro-actuator mechanism is also preferably adjusted while observing the transfer of optical signals between different terminal ports of the optical system to optimize the transfer. This transfer is preferably optimized for at least one of maximum optical power transfer, minimum insertion loss, and minimum power reflection. Alternatively, transfer may be optimized by viewing the physical location of the end of the connecting waveguide relative to the terminal port of the optical system. Finally, the terminal ports may be a fiber V-groove interface, the waveguide of an optical chip, an integrated optics connector port, a laser source output, or a hybrid component connector port.

There is further provided in accordance with still another preferred embodiment of the present invention, apparatus for locking the position of the operating arm of a micro- actuator, comprising a first clamping plate on one side of the operating arm, the first plate being movable by means of a first micro-actuator, operative when powered to pull the plate away from the arm, and a second clamping plate on an opposite side of the operating arm, the second plate being movable by means of a second micro-actuator, and which is initially held away from the arm by means of a latching plate held behind a shutter, and wherein the second micro-actuator is activated to pull the latching plate irreversibly through the shutter, such that the second clamping plate is latched close to the operating arm, such that when the first micro-actuator is released, the first clamping plate locks the operating arm against the second clamping plate.

In the above mentioned apparatus for locking the position of the operating arm of a micro-actuator, the shutter is preferably opened by the operation of a third micro-actuator in order to allow the second micro-actuator to pull the latching plate through the shutter. Additionally, the locking is maintained without the application of voltages to any of the micro-actuators. Any of the micro-actuators may be electrostatically operated or thermally operated.

In accordance with a further preferred embodiment of the present invention, there is also provided a thermal actuator for providing motion out of the plane in which the actuator is disposed, comprising a beam conveying an operating current, the beam having a doping profile, such that its electrical conductivity varies with depth, such that the operating current causes the beam to bend relative to the plane.

There is also provided in accordance with yet a further preferred embodiment of the present invention, a thermal actuator for providing motion out of the plane in which the actuator is disposed, comprising a beam conveying an operating current, the beam being clamped at an end, and having a relieved profile on one edge, such that as it heats up under the influence of the operating current, the beam bends relative to the plane in the direction of the relieved profile.

There is even further provided in accordance with a preferred embodiment of the present invention, a method of improving the verticality of faces in a monolithic structure fabricated in a substrate, comprising the steps of generating an approximately vertical face in the structure by means of reactive ion etching, wet etching the vertical face in order to expose preferentially vertical crystallographic planes in the substrate. In this method, the improvement of verticality of faces is operative to improve the parallelism of faces of optical gaps or other planes in monolithic waveguide structures, such that the efficiency of transfer of light across the faces is improved. The substrate may be a silicon substrate having a 110 plane of orientation, and the wet etching step is then operative to expose the 111 planes of the substrate, the 111 planes being accurately perpendicular to the 110 plane of orientation of the substrate. Alternatively and preferably, the substrate is a silicon substrate having a 010 plane of orientation, and the wet etching step is operative to expose the 100 planes of the substrate, the 100 planes being accurately perpendicular to the 010 plane of orientation of the substrate. In either case, the wet etching is preferably performed by a potassium hydroxide solution, after protection of areas not to be wet etched by means of a protective mask.

Furthermore, in accordance with yet another preferred embodiment of the present invention, there is provided a multifunctional line protection chip, comprising a substrate comprising a coupler for sampling an input signal, an input monitoring photodiode detecting the sampled input signal, a two way, one pole optical switch, a variable optical attenuator, and a transceiver for detecting the input signal, wherein the variable optical attenuator is controlled by the output of the photodiode, such that a signal which would saturate the transceiver is attenuated., and wherein at least the coupler, the two way, one pole optical switch and the variable optical attenuator are fabricated monolithically. Preferably, the input monitoring photodiode is also fabricated in the substrate. Even more preferably, the multifunctional line protection chip also comprises monolithically fabricated ring resonators, operative for filtering wavelengths of light from the input signal, the wavelengths being selected by means of micro-actuators attached to the rings. Alternatively and preferably, these filters can be monolithically fabricated Mach-Zehnder interferometric filters, for filtering wavelengths of light from the input signal, the wavelengths being selected by means of micro-actuators attached to arms of the Mach-Zehnder filters.

There is also provided in accordance with a further preferred embodiment of the present invention, a light coupler comprising a multimode interferometer comprising two output waveguides disposed at an exit plane of the multimode interferometer, and wherein the multimode interferometer is such as to have interference peaks coupled out at the exit plane having predefined intensities, one peak containing transmitted light at one of the output waveguides, and the other, a small fraction of the transmitted light, at the second output waveguide, and wherein the light coupler is fabricated monolithically on a substrate, and is utilized in any of the preferred embodiments of the present invention where a monolithic coupler is required.

In accordance with yet another preferred embodiment of the present invention, there is provided a variable optical coupler, comprising a suspended first waveguide section having a first end, a fixed second waveguide section having a second end disposed generally opposite the first end and in close proximity thereto, a micro-actuator attached to the first waveguide section imparting lateral motion to the waveguide, such that change of alignment of the first end with the second end changes the attenuation of light traversing between the first and second waveguide sections, and a fixed third waveguide section disposed with its end close to the gap between the first end and the second end, such that the third waveguide section collects coupled light from the first waveguide section not transferred to the second waveguide section. The level of the coupled light generally increases with increased attenuation of light traversing between the first and second waveguide sections. In this variable optical coupler, the light coupler is preferably fabricated monolithically on a substrate.

There is further provided in accordance with yet another preferred embodiment of the present invention, a variable optical coupler, comprising a suspended first waveguide section having a first end, a fixed second waveguide section having a second end disposed generally opposite the first end and in close proximity thereto, and a micro-actuator attached to the first waveguide section imparting lateral motion to the waveguide, such that change of alignment of the first end with the second end changes the attenuation of light traversing between the first and second waveguide sections, wherein the fixed second waveguide section has a discontinuity in its path, at which is connected an output waveguide operative to sample the light traversing the discontinuity. In this variable optical coupler, the discontinuity is preferably disposed such that it samples light from the center of a zero order mode propagating in the second waveguide section. The discontinuity may be a right angle bend, and the output waveguide is connected at the center of the apex of the right angle bend.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be understood and appreciated more fully from the following detailed description, taken in conjunction with the drawings in which:

Fig.l illustrates schematically a VOA, constructed and operative according to a first preferred embodiment of the present invention; Fig. 2 illustrates how the light is coupled between the two waveguides of the embodiment shown in Fig. 1;

Figs. 3A and 3B are schematic illustrations of alternative methods of actuating the waveguide movement in the VOA of Fig. 1, using bridge connections between the waveguides and the actuators;

Fig. 4 is a schematic drawing of a VOA, constructed and operative according to another preferred embodiment of the present invention;

Fig. 5 schematically illustrates a 1 x 2 optical switch, constructed and operative according to a preferred embodiment of the present invention;

Fig. 6A and Fig 6B schematically illustrate multi-pole switches; Fig 6A illustrates a 1 x 3 optical switch, similar in construction and operation to the 1 x 2 switch shown in Fig. 5, and Fig. 6B schematically illustrates a degenerate 2 x 2 optical switch, constructed and operative according to another preferred embodiment of the present invention;

Fig. 7 schematically illustrates a Guided Mode Mixing VOA, constructed and operative according to another preferred embodiment of the present invention;

Fig. 8 is a schematic diagram of a receiver variable optical attenuator constructed and operative according to a preferred embodiment of the present invention;

Fig. 9 is an optical intensity monitor constructed and operative according to yet another preferred embodiment of the present invention;

Fig. 10 is a digital variable optical attenuator (DVOA), constructed and operative according to a preferred embodiment of the present invention;

Fig. 11 is a DVOA, constructed and operative according to another preferred embodiment of the present invention;

Fig. 12 is a graph showing the relationship between the applied voltage V and the stiffness k_n overcome by that voltage, for the DVOA of Fig. 11; Fig. 13 shows a DVOA, constructed and operative according to yet another preferred embodiment of the present invention, and which overcomes the problem of the large number of actuators in the embodiment of Fig. 11;

Fig. 14 is a schematic drawing of an optical motherboard system, constructed and operative according to a preferred embodiment of the present invention;

Fig. 15 is a schematic drawing of a micro-actuator latching and locking system, according to another preferred embodiment of the present invention;

Fig. 16 is a representation of a monolithic pattern of the schematic latching mechanism shown in Fig. 15, to illustrate a preferred example of a real-life implementation of the embodiment shown in Fig. 15;

Fig. 17 is a schematic view of a detailed pattern in a monolithic substrate, according to a further preferred embodiment of the present invention, for executing motion in two dimensions;

Fig. 18A and 18B are schematic drawings illustrating thermal actuators for out-of-plane motion; Fig. 18A utilizes a doping profile to achieve such bending, while Fig. 18B utilizes a castellated profile cut into one surface of the actuator beam;

Fig. 19A is a schematic drawing of a dynamic fiber ribbon connector, constructed and operative according to another preferred embodiment of the present invention, as an example of an application of the three-dimensional actuator embodiment of Fig. 17;

Fig. 19B illustrates schematically a complete Optical Mother Board 0MB with integrated optics assembly, constructed and operative according to another preferred embodiment of the present invention;

Figs. 20 A to 20C illustrate schematically stages in RIE etching followed by wet etching of a (110) plane silicon wafer to produce accurately vertical and parallel walls, according to another preferred embodiment of the present invention; Fig. 21 illustrates schematically an industry standard (100) wafer, with the desired etched features having a 45° rotation to the {110} wafer locating edge;

Fig. 22 illustrates schematically one of the features of Fig. 21 in magnification, showing simulation results of the exposed planes in the feature, selectively etched with a KOH wet etch, and using a rectangular 45°-tilted mask;

Fig. 23 illustrates the rate of progress of the various etching steps according to the procedure shown in Figs. 21 and 22;

Figs. 24A to 24E illustrate the progressive stages, using the process described in Figs. 21 to 23, in etching a complete feature into a silicon on insulator (SOI) substrate;

Fig. 25A is a schematic drawing of a monolithic line protection chip (LPC), constructed and operative according to another preferred embodiment of the present invention;

Fig. 25B shows preferred embodiments of ring filters, which can be integrated onto the same substrate as the rest of the OADM;

Fig. 26 is a schematic drawing of a 1% output coupler using a multimode interferometer (MMI), constructed and operative according to another preferred embodiment of the present invention;

Fig. 27 shows a further variable coupler, constructed and operative as part of a VOA, according to another preferred embodiment of the present invention; and

Fig. 28 shows a further schematic VOA embodiment, constructed and operative according to yet another preferred embodiment of the present invention.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

Reference is now made to Fig. 1, which illustrates schematically a variable optical attenuator (VOA), constructed and operative according to a first preferred embodiment of the present invention. In the embodiment shown in Fig. 1, the light from the input fiber 10, is directed by means of an adapter section 12, to a suspended waveguide 16 section, preferably having an abrupt well-defined end 18. Opposite this end, and in close proximity thereto, is disposed another waveguide 20, connected to the output fiber 22. One of the waveguides, in the example shown in Fig. 1 the input waveguide 16, is preferably connected to a MEMS micro-actuator 24, that can move the waveguide end laterally to misalign it from head-on alignment. This misalignment controls the attenuation of the light passing between the two fibers. The ends of the waveguide, both the free ends and those attached to the fibers, are generally coated with an anti-reflection coating 14.

The input light power is preferably detected by coupling out 25 a small portion of the input light, typically 1%, and detecting it preferably by means of an on-chip or a hybrid-mounted photodiode 26. Likewise, the output light is preferably determined by using a second coupler 27 to direct a sample of the output light, also typically 1%, to an on-chip or hybrid photodiode 28. Signals from both of these detectors are used by the control circuits 29 to set the attenuation iteratively to the value required, according to user commands.

An important advantage of the VOA of the present invention is that the waveguides 12, 20, the MEMS actuator 24, the couplers 25, 27, and the photo- detectors 26, 28 can all be constructed on-chip, such that the entire VOA can preferably be of monolithic construction. Furthermore, the control circuits 29 can also preferably be constructed on the same chip, such that a complete, digitally controlled, monolithic VOA is thus attained. The fabrication process for such devices is preferably performed using a Spin-On-Dopant (SOD) process on Silicon-On-Insulator (SOI) substrate, such as is described in the article entitled "A Novel Boron Spin-On Dopant," by B. Justice, et al., published in "Solid State Technology," October, 1984, pp 153-15, and in the article "Self-Alignment of Optical Fibers with Optical Quality End-Polished Silicon Rib Waveguides using Wet Chemical Micromachining Techniques" by M.A. Rosa et al., published in IEEE Journal of Selected Topics in Quantum Electronics, Vol. 5, No. 5, pp.l249-1254 (1999). Reference is now made to Fig. 2, which illustrates how the light is coupled between the two waveguides for low attenuation settings of the VOA. The waveguides used can preferably be multi-mode, such that when they are slightly misaligned, zero-order mode light in the input waveguide 16, as shown schematically by the wavefront 30, excites higher order modes in the output waveguide 20, as shown schematically by the wavefront 32. Since the fibers are generally single mode fibers, and can only transmit the zero order mode, as shown schematically by the wavefront 34, the higher order modes 32 are rejected from entering the second fiber, and thus the intensity drops, according to the extent of waveguide end misalignment. For this reason, the VOA of the present invention is known as a VOA based on mode mixing. Once the waveguide is shifted significantly, the large misalignment between the two cross sections of the two waveguides increase illustrates schematically an industry standard (100) wafer 215, and with the desired waveguide features having a 45° rotation to the {110} wafer locating edges, and reduces overall propagation of light across the gap, resulting in the higher attenuation loss range of the VOA.

Reference is now made to Figs. 3A and 3B, which are schematic illustrations of alternative and preferred methods of actuating the waveguide movement to adjust the level of attenuation in the VOA's of the present invention. In these embodiments, instead of the actuator operating directly on the input waveguide 16 or the output waveguide 20, the waveguides are connected to each other through a bridge, and it is the bridge that is connected to the actuator 24. Upon deflection of the bridge, the angle between the two waveguides changes, leading to attenuation. Two preferred options are shown. In Fig 3A is shown an embodiment with two bridges 36, while in Fig. 3B is shown a one-bridge 38 embodiment. The form of bridge attachment to the waveguides can be used to control the center of rotation of the waveguide ends, and thereby, the degree of the attenuation. In both of these embodiments, a stiffening member is preferably provided to prevent undesired motion of the waveguide ends out of the plane of motion of the micro-actuator. The VOA's according to these various embodiments of the present invention have a number of additional advantages over prior art VOA's. In the VOA of the present invention, non-similar waveguides can preferably be used for input and output. For example, the input waveguide can be narrower than the output waveguide, thus allowing better control of the attenuation parameters.

Furthermore, although the waveguides described in the embodiments above are multimode waveguides, these VOA's and any of the other preferred embodiments of the present invention described hereinbelow, can be fabricated in few-mode or even single-mode waveguide, by using rib waveguide structures, as described in U.S. Patent No. 5,078,516 to E. Kapon et al., for "Tapered Rib Waveguides".

Reference is now made to Fig. 4, which is a schematic drawing of another VOA implementation, constructed and operative according to another preferred embodiment of the present invention. The attenuation is varied by adjusting the lateral position of the end of input waveguide 40 relative to the end of the output waveguide 42 by means of micro-actuator Al, as in the previous embodiment shown in Fig. 1. However, because of the parallelism of the ends, and because of the small dimensions of the gap between the ends, which is typically of the order of a few microns, interference effects are generated between the light transmitted across the gap, and any light reflected from the waveguide ends. Such effects cause the intensity of the light transmitted across the gap to show the typical interference pattern of a sinusoidal variation as a function of gap width and of wavelength. For the gap dimensions typically used in such a VOA, and for the reflectivity of the silicon preferably used in such a VOA, this variation can be as much as ±5db, which is reflected directly on the value of the insertion loss of the VOA at its minimum setting, and is thus unacceptable, not only because of the absolute level which it reaches, but also because this level itself is changeable, depending on small production variation of the gap size, changes in the gap size during use, changes in the ambient temperature which could cause change in the gap or in any other part of the attenuator, and changes in the wavelength of the light that is being attenuated. The use of anti-reflection (AR) coatings 14 on the fiber ends, as preferably used in the embodiment of Fig. 1, can reduce this variation to less than ±0.5db, depending on the quality of the AR coating. However, for many applications, this level too needs to be reduced.

According to the preferred embodiment shown in Fig. 4, such interference effects are overcome by the use of a second microactuator A2, which is operative to adjust the gap size itself, by moving an end of the waveguide longitudinally, such that it is closer to or further from the other end. The use of A2 is mandated when it is necessary to get to low attenuation settings. To achieve attenuations of from approximately 2db, depending on the efficiency of the AR coating, to the VOA maximum, the attenuation is set by sampling through the 1:100 couplers, the input and output power, or at least the output power, as described in relation to the embodiment of Fig. 1, while adjusting Al to reach the desired attenuation level. For attenuations of less than approximately 2db, or, if the AR coating is very efficient, for even lower values of attenuation, the microactuator A2 is then actuated while monitoring the input and output power, or at least the output power, and the interference effects are reduced by adjustment of the gap size, until the attenuation reaches the desired point. By adjusting A2 to use points of constructive interference in the gap, the VOA according to this embodiment is thus able to reduce the insertion loss of the VOA to almost zero, to within 0.1 to 0.2db.

Furthermore, since the microactuator A2 is able to reduce effects arising from light reflected from the waveguide ends, application of an anti-reflection coating to the waveguide ends is not essential. If an anti-reflection coating is not used, the microactuator Al alone is preferably used for attenuations of over 4db, and A2 to achieve values below 4db. Since application of an anti-reflection coating is not a simple procedure on such a small scale, this VOA can thus be constructed more economically than that of Fig. 1. Furthermore, the use of active gap manipulation enables production tolerances of the VOA to be eased, also reducing production costs and enabling compensation for ambient temperature changes. Although operation in closed loop enables compensation for temperature changes in the set attenuation level, the zero attenuation insertion loss specification may be degraded by temperature change, and VOA's constructed according to this embodiment of the present invention enable such degradation with temperature to be overcome.

In addition, according to further preferred embodiments of the present invention, it is possible to fabricate on-chip, arrays of such moving gap VOA structures, that can be used in multi-channel applications. Reference is now made to Fig. 5, which illustrates a 1 x 2 optical switch, constructed and operative according to a preferred embodiment of the present invention. This embodiment is shown in detail to illustrate the patterns preferably used on a Silicon On Insulator (SOI) substrate, or another suitable substrate system used to implement such a component. The switch preferably uses only rigid waveguide elements which are moved in unison by means of micro-actuators.

The input waveguide 50 terminates at an end 51, which delineates one side of a gap similar to that of the above-described embodiments of the VOA of the present invention. A first microactuator 52 is operative to move one section of bent waveguide 54, known as a transfer waveguide, whose first end 55 can move opposite the end 51 of the input waveguide and in close proximity to it. Movement of the microactuator is operative to either bring the end 55 of the transfer waveguide opposite the input waveguide to transfer input light thereto, or to move it away from the coupling position. The other end 56 of the transfer waveguide 54 is disposed opposite and in close proximity to the end 57 of a first output waveguide 58. The microactuator arm 59 is attached to the transfer waveguide by means of a bridge structure at a number of points 60, to maintain its dimensional stability.

A second microactuator 62 is preferably operative to move a second section of bent transfer waveguide 64, whose first end 65 can move opposite the end 51 of the input waveguide and in close proximity to it. The other end 66 of the transfer waveguide 64 is disposed opposite and in close proximity to the end 77 of a second output waveguide 68.

According to a first preferred embodiment of this switch array, the two microactuators are operated in unison in the same direction and between two predefined extreme positions, such that both transfer waveguides 54, 64, also move in unison in the same direction and between two predefined extreme positions. The ends of the transfer waveguides are disposed such that concurrent motion of both of the transfer waveguides is operative, when the waveguides are in the first of their predefined end positions (the lower position in the drawing), to guide light from the input waveguide 50, across the gap between ends 51 and 55, through transfer waveguide 54, across the gap between ends 56 and 57 to output waveguide 58. When the waveguides are in their other predefined end position, the upper position in the drawing, light is guided from the input waveguide 50, across the gap between ends 51 and 65, through transfer waveguide 64, across the gap between ends 66 and 67 to output waveguide 68. In Fig. 5, for clarity, the waveguides are shown in their resting between the two switch positions, although this position generally has no utilizable function in a switch embodiment. Furthermore, although two actuators are shown in the embodiment of Fig. 5, such a switch could also be implemented using a single actuator, provided that suitable connecting structures are provided to support the transverse waveguides in a mutually fixed position.

Alternatively and preferably, according to a second preferred embodiment, each of the switches of Fig. 5 can be operated independently by its own microactuator, such that motion of transfer waveguide 54, connecting input waveguide 50 with output waveguide 58 is performed independently of motion of transfer waveguide 64, connecting input waveguide 50 with output waveguide 68. This embodiment thus allows the construction of rigid waveguide, single pole l l switches. Though based on similar geometry to the VOA's described in the embodiments of Figs. 1 to Fig. 3, and constructed using similar techniques, the switch embodiment of Fig. 5 differs in two aspects:

(a) The microactuators preferably move the waveguide ends between discrete predefined end positions to provide as good a transfer of light across the gap as possible, and no degrees of attenuation need be involved for intermediate positions; and

(b) No bending of the waveguides is involved in these embodiments; on the contrary, the insertion loss of the switch is determined by the maintenance of the rigidity of the waveguides to ensure accurate meeting of the waveguide ends at the gaps.

However, according to a preferred variation of the embodiment of the switch of Fig. 5, if the microactuators are designed to provide controlled and graduated motion as well as discrete motion to their extremities, the switch of this embodiment can provide both a VOA function to the intensity of light passing, and a complete switching function, when the actuator(s) reach the designed switching limit of motion.

The switch geometry shown in Fig. 5 can also be used to produce 1 x n switches. A first example of this extension geometry is shown in Fig. 6A, which illustrates a 1 x 3 switch, constructed and operative according to another preferred embodiment of the present invention. In Fig. 6A, only the outlines of the waveguide are shown, the other details being similar to those of the 1 x 2 switch embodiment of Fig. 5. In Fig. 6 A, opposite the end of the input waveguide 70 are disposed the ends of three transfer waveguides, 71, 72, 73, which are moved in unison 74 by means of one or more micro-actuators, not shown in Fig. 6A, to transfer the optical signal to any one of the output waveguides, 75, 76, 77. In a similar manner, 1 x n switches can also be implemented. However, for large n, such switches may become cumbersome, and a degenerate switching array, of the type schematically illustrated in Fig. 6B below is preferably used. Reference is now made to Fig. 6B, which is a schematic drawing of a 2 x 2 optical switch, constructed and operative according to another preferred embodiment of the present invention. The switch is constructed of four switches, 80, 81, 82, 83, preferably of the 1 x 2 type shown in the embodiment of Fig. 5, connected so as to implement a 2 x 2 array. The micro-actuators of each of the four component 1 x 2 switches are operated in a synchronized manner which ensures that the correct path is generated for each switch setting selected, and that no path conflicts are generated.

Reference is now made to Fig. 7 which is a schematic diagram of a VOA, constructed and operative according to another preferred embodiment of the present invention. The VOA's described in the embodiments of Fig. 1 to Fig. 6B involve the use of a multi-mode waveguide in transferring energy from the zero order mode of the input waveguide to higher modes in the output waveguide, or for higher attenuation levels, to straightforward misalignment of the waveguides, as illustrated in Fig. 2. Since, however, this transfer is done through the free space between the two waveguide, this may result in some energy loss to the surrounding, quite apart from the problem of interference effects due to parasitic reflections from the ends of the waveguides. The embodiment of Fig. 7 attempts to overcome such losses by keeping the light propagating as a guided mode in a single continuous waveguide 84, and is hence known as a Guided Mode Mixing VOA (GMM-VOA). Such GMM- VOA's can have insertion losses of down to below ldb, without the need to provide special effects to prevent interference effects.

Referring back now to Fig. 7, the input and output fibers and waveguides, with their monitor couplers are preferably the same as those used in the embodiment of Fig. 1, and are so labeled. However, unlike the embodiment of Fig. 1, the zero order light is propagated within the continuous waveguide 84, which preferably has a number of corner bends in its length. 45° facets are located preferably on the outside walls of the bends, and are operative to reflect the light round each corner bend. A microactuator 88 is attached to one point of the continuous waveguide 84, and on actuation, is operative to distort the continuous waveguide 84, so that the propagating light is shifted from its previous propagation path, and undergoes mode mixing because of the convoluted propagation path. As a results, higher order modes of propagation are excited, which are then unable to propagate down the single-mode output fiber 22. The bending can be simple bending, torsion or any combination of stress/strain that optimizes the transfer of energy from the zero order mode to higher modes. The extent of the shifting of light into these higher order modes is dependent on the extent of the movement applied to the continuous waveguide 84, and the actuator 88 thus controls the attenuation level of the VOA. Though the embodiment of Fig. 7 has been explained in terms of geometrical effects, it is to be understood that photoelastic or other mechanical effects which interfere with the propagation of the light, may also be operative to shift energy into higher order modes, that are blocked when entering the output fiber.

Reference is now made to Fig. 8, which is a schematic diagram of a receiver variable optical attenuator (RVOA), constructed and operative according to another preferred embodiment of the present invention. The RVOA can perform the dual function of an equalizer, maintaining a desired input level to the optical receiver regardless of input signal fluctuations, and of a limiter, protecting the receiver from excessive input power which could damage the sensitive receiver.

The receiver generally comprises a sensitive optical signal detector 90, such as a high sensitivity photo-diode, capable of detecting the high frequency modulation on the optical signal, which is typically modulated at 2.5GHz, though uses at 10 GHz and 40 GHz are becoming commoner. This receiver detector outputs an electronic signal 92 of the modulation on the optical signal. Because of its high sensitivity and fast response time, such a detector is a costly and easily damaged component, and is thus preferably constrained to work under constant input power, typically 1 mW, by means of an equalizer at its input. In the RVOA embodiment of the present invention, the equalizing function is performed by means of a VOA 94, preferably of the type shown in the embodiment of Fig. 1 above. The level set by the VOA is determined by sampling the input optical signal 95 by means of a coupler 96, which preferably couples out 1% of the input signal, detecting this sampled signal by means of a photodiode 98, and using the detected signal from the photodiode 98 to adjust the level of attenuation provided by the VOA 94. Since the photodiode 98 needs only to detect the optical signal level, and not the signal modulation, it can be a comparatively inexpensive detector, unlike the sensitive optical signal detector 90, which is an expensive and easily damaged component. The RVOA embodiment shown in Fig. 8 thus provides low cost receiver protection and equalization.

Reference is now made to Fig. 9, which is an optical intensity monitor or tap 100, constructed and operative according to another preferred embodiment of the present invention. The monitor differs from prior art monitors in that it can be implemented in monolithic construction. The signal in the input fiber 102 is passed into a waveguide 103 constructed on the monitor chip. A sampling coupler 104, which preferably couples off a 1% fraction of the signal, is provided in the main waveguide 103. The main signal is passed on to the output fiber 105. The preferably 1% sampled signal is passed to a photodiode 106, from which is derived an output electronic monitor signal 108, proportional to the intensity of the input optical signal.

Reference is now made to Fig. 10, which is a digital variable optical attenuator (DVOA), constructed and operative according to another preferred embodiment of the present invention. This DVOA is an extension of the switching concept described in the PCT application published as WO/0217004 for "Mode Coupled Optomechanical Devices" to one of the inventors of the present application. A fixed waveguide 110 connects the input 112 of the DVOA with the output port 114. A series of suspended waveguide sections 116 are disposed along the fixed waveguide 110. These suspended waveguide sections 116 attenuate light propagating in the fixed waveguide by acting as attenuating mode couplers. Each such attenuating coupler extracts a part of the light by means of an evanescent field interaction when the suspended waveguide section is brought into close proximity to or contact with the fixed waveguide. Each section 116 is designed to attenuate light propagating in the waveguide by a predetermined amount, which is set by the length of each such attenuating coupler. A small sample of light at the output of the main waveguide is directed to a photodiode 117, and the electronic output signal therefrom is input to a controller 118. The controller controls the intensity along the main waveguide by driving the actuators 119 to activate different combinations of attenuating couplers according to a lookup table, which relates desired attenuation level to combinations of couplers. In the preferred embodiment of Fig. 10, a combination of different attenuating couplers is used which provides attenuation levels of from zero to 30db in steps of 0. ldb.

Reference is now made to Fig. 11, which is a DVOA, constructed and operative according to yet another preferred embodiment of the present invention. This DVOA differs from that shown in Fig. 10 in that the attenuation is generated by a predetermined number of attenuating couplers, each of the same attenuation level, but activated by means of actuators having different stiffnesses. The stiffnesses of the individual actuators are arranged in a serial progression, such that the application of a specific drive voltage activates all of the attenuating couplers up to and including that operated by the applied voltage. In the preferred embodiment shown in Fig. 11, which is for a 30db DVOA, the coupling attenuators 120 are each of value O.ldb, such that 300 are required. These 300 coupling attenuators are arranged along the fixed waveguide 122, and are connected to a common voltage source, V. The actuators 124 are preferably in the form of leaf actuators or pp actuators, as are known in the art. Each actuator has different stiffness in a progressive series, k_x to k₃₀o, where ki is the most flexible and k₃₀o is the stiffest.

Reference is now made to Fig. 12, which is a graph showing the relationship between the voltage V and the stiffness k_n overcome by that voltage. According to the preferred embodiment shown, there exists a linear correlation between V and k_n, such that a linear relation exists between the voltage V and the number of couplers that are activated, and hence between the voltage V and the attenuation value obtained. The higher the voltage V, the greater the number of coupling attenuators actuated, and the higher the attenuation. This arrangement is thus inherently digitized, each voltage being related linearly to a predefined attenuation level, as shown in Fig. 12.

The embodiment of the DVOA shown in Fig. 11 has the disadvantage of requiring a large number of actuators to provide good attenuation resolution over a large range. Reference is now made to Fig. 13, which is a DVOA, constructed and operative according to yet another preferred embodiment of the present invention, and which overcomes the problem of the large number of actuators in the embodiment of Fig. 11.

In the embodiment of Fig. 13, the coupling attenuators are divided into three groups, according to the coupling level provided. Thus, for the preferred example of a 30db attenuator, the first group 130, would preferably be of 3 attenuating couplers, each of 10 db, to provide the full 30 db. The stiffnesses ki to k₃ are serially and linearly increased proportionally to the attenuation value of each coupler. The stiffnesses ki to k₃ are serially and linearly increased proportionally to the attenuation value of each coupler. The second group 132, would preferably be of 10 attenuating couplers, each of 1 db, to provide 10 db of attenuation in increments of 1 db. The third group 134, would preferably be of 10 attenuating couplers, each of 0.1 db, to provide 1 db of attenuation in increments of 0.1 db. The stiffnesses ki to kio of these latter two groups are serially and linearly increased proportionally to the attenuation value of each coupler.

Thus, in the preferred embodiment shown, three banks of actuators are required: Core actuators, to which are applied voltages VI, regular actuators, to which are applied voltages V2, and fine actuators, to which are applied voltages V3. For the 30db attenuator described above, 23 actuators are thus required, ten for each full decade of attenuation. This significant reduction in the number of actuators is achieved by the simple expediency of increasing the number of applied voltages to one per decade of attenuation, which means 3 voltages for the preferred example shown.

The actuators can be preferably operated either in a look-up table mode, or in a feedback mode using a power sampling monitor in the output of the waveguide. In either embodiment, as an example, if 23.5db attenuation is required, VI is increased to 20db, V2 is then increased until 3 attenuating couplers are activated, and then V3 is turned up until 5 actuators are active. In the feedback mode, if for example, because of a weakening of the stiffness of one of the actuators, V3 were operative to activate 6 actuators in the last stage, the value of attenuation obtained would be 23.6 db instead of 23.5 db, and the feedback loop would command the control system to lower the voltage slightly. The feedback mode of operation is thus able to compensate for changes in the stiffnesses of the MEMS actuators over time, or for changes in the attenuation values of any of the coupling attenuators with time, by applying a higher or lower drive voltage as required.

For the Lookup mode of operation, for the 23.5db attenuation example, the controller would simply provide the command Go (V1,V2,V3) = (2, 3, 5), where the function (2,3,5) represents the number of actuators to be operated in each bank.

The preferred embodiment of Fig. 13 has a number of advantages over that of Fig. 11 and that of other prior art DVOA's:

(a) Although there are 23 actuators there are only 3 voltages to control.

(b) The DVOA may be implemented with a look-up table or under closed loop control.

(c) The use of a straight waveguide results in low polarization dependent losses, and low absolute losses.

(d) Simple, small footprint (10-50 micron) digital actuators can be used.

A problem which arises in many integrated optical systems, subsystems or components is the need to align their on-chip inputs and outputs to the input and output optical fibers connecting the system to the outside world. In the prior art, this is generally done by laborious procedures involving micro-manipulation and manual clamping techniques. Such methods are slow, unable to compensate for changes in dimensional accuracy of the parts with time, and their accuracy is often subject to operator skills.

The problem can be defined more quantitatively by noting that the diameter tolerance of a typical fiber is typically of the order of ±lμm, and the tolerance for the eccentricity is in the range of ±0.5μm. In addition, the accuracy of any locating feature on the substrate itself, such as a V-groove, has a tolerance of typically ±0.5 μm. Therefore when placing a fiber in a well-defined locating feature, such as V groove, an accumulative misalignment of the order of 2-3 μm can be expected. Such a level of misalignment results in excessive insertion loss in the system, as well as spurious signals due to reflection at the inputs and outputs. This problem increase when dealing with fiber ribbon inputs and outputs, which, in addition to the inherent losses due to this misalignment, also generate undesirable non-uniformity in the insertion losses between adjacent channels. This explains the need for careful and accurate alignment of the input and output fibers with the integrated optics substrate.

Reference is now made to Fig. 14, which is a schematic drawing of an optical mother board system (OMB), constructed and operative according to a preferred embodiment of the present invention, which largely overcomes many of the above-mentioned problems in the alignment of the inputs and outputs of integrated optics systems. The OMB system is so-called because of its resemblance to a computer mother board, where all of the essential components are mounted on the mother board, and all connections thereto are performed simply and repeatably. The OMB is a micro-aligning system, which aligns a passively placed fiber with the waveguide of the optical chip or substrate to which it is intended to interface, with two or even three degrees of freedom. The system is able to find the position of optimum alignment between the fiber and the waveguide, by providing the ability to adjust these mutual positions while testing the transfer of optical power between them, or by physically viewing their mutual position by means of a camera. Once this alignment has been achieved, the system, according to further preferred embodiments of the present invention, is able to latch and lock the selected alignment such that it is maintained for as long as is desired. Furthermore, the locking system is fail-safe in that a fall in power leaves the locking intact.

In Fig. 14, there is shown schematically a preferred embodiment of the OMB 140 with a single input fiber 142 and output fiber 144 attached thereto. The optical system chip or substrate 146 is fixed on the OMB, and in the preferred example shown in Fig. 14, has one input waveguide 148 which is to be fed by the input fiber 142, and one output waveguide 150 which is to feed the output fiber 144. According to further preferred embodiments, the OMB can be applied to devices with only one output, such as laser sources. On the OMB, and disposed in a position that approximately connects between the locations of the input fiber end and the input waveguide end, there is situated a short OMB input waveguide section 152. This OMB input waveguide section is equipped at both of its ends with three dimensional micro-actuators 154, 155. Similar three dimensional actuators are also preferably disposed at the ends of a similar OMB output waveguide section 156. Each of the micro-actuators is preferably provided with a latching and locking mechanism, which is to be further explained in the embodiments of Figs. 15 to 17 hereinbelow. In use, the microactuators are sequentially adjusted while the optical system is operating, the adjustment being performed to provide optimum power transfer, or minimum insertion loss, or minimum power reflection, or a combination thereof, as appropriate for the application in hand. Once the optimum position has been achieved, this position is latched and locked, as described in the preferred embodiments hereinbelow.

Reference is now made to Fig. 15, which is a schematic drawing of an actuator latching and locking system, according to another preferred embodiment of the present invention. The preferred embodiment shown in Fig. 15 is suitable for a one-dimensional locking system, such as would preferably be used in the VOA embodiments of Fig. 1 of the present invention, with two waveguides 16, 20. If it were to be applied for one direction of the two dimensional micro-actuators of the OMB embodiment of the present invention, the waveguide 16 would be adjusted opposite a fiber end, rather than opposite another waveguide 20 as shown in Fig. 15. For the example, of the VOA of Fig. 1, the position of the input waveguide 16 is adjusted opposite the output waveguide 20 to provide the required attenuation level, by means of the actuator Al, designated 24 in Fig. 1. The actuating arm 160 of the micro-actuator Al passes between a pair of clamp plates 162, 164, connected respectively to the actuating arms of two further micro-actuators A3, A4, which are of the normally closed type. The actuator A4 is optional, and its function can be performed alternatively by the spring action of the latching lips or shutter, 168. The direction of operation of each actuator when activated is indicated in Fig. 15 by arrows.

The operation of this actuator latching and locking system preferably comprises two separate stages. In the first stage, the latching mechanism is armed, this being a one time operation performed only the first time the mechanism is used, and in the second stage, it is adjusted and locked, which operation can be performed thereafter as often as is necessary. In the first stage, actuator A3 is actuated to allow space for the movement of actuator A2. A4 is then actuated to pull the latching lips 168, also known as the shutter, away from the path of motion of the stop plate 166. Then A2 is operated to pull the stop plate 166 to the left of the shutter 168. Once the mechanism is armed by having the stop plate to the left of the shutter, A4 followed by A2 can be released, and the stop plate 166 returns to the right to rest against the shutter 168. At this stage when A3 is released 164 and 162 clamp 160.

In the alternative and preferred embodiment mechanism without the actuator A4, the latching lips 168 are either spring loaded or fixed, and the stop plate 166 is pulled past their chamfered edges by the action of A2, either pushing the spring loaded latching lips slightly apart in the process, or slightly bending the edges of the lips 168 in the process, and then returns to rest on the latching lips when A2 is released. Once the arming procedure has been completed, clamping plate 162 is very close to arm 160, while clamping plate 164 is held away from the arm 160 by actuator A3.

In the adjustment stage, the correct position of the waveguide is established by means of actuator Al, and then the drive voltage is released from A3, such that clamping plate 164 clamps the operating arm 160 against plate 162, and the position of waveguide 16 is thus fixed. The voltage can then be released from Al. One advantage of this clamping mechanism is that in the clamped position, none of the actuators require powering, as they are all normally closed. This situation is an important improvement over prior art clamping mechanisms, in that in the event of a power failure, the adjustment is maintained in its intended position, since the clamp actuators A3, Al, need powering only when it is necessary to adjust the waveguide position. The arming stage is necessary since it is to manufacture a mechanism that is normally closed, since there is no way of fabricating 164 and 162 such that they touch and lock 160 when in an unpowered position. The shutter mechanism, with or without actuator A4, is to move plate 162 physically into a position close to 160 during the arming process.

Reference is now made to Fig. 16, which is a representation of a monolithic pattern of the schematic latching mechanism shown in Fig. 15, to illustrate one preferred example of a real-life implementation of the embodiment shown in Fig. 15. The four micro-actuators Al to A4 are shown, as are the other operative parts of Fig. 15, with their same nomenclature. In the embodiment of Fig. 16 the suspended waveguide 16 is shown being aligned against a fiber 17, as would be used in the OMB application of Fig. 14. In Fig. 16, electrostatic actuators are used, but it is to be understood that other type of actuators such as thermal actuators can generally also be used to perform the same functions as the electrostatic actuators.

In the OMB embodiment described in Fig. 14, there is a need for two dimensional micro-actuators, shown schematically as 154, 155 in Fig. 14, such that the component waveguides can be moved and locked in the plane perpendicular to the direction of the waveguide itself. In general, there is no need to provide adjustment in the direction of the waveguide, since the accuracy of placement in this direction, typically better than 15 um, is sufficiently good to provide efficient transfer of the light between the waveguide and the fiber. In cases where a latching and locking mechanism is provided, as shown in the embodiments of Figs. 15 and 16, a micro-actuator operating in the third direction is also required to move the waveguide lengthwise to its locking position.

Reference is now made to Fig. 17, which is a schematic view of a detailed pattern in a monolithic substrate, according to a further preferred embodiment of the present invention, for executing such motion in two dimensions. The preferred application illustrated in Fig. 17 is that of aligning a waveguide 170 to an external fiber 172 through the silicon package wall 174 of the chip. The fiber is located in a V/U groove, and thus has a fixed position. The waveguide is suspended such that it has movement with three degrees of freedom (DOF), two for adjusting the position of the waveguide in the cross section of the fiber, and one to allow latching of the waveguide-fiber interface. A latching plate 175 is attached near the end of the suspended waveguide 170, and when shutter actuators 176 are activated, moving in the ±Y direction, the shutter 178 opens allowing the latching plate to move in the X-direction towards the fiber interface.

Motion of the waveguide itself is generated by the 3 -DOF actuator 180. Three biasing voltages with reference to the ground, control this actuator. VI creates an electrostatic field between the vertically drawn walls of the outer frame 182 of the actuator and the inner frame 184. This force is responsible for moving the waveguide in the +X direction, and is actuated only after the shutter 178 is opened. V2 creates an electrostatic field between the horizontally drawn walls of the outer frame 186, and upper and lower fixed electrodes 188. This force is responsible for moving the waveguide in the ±Y direction. The vertically-drawn walls 182 and the horizontally-drawn walls 186 of the outer frame are insulated from each other by a region of silicon with low doping 185. Finally, V3 is applied between the substrate 189 and the inner frame 184, as shown in the insert drawing, which is a view in the plane of the substrate, and is responsible for moving the waveguide in the ±Z direction, i.e. into and out of the plane of the drawing. The waveguide is preferably made of low doped or intrinsic silicon, while the actuators have a high doping level to provide them with good conductivity. The different doping is provided by using SOI wafers with intrinsic silicon for the device layer, that is selectively subjected to SOD or ion implantation to create conductive areas for the actuators.

In the embodiment described in Fig. 17, the micro-actuators for motion in all three directions are conventional electrostatic motors. According to another preferred embodiment of the present invention, thermal actuators or motors can be used for the adjustment directions out of the plane of the substrate or chip surface. Such actuators, being generally simpler and smaller than electrostatic actuators, can also preferably be used for the in-plane movements.

Reference is now made to Fig. 18 A, which is a schematic drawing illustrating a method of constructing a thermal actuator for out-of-plane motion, according to another preferred embodiment of the present invention. Fig. 18A shows a side view of a suspended beam section 190, having a doping profile 191 in the direction into the depth of the substrate 192, such as can be readily formed by diffusion into the substrate either by SOD or ion implantation. A variable resistance profile is thus generated through the depth of the substrate, depending on the doping profile generated. The beam is etched into the selectively doped region, and a current running through such a beam element thus also acquires a depth profile, as shown in the plot to the left of the drawing of the beam, showing the current I in any layer of the beam, as a function of the depth d of that layer. As a result of this current profile, the element heats up selectively in such a way that it bends into or out of the plane of the substrate, like a bi-metal strip, depending on the profile generated. In general, such a doped profile beam will tend to bend upwards under the effect of current flow, since the resistance will be higher at its lower edge where the doping is lower, and it will thus get hotter there, and bend upwards. By means of mechanical structures, this upwards motion of the beam can be linked to a downwards motion of another beam. Reference is now made to Fig. 18B, which is a schematic drawing illustrating a method of constructing a thermal actuator for out-of-plane motion, according to another preferred embodiment of the present invention. Fig. 18B shows a side view of a suspended beam section similar to that of Fig. 18 A, but in this embodiment, a preferably castellated profile has been etched into the top surface of the beam. As a result, when the beam is heated by passage of a current, the expansion at the top edge is relieved by the additional incursions of the castellated edge, these being missing from the bottom edge, such that the bottom edge expands more, and bends the beam upwards. This embodiment is simpler to construct than that of Fig. 18A, since no doping profile needs to be generated, just the etched form. Though a castellated profile is shown in Fig. 18B, it is to be understood that any suitable similar profile is also suitable for this embodiment, such as a sawtooth profile, or a sinusoidal profile.

Reference is now made to Fig. 19A, which is a schematic drawing of a dynamic fiber ribbon connector, constructed and operative according to another preferred embodiment of the present invention, as an example of an application of the three-dimensional actuator embodiment of Fig. 17. As has been mentioned above, the accuracy of alignment of a fiber ribbon with respect to the substrate to which it is attached can be 2 to 3 microns. High accuracy passive fiber ribbons are available with accuracy of 0.5um but these fibers are very costly. Though a 2 to 3 micron misalignment may be marginally acceptable when feeding a standard 9 x 9 micron waveguide, if the ribbon fiber is to be fed by or to feed a small cross section waveguide, with cross section smaller than 5um x 5um, such as a laser output, or semiconductor optical amplifier (SOA), then a 2 to 3 micron misalignment is totally unworkable. Furthermore, in many cases, a spot size converter (SSC) for mode matching is required at the end of the waveguide, making the end of the waveguide even smaller. Under such conditions, the use of the OMB embodiment, as shown in Fig. 19 A, to dynamically match the waveguides to the ribbon fiber ends, greatly increases the efficiency of the connector. In the embodiment of Fig. 19A, the end of each of the suspended waveguides 195 is fitted with a multi-dimensional actuator 196, such that dynamic matching to the ribbon fiber inputs 197 is optimized. The other ends of the waveguides are connected to high accuracy pitch fixed waveguides 198.

Reference is now made to Fig. 19B, which illustrates schematically a complete OMB with integrated optics assembly, 200, comprising input and output ribbon fibers 202, and monolithic 203 and hybrid 204 function chips mounted on the substrate. All of the connections are made using suspended waveguide 205 with SSC's 206 where necessary, and dynamic manipulators 208 on the end of each one to provide optimum loss free inter-component connection.

Th efficiency of operation of the above-described embodiments of the present invention, such as the VOA's and the switches, is dependent to a large extent on the parallelism between the facing walls of the waveguide gaps therein, or of the parallelism of other planes between which light is transferred. Furthermore, if the input and output facets of the substrate are not vertical, the connecting fibers need to be tilted, so that only the basic mode is input or output. It would therefore be convenient if these facets too, were perpendicular to the substrate plane. The Reactive Ion Etching (RLE) process, by which such integrated optical circuits are preferably produced, results in fairly parallel walls, with angles of 89° to 89.5° to the substrate plane. However, for high quality optical waveguide applications, the parallelism of the walls should be less than 0.2°, and the perpendicularity thus better than 89.8°.

According to yet further preferred embodiments of the present invention, there is also provided a method of producing facets and walls with high parallelism, of the order stated above or better. The process involves exposing specific crystallographic planes in the silicon substrate, such as the (100) plane or the (111) plane. Walls using these planes, once exposed, results in very vertical and parallel facets that are also of extremely high quality, with surface roughness of the order of a few nanometers. Though the process is described hereinbelow for a silicon substrate, it is to be understood that the invention is equally applicable to single crystal substrates made of any other suitable integrated optics substrate, as is known in the art.

Reference is now made to Fig. 20A, which illustrates schematically a (110) plane silicon wafer after a waveguide feature has been etched therein by RLE, showing a small tilt of the sidewalls. In such (110) wafers, the (111) plane is vertical to the substrate. A Si0₂ or Si₃N₄ mask is applied to the substrate, as is known in the art, in order to perform a wet etch of the substrate, preferably using KOH solution. The mask protects those areas which are not to be etched. The KOH etches through the (110) plane, attacking the material in all directions, as shown in Fig. 20B. However, since the rate of etching of the (111) plane is much slower than that of the other planes in the silicon, the (111) plane at the edge of the mask is gradually exposed, as shown in Fig. 20C, revealing a smooth and accurately vertical wall feature.

Since (110) plane silicon wafers are not the standard wafers used in the industry, and are not always easy to obtain, alternative and preferred methods are now described, using the industry standard (100) wafers. Two different methods of so doing are now described. According to a first method, the standard (100) wafers are used with the wafer locating flat in the {110} direction. In this case, the design is rotated 45°, such that the facets, once etched, are in the (100) planes. Alternatively and preferably, these standard wafers are ordered with the wafer locating flat in the {100} direction, such that the design does not need to be rotated by 45°, which simplifies the fabrication process. However, although such wafers are of the industry standard (100) plane, the direction of the locating flat is not standard, and is subject to special order. Consequently, use of the standard (100) wafers but with the wafer locating flat rotated 45° to the {110} direction is clearly preferable for convenience of substrate supply, but is more complicated from a fabrication point of view.

Reference is now made to Fig. 21, which illustrates schematically an industry standard (100) wafer 215, and with the desired waveguide features having a 45° rotation to the {110} wafer locating edge. In the drawing are shown an input facet 210, a gap between waveguides 212, and an output facet 214, all aligned in the {100} directions.

Reference is now made to Fig. 22, which illustrates schematically one of the features of Fig. 21 in magnification, such as gap 212, showing simulation results of the exposed planes in the feature, selectively etched with a KOH wet etch, and using a rectangular 45°-tilted mask. The (100) planes that are exposed as sidewalls or facets, retain their vertical orientation by the effect of the preferential etching of the (100) planes, even though the cavity that is being etched expands as the etching proceeds. The (110) planes are not visible in the plan view of Fig. 22, as they stand out of the plane of the drawing, and parallel to the line so marked.

However, as the etching process is continued, the (111) planes begin to appear at the corners of the etched cavity, since they are etched much more slowly than the (110) and the (100) planes, and only appear after the (100) and the (110) planes are exposed. The etch rates for the three different planes (110), (100) and (111) are approximately in the ratio 300:100:1, depending on the specific etchant concentration and temperature. The presence of these (111) planes results in slight corner damage using this procedure.

Reference is now made to Fig. 23, which illustrates the rate of progress of the various etching steps according to the procedure shown in Figs. 21 and 22. The shape of the sidewall 211 of the feature is non-vertical, and, in the example shown, an undercut as a result of the specific RIE process used. Etching with a KOH solution results in the vertical (100) planes from being exposed. Since the top corner 216 of the feature is more exposed to the etchant than the bottom corner 218, the top of the feature is preferentially etched more quickly than the bottom of the feature, and (100) planes are successively more and more exposed as the etching proceeds until the sidewall eventually attains a completely vertical (100) plane surface.

Reference is now made to Figs. 24A to 24E, which illustrate the progressive stages, using the process described in Figs. 21 to 23, in etching a complete feature into a silicon on insulator (SOI) substrate. The substrate is made of silicon 220, on which are layers of silicon oxide 222, silicon 224 which will eventually become the desired waveguide, and silicon oxide 226. The right hand side of each drawing shows a top view of the feature itself, which is a straight section of waveguide 224, which will eventually become a suspended section, and with an actuator 230 attached to the center of the waveguide. In order to prevent the exposure of the (111) planes from reaching the waveguide itself, and causing damage to the shape of the waveguide, side arms 228 are designed into the structure, onto the corners of which the (111) plane exposure can form harmlessly. On the left of each drawing is shown a cross section of the appearance of the etched waveguide at each step, as shown in section A-A, to illustrate the various layers applied and etched. The separate stages can be described by three main steps:

(i) Etching the entire structure shape using RIE, resulting in nearly vertical sidewalls.

(ii) Protecting the areas whose sidewalls do not need correcting.

(iii) Exposing the chip to KOH and correcting the sidewalls such that the (100) planes are exposed by preferential etching, as shown in Fig. 23.

In Fig. 24 A, the outline of the feature is etched by RIE down to the silicon substrate. In Fig 24B, a protective mask of Silicon Nitride Si₃N₄ 231 is applied to the entire feature area. In Fig. 24C, the protective mask has been lithographically defined and those areas to be removed, such as the waveguide side walls or ends 232 and any other desired parts, are etched away preferably using a phosphoric acid etch bath. In Fig. 24D, a deep potassium hydroxide (KOH) etch is performed for achieving the desired verticality and smoothness of the (100) sidewalls or ends 234, in accordance with the preferred method of this invention. It is at this stage that the slight damage is produced at the corners 236 of the feature, as mentioned above, because of the preferentially faster etching of the (110) planes at these corners, but since these corners are distant from the operative light conveying element of the waveguide, this superfluous etching is inconsequential. In Fig. 24E, the protecting Si₃N layer from the step in Fig. 24C is removed, preferably by means of another phosphoric acid etch dip, to leave the clean waveguide feature, with accurately vertical and smooth ends, ready for the next stage of the fabrication process.

Reference is now made to Fig. 25 which is a schematic drawing of a monolithic line protection chip (LPC), constructed and operative according to another preferred embodiment of the present invention. The LPC is generally part of an Optical Add and Drop Multiplexer (OADM) system, which comprises combinations of filters, switches, splitters, VOA's and detectors and is used to channel information to and from line cards that are positioned in the path of an optical ring. Prior art LPC's are generally constructed by using discrete components, such as the VOA, couplers and 1 x 2 switches, and connecting them on a board in hybrid technology. The LPC of the present invention differs from prior art LPC's in that it can be constructed monolithically on a single chip, including the VOA, the power splitters, the 1 x 2 switches, and also optionally the photodiodes and the filters, to produce a complete, multifunctional, monolithic OADM.

Referring now in detail to Fig. 25A, there are two optical information rings to provide redundancy to the information, and they are designated East and West, as is generally done in the art. Describing now the Drop function, and starting from the East ring, λj is dropped using a filter 270, and is input into the LPC 280. The filters can preferably be constructed using deformable ring resonators, such as those described in the PCT application, published as document WO 02/17004 entitled "Mode Coupled Optomechanical Device" to one of the inventors of the present invention, and hereby incorporated by reference in its entirety. In Fig. 25B, there are shown preferred embodiments of such ring filters 271, which can be integrated onto the same substrate as the rest of the OADM. The MEMS actuators 273 change the ring diameter, and thus change the wavelength to be dropped or added, as described fully in the above-referenced WO 02/17004. Alternatively and preferably, monolithically fabricated Mach-Zehnder interferometric filters can be used for filtering wavelengths of light from the input signal, the wavelengths being selected by means of micro-actuators attached to one or more arms of the Mach-Zehnder filter.

A part of the signal is channeled to a monitoring photodiode PD1 using an integrated coupler 272. This photodiode records the intensity of the incoming signal and is used to set the state of a VOA 274, to ensure that the signal does not saturate the receiver section 276 of the transceiver. The signal is routed through a 1 x 2 switch 278 which is operative to select between signals dropped from the East or West rings. Because of the redundancy of the system, the information coming from East or West is the same, and the switch is set to choose that with the higher quality. The λj signal from the West ring with the same information would be dropped by means of filter 282, and its intensity monitored on photodiode PD-2. This signal would reach the same 1 x 2 switch 278, and the higher quality of the two would be channeled to the transceiver.

The Add function operates in a very similar manner to that of the drop function, except that the Add signal is generated at the transmitter side 284 of the transceiver and traverses a VOA 286 to regulate its intensity, as determined by the monitor photodiode PD-3. It is next channeled to a splitter 288 which divides its intensity equally for adding to the West and East rings by means of filters 290, 292. In some applications, the splitter 288 is replaced with a 1 x 2 switch 294, as shown in the insert drawing, which is operative to direct the whole signal either to East or to West. In that case only the East or the West ring operates, while the other remains as a back-up ring.

The LPC of the present invention is constructed on a single motherboard 280 such as a silicon substrate where each of the components is fabricated monolithically. Alternatively and preferably, all or some of the photodiodes PD-1 to PD-3 may be placed on the motherboard in hybrid technology, in cases where they cannot be grown monolithically. The basic components of the LPC 180 chip shown in Fig. 25A can be manufacture monolithically on silicon substrate. According to one preferred embodiment, the 1 x 2 switches and the VOA's are constructed by the methods of the present invention. Alternatively and preferably, the 1 x 2 switch and the VOA are based on moving waveguides techniques, either using mode matching and mismatching coupling, or using evanescent or mode coupling as described in the above referenced co-pending patent application, published as WO 02/17004.

Reference is now made to Fig. 26, which illustrates a further use of the waveguide technology, according to further preferred embodiments of the present invention, for use in applications for the coupling of light between main waveguide and a sampling port. Such a component is required for the VOA embodiments of the present invention, and in numerous other integrated optics applications. In Fig. 26, there is schematically shown a 1% output coupler using a multimode interferometer (MMI) of the type that is known in the art, constructed and operative according to another preferred embodiment of the present invention. The MMI is constructed such that the interference peaks coupled out at the exit plane have predefined intensities, one preferably containing 99% of the input energy, and the other l%.The input waveguide 300 leads the optical signal into the MMI 302, and the two outputs are arranged such that the main power is output from one waveguide 304, and a preferably 1% sample of the power is coupled out of the other waveguide 306. Such an MMI coupler in monolithic form may be used in any of the preferred embodiments of the present invention, where a monolithic coupler is preferably used.

Reference is now made to Fig. 27, which shows a further coupler, in this example, a variable coupler, constructed and operative as part of a VOA, according to another preferred embodiment of the present invention. In this embodiment, the input waveguide 310 is of suspended design, and ends at a gap 314, having properties and constructed in a manner similar to those described hereinabove. Opposite the gap 314 is disposed an output waveguide 312, into which most of the input light I_in is propagated across the gap 314, and exits the coupler as I_out. This input and output waveguide pair thus constitute a VOA similar to that shown in the embodiment of Fig. 1. As mentioned previously in connection with these VOA embodiments, a small percentage of the light leaks away from the coupling gap, and in this preferred embodiment of the present invention, is collected by the output coupling waveguide 316, as I_COupi_ed- As previously, the suspended input waveguide 310 is moveable by means of an actuator 312, and motion downwards causes the attenuation to increase. However, at the same time, this motion also causes the percentage of light coupled into the output coupling waveguide 316 to change. In general, the higher the attenuation, the higher the level of stray light in the gap, and the larger the coupled light signal. This coupled light signal can therefore be detected and measured, and used as a control signal for the actuator to set the attenuation level of the VOA, either on its own, or in conjunction with the output light level I_out. This coupled control light output differs from those of most attenuator control outputs, in that as the input waveguide moves down away from the gap, the attenuation increases, such that the output light decreases, but at the same time, the coupled control signal level increases. In most controlled VOA's, using fixed output couplers to provide a measure of the output light power, as for instance shown in the embodiment of Fig. 1, increased attenuation levels generally results in reduced control signals, and therefore in loss of control sensitivity. In the preferred VOA embodiment shown in Fig. 27, control sensitivity is increased with increasing attenuation setting.

Reference is now made to Fig. 28, which shows a further VOA embodiment, constructed and operative according to yet another preferred embodiment of the present invention. In this embodiment, the input waveguide 320 is moved by means of an actuator 322 opposite the coupling gap 324, and the attenuated light passed into the output waveguide 326. Disposed in this output waveguide there is preferably a 45° bend, and at the centerpoint of the wall of this bend 328, a sampling output waveguide is connected, and the sampled signal therein is detected, preferably by a photodiode 330. The sampling waveguide is of dimensions such that preferably 1% of the light in the main waveguide is coupled out. The output fraction, at whatever level is chosen, remains accurate regardless of the power level in the waveguide, because the location of the coupling hole near the center of the bend ensures that it samples only essentially from the center of the zero order mode. As with the previous embodiments shown, this output coupler can be used to control the attenuation level of the VOA, or alternatively and preferably, the part of this embodiment with the coupling waveguide at the center of the bend can operate as a stand alone output coupler.

It is appreciated by persons skilled in the art that the present invention is not limited by what has been particularly shown and described hereinabove. Rather the scope of the present invention includes both combinations and subcombinations of various features described hereinabove as well as variations and modifications thereto which would occur to a person of skill in the art upon reading the above description and which are not in the prior art.

Claims

CLAIMSWe claim:

1. A variable optical attenuator comprising: a suspended first waveguide section having a first end; a fixed second waveguide section having a second end disposed generally opposite said first end and in close proximity thereto; and a micro-actuator attached to said first waveguide section imparting lateral motion to said waveguide, such that change of alignment of said first end with said second end changes the attenuation of light traversing between said first and second waveguide sections.

2. A variable optical attenuator according to claim 1, and also comprising: a first coupler sampling light traversing said first waveguide; a second coupler sampling light traversing said second waveguide; and detectors detecting said sampled light.

3. A variable optical attenuator according to claim 2, and also comprising control circuitry receiving inputs from at least one of said detectors and providing control signals for said micro-actuator.

4. A variable optical attenuator according to claim 3, and wherein at least said first and second waveguides, said micro-actuator and said couplers are fabricated monolithically on a substrate.

5. A variable optical attenuator according to claim 4, and wherein at least one of said control circuitry and said detectors are also fabricated monolithically on said substrate.

6. A variable optical attenuator according to any of claims 1 to 5, wherein said first end and said second end are separated by a gap, and wherein said variable optical attenuator also comprises a second micro-actuator attached to said first waveguide section and imparting longitudinal motion thereto to change said gap, thus altering the attenuation of light traversing between said first and second waveguide sections.

7. A variable optical attenuator according to claim 6, and wherein said micro-actuator is controlled to change said gap so as to reduce interference effects of light reflected between said ends of said waveguides.

8. A variable optical attenuator according to claim 6, and wherein said micro-actuator is controlled to change said gap so as to reduce the effects of temperature on said variable optical attenuator.

9. An optical switch, comprising: an input waveguide section having a first end; an output waveguide section, having a second end; a rigid suspended transfer waveguide section having a microactuator attached thereto, imparting lateral motion to said rigid suspended transfer waveguide section such that it switches between a bridging position between said first and second ends, such that light passes between said input and output waveguides, and a non-bridging position, such that light cannot pass between said input and output waveguide sections.

10. An optical switch according to claim 9, and wherein said rigid suspended transfer waveguide section has third and fourth ends, and wherein in said bridging position, said third end is disposed opposite to and in close proximity to said first end, and said fourth end is disposed opposite to and in close proximity to said second end, so that light can pass between said input and output waveguide sections.

11. An optical switch according to claim 9, and wherein said rigid suspended transfer waveguide section has third and fourth ends, and wherein in said non-bridging position, said third end is disposed laterally distant from said first end, and said fourth end is disposed laterally distant from said second end, so that light cannot pass between said input and output waveguide sections.

12. An optical switch according to any of claims 9 to 11, and' wherein said input waveguide section, said output waveguide section, said rigid suspended transfer waveguide section and said micro-actuator are fabricated monolithically on a substrate.

13. A multi- way optical switch, comprising: an input waveguide section having a first end; a plurality of output waveguide sections, each having a second end; and a plurality of rigid suspended transfer waveguide sections, one for each of said plurality of output waveguide sections, said plurality of transfer waveguide sections having at least one micro-actuator attached, imparting lateral motion between predetermined positions to at least one of said transfer waveguide sections, such that at each of said predetermined positions, one of said transfer waveguide sections bridges between said first end of said input waveguide section, and said second end of one of said plurality of output waveguide sections, such that light passes between said input waveguide section and one of said plurality of output waveguide sections.

14. A multi-way optical switch according to claim 13, and wherein said input waveguide section, said output waveguide sections, said rigid suspended transfer waveguide sections and said at least one micro-actuator are fabricated monolithically on a substrate.

15. A variable optical attenuator comprising: an input waveguide section; an output waveguide section; a suspended waveguide section connecting said input and output waveguide sections; a micro-actuator attached to said suspended waveguide section imparting distortion to said suspended waveguide section from its initial shape, such that high order modes are excited in light traversing through said suspended waveguide section, and a serial component which attenuates the propagation of higher order modes, such that said high order modes excited in light traversing through said suspended waveguide section are attenuated.

16. A variable optical attenuator according to claim 15 and wherein said distortion is at least one of lateral bending distortion and torsional distortion.

17. A variable optical attenuator according to claim 15 and wherein said serial component is a single mode output fiber.

18. A variable optical attenuator according to claim 15 and wherein said suspended waveguide section has at least one bend in its path, said bend having a mirror disposed such that light traversing through said suspended waveguide section is reflected therefrom at an angle of reflection, and distortion to said suspended waveguide section changes said angle of reflection, such that said light is deviated from its path in said undistorted waveguide.

19. A variable optical attenuator according to claim 15 and wherein said high order modes are excited by geometrical distortion in the propagation path of light passing therethrough.

20. A variable optical attenuator according to claim 15 and wherein said high order modes are excited by photoelastic effects on the material of said suspended waveguide section.

21. An optical receiver protection device, comprising: a receiver detector receiving an optical signal; an input waveguide supplying said optical signal; a coupler for extracting a sample of said optical signal; a monitor detector receiving said sample of said optical signal, and providing an electronic signal corresponding thereto; a variable optical attenuator controlled by a micro-actuator disposed between said input waveguide and said receiver detector, said micro-actuator being controlled by said electronic signal, such that when said electronic signal is indicative of an excessive optical signal in said input waveguide, said micro-actuator is operative to attenuate said input optical signal.

22. An optical receiver protection device according to claim 21, and wherein at least said input waveguide, said coupler, and said variable optical attenuator are fabricated monolithically on a substrate.

23. An optical receiver protection device according to either of claims 21 and 22, and wherein said variable optical attenuator comprises a suspended first waveguide section and a fixed second waveguide section, the ends of said sections being mutually laterally positioned by means of said micro-actuator, said micro-actuator being attached to said first waveguide section, such that the attenuation of said attenuator is varied.

24. An optical receiver protection device according to either of claims 21 and 22, and wherein said variable optical attenuator comprises a suspended waveguide section connected between said input waveguide and said receiver detector, said micro-actuator imparting distortion to said suspended waveguide section, such that high order modes are excited in said optical signal traversing through said suspended waveguide section, such that the attenuation of said attenuator is varied.

25. An optical receiver protection device according to any of claims 21 to 24, and wherein said receiver detector is significantly more costly than other components of said receiver protection device.

26. A monolithic in-line optical power monitor, comprising: an input fiber conveying an optical signal to be monitored; an output fiber outputting said optical signal after monitoring; a waveguide section connecting said input fiber and said output fiber; a coupler disposed within the path of said waveguide, and sampling a part of said optical signal; and a detector measuring said sampled part of said optical signal, and providing an electrical signal in accordance with the intensity of said optical signal; wherein at least said waveguide section and said coupler are fabricated monolithically on a substrate.

27. A monolithic in-line optical power monitor according to claim 26, and wherein said detector is also fabricated monolithically on said substrate.

28. A digitally controlled variable optical attenuator, comprising: a fixed waveguide receiving an input optical signal; a plurality of suspended waveguide sections disposed serially along said fixed waveguide, and spaced therefrom; micro-actuators operative to move said suspended waveguide sections towards said fixed waveguide, such that said suspended waveguide sections attenuate said optical signal propagating in said fixed waveguide by evanescent field interaction; a coupler for extracting a sample of said optical signal after passing said plurality of suspended waveguide sections; and a monitor detector receiving said sample of said optical signal, and providing an electronic signal for controlling said micro-actuators, such that attenuation of said input optical signal is digitally determined.

29. A digitally controlled variable optical attenuator according to claim 28, and wherein said micro-actuators are such as to move said suspended waveguide sections into close proximity with said fixed waveguide.

30. A digitally controlled variable optical attenuator according to claim 28, and wherein said micro-actuators are such as to move said suspended waveguide sections into contact with said fixed waveguide.

31. A digitally controlled variable optical attenuator according to claim 28, and wherein said electronic signal for controlling said micro-actuators is processed to operate a combination of micro-actuators which generate a desired level of attenuation.

32. A digitally controlled variable optical attenuator according to any of claims 28 to 31, and wherein individual ones of said plurality of suspended waveguide sections generally have different attenuating effects.

33. A digitally controlled variable optical attenuator, comprising: a fixed waveguide receiving an input optical signal; a plurality of suspended waveguide sections disposed serially along said fixed waveguide, and spaced therefrom; micro-actuators operative to move said suspended waveguide sections towards said fixed waveguide, such that said suspended waveguide sections attenuate said optical signal propagating in said fixed waveguide by evanescent field interaction, and wherein said micro-actuators generally have differing stiffnesses; a coupler for extracting a sample of said optical signal after passing said plurality of suspended waveguide sections; and a monitor detector receiving said sample of said optical signal, and providing an electronic signal corresponding to said optical signal for applying to said micro-actuators, such that attenuation of said input optical signal is digitally determined.

34. A digitally controlled variable optical attenuator according to claim 33, and wherein individual ones of said suspended waveguide sections generally have the same attenuating effect, and wherein said electronic signal is applied to all of said micro-actuators, and is operative to actuate those micro-actuators having a stiffness up to that overcome by said electrical signal, such that a desired level of attenuation is generated.

35. A digitally controlled variable optical attenuator according to claim 33, and wherein said suspended waveguide sections are divided into groups, each group covering a different range of attenuation, and each group having a common applied voltage, each of the suspended waveguide sections in each group generally having a range of attenuating effects, and said suspended waveguide sections having stiffnesses corresponding to their attenuating effect; and wherein said electronic signal is processed into different signals for application separately to all of said micro-actuators in each of said groups, and is operative to actuate those micro-actuators in each group having a stiffness up to that overcome by said different signals, such that a desired level of attenuation is generated.

36. A digitally controlled variable optical attenuator according to claim 35, and wherein each of said groups covers a decade of attenuation.

37. Apparatus for the transfer of optical signals between different terminal ports of an optical system, comprising: a connecting waveguide having ends for connecting between two of said different terminal ports; and at least one micro-actuator mechanism attached to at least one end of said connecting waveguide, said micro-actuator mechanism being adjustable in at least one dimension to manipulate the position of said end of said connecting waveguide relative to at least one of said terminal ports; wherein at least said connecting waveguide and said micro-actuator mechanism are constructed monolithically on a substrate.

38. Apparatus according to claim 37, and wherein said at least one dimension is two dimensions.

39. Apparatus according to claim 37, and wherein said at least one dimension is three dimensions.

40. Apparatus according to claim 37, and wherein said at least one micro-actuator mechanism is lockable, such that said position of said end of said connecting waveguide can be fixed in its optimum position.

41. Apparatus according to claim 37, and wherein said at least one micro-actuator mechanism is adjusted while observing the transfer of optical signals between different terminal ports of said optical system to optimize said transfer.

42. Apparatus according to claim 41, and wherein said transfer is optimized by providing at least one of maximum optical power transfer, minimum insertion loss, and minimum power reflection.

43. Apparatus according to claim 37, and wherein said transfer is optimized by viewing the physical location of said end of said connecting waveguide relative to said terminal port of said optical system.

44. Apparatus according to claim 37, and wherein said terminal ports are any one of a fiber V-groove interface, the waveguide of an optical chip, an integrated optics connector port, a laser source output, and a hybrid component connector port.

45. Apparatus for locking the position of the operating arm of a microactuator, comprising: a first clamping plate on one side of said operating arm, said first plate being movable by means of a first micro-actuator, operative when powered to pull said plate away from said arm; and a second clamping plate on an opposite side of said operating arm, said second plate being movable by means of a second micro-actuator, and which is initially held away from said arm by means of a latching plate held behind a shutter; and wherein said second micro-actuator is activated to pull said latching plate irreversibly through said shutter, such that said second clamping plate is latched close to said operating arm, such that when said first micro-actuator is released, said first clamping plate locks said operating arm against said second clamping plate.

46. Apparatus for locking the position of the operating arm of a microactuator according to claim 45, and wherein said shutter is opened by the operation of a third micro-actuator in order to allow said second micro-actuator to pull said latching plate through said shutter.

47. Apparatus for locking the position of the operating arm of a microactuator according to claim 45, and wherein said locking is maintained without the application of voltages to any of said micro-actuators.

48. Apparatus for locking the position of the operating arm of a microactuator according to any of claims 45 to 47, and wherein at least one of said micro-actuators is electrostatically operated.

49. Apparatus for locking the position of the operating arm of a microactuator according to any of claims 45 to 47, and wherein at least one of said micro-actuators is thermally operated.

50. A thermal actuator for providing motion out of the plane in which said actuator is disposed, comprising a beam conveying an operating current, said beam having a doping profile, such that its electrical conductivity varies with depth, such that said operating current causes said beam to bend relative to said plane.

51. A thermal actuator for providing motion out of the plane in which said actuator is disposed, comprising a beam conveying an operating current, said beam being clamped at an end, and having a relieved profile on one edge, such that as it heats up under the influence of said operating current, said beam bends relative to said plane in the direction of said relieved profile.

52. A method of improving the verticality of faces in a monolithic structure fabricated in a substrate, comprising the steps of: generating an approximately vertical face in said structure by means of reactive ion etching; wet etching said vertical face in order to expose preferentially vertical crystallographic planes in said substrate.

53. The method of claim 52, and wherein said improvement of verticality of faces is operative to improve the parallelism of faces of optical planes in monolithic waveguide structures, such that the efficiency of transfer of light across said faces is improved.

54. The method of claim 52, and wherein said substrate is a silicon substrate having a 110 plane of orientation, and wherein said wet etching step is operative to expose the 111 planes of said substrate, said 111 planes being accurately perpendicular to said 110 plane of orientation of said substrate.

55. The method of claim 52, and wherein said substrate is a silicon substrate having a 001 plane of orientation, and wherein said wet etching step is operative to expose the 100 planes of said substrate, said 100 planes being accurately perpendicular to said 001 plane of orientation of said substrate.

56. The method of claim 55, and wherein said etching takes place at 45 degrees relative to the 110 direction in said substrate.

57. The method of claim 52, and wherein said substrate is a silicon substrate, and said wet etching is performed by a potassium hydroxide solution, after protection of areas not to be wet etched by means of a protective mask.

58. A multifunctional line protection chip, comprising a substrate comprising: a coupler for sampling an input signal; an input monitoring photodiode detecting said sampled input signal; a two way, one pole optical switch; a variable optical attenuator; and a transceiver for detecting said input signal; wherein said variable optical attenuator is controlled by the output of said photodiode, such that a signal which would saturate said transceiver is attenuated., and wherein at least said coupler, said two way, one pole optical switch and said variable optical attenuator are fabricated monolithically.

59. A multifunctional line protection chip, according to claim 58, and wherein said input monitoring photodiode is also fabricated in said substrate.

60. A multifunctional line protection chip, according to claim 58, and also comprising monolithically fabricated ring resonators, operative for filtering wavelengths of light from said input signal, said wavelengths being selected by means of micro-actuators attached to said rings.

61. A multifunctional line protection chip, according to claim 58, and also comprising monolithically fabricated Mach-Zehnder interferometric filters for filtering wavelengths of light from said input signal, said wavelengths being selected by means of micro-actuators attached to an arm of said filters.

62. A variable optical coupler, comprising: a suspended first waveguide section having a first end; a fixed second waveguide section having a second end disposed generally opposite said first end and in close proximity thereto; a micro-actuator attached to said first waveguide section imparting lateral motion to said waveguide, such that change of alignment of said first end with said second end changes the attenuation of light traversing between said first and second waveguide sections, and a fixed third waveguide section disposed with its end close to the gap between said first end and said second end, such that said third waveguide section collects coupled light from said first waveguide section not transferred to said second waveguide section.

63. A variable optical coupler according to claim 62, and wherein the level of said coupled light increases with increased attenuation of light traversing between said first and second waveguide sections.

64. A variable optical coupler according to either of claims 62 and 63 and wherein said light coupler is fabricated monolithically on a substrate.

65. A variable optical coupler, comprising: a suspended first waveguide section having a first end; a fixed second waveguide section having a second end disposed generally opposite said first end and in close proximity thereto; and a micro-actuator attached to said first waveguide section imparting lateral motion to said waveguide, such that change of alignment of said first end with said second end changes the attenuation of light traversing between said first and second waveguide sections; wherein said fixed second waveguide section has a discontinuity in its path, at which is connected an output waveguide operative to sample the light traversing said discontinuity.

66. A variable optical coupler according to claim 65, and wherein said discontinuity is disposed such that it samples light from the center of a zero order mode propagating in said second waveguide section.

67. A variable optical coupler according to claim 66, and wherein said discontinuity is a right angle bend, and said output waveguide is connected at the center of the apex of said right angle bend.