CROSS REFERENCE TO RELATED APPLICATIONS This application claims priority under 35 USC 119 (e) (1) of United States Provisional Patent Application Serial No. 60/302, 582 and U. S. Provisional Patent Application 60/302, 246 both filed June 29,2001.

BACKGROUND As set forth herein, items in the figures are referred to in the form #1-#2 where #1 denotes the figure and #2 denotes the item in that figure. For example, 3-107 refers to item 107 of FIG. 3.

Typically, connectors used in fiber optics are keyed and formed to snap together in a very precise way to ensure that light is properly coupled between the two sides of the connectors and to minimize losses. As a result, the tolerances of the individual components before, and after, assembly is very important.

In the case of connectors, the connectors typically mounts onto a mating part that, for example, is mounted on a module in a precise way and the connector either'snaps'onto the mating part to self lock (and may be easily removable) or it is crimped onto the mating part.

Often, the mating part is manufactured to a particular specification so that connectors of a particular type, typically from other manufacturers, can connect to it.

The snapping type mating part typically requires high precision in its design, placement and attachment, as well as most, if not all, other critical components on the mating side. If the mating part lacks the requisite precision in either its manufacture or placement, it can cause a misalignment which, in the particular case, may still be acceptable, may be wholly unacceptable, or may be somewhere in between, depending upon the particular application in which it is being used.

The crimping type mating part has the same problems and further, is subject to the introduction of errors in alignment during crimping.

Moreover connectors are typically crimped or epoxied in a manner which may allow for adjustment in the X or Y planes, but provides little or no flexibility in adjusting for imprecision in the Z direction (i. e. the distance along the direction the connector is typically inserted) or for inaccuracies in"roll","pitch"or"yaw".

In either case, this variability leads to increased cost, both in labor and materials, because the tolerance error in any two given pieces may be within their respective specifications but their combined errors are sufficiently great so as to cause problems in mating or with the efficiency of the coupling of light. Thus, even if such problems are not fatal to use of the components, they may nevertheless adversely affect performance and/or reliability.

SUMMARY OF THE INVENTION We have recognized the above problems and created a way to: 1) ensure that the mating part to which connectors are attached are put on with high alignment precision, 2) ensure mating parts are secured permanently (absent unusual circumstances) on a module, and 3) allows flexibility in at least the three dimensions so as to reduce the precision required on pieces inside the connector and/or the mating part.

In general, the invention relates to the integration of connectors. More particularly, we have devised an optical connector mating unit, that can be mounted on another component, for example, an opto-electronic module, in a way that provides for high alignment accuracy between the two while reducing the accuracy requirements needed for those parts while ensuring that a good connection between an optical connector and the other component will be created.

Depending upon the particular implementation, our approach may have one or more of a number of advantages, including simplicity of manufacture, significantly lower tolerance on parts for connection, real-time flexibility on distance in the'Z'dimension, flexibility in roll, pitch or yaw positioning, and an ability to compensate for slight variations in individual components on which the mating portion is mounted or to which the mating portion connects.

The advantages and features described herein are a few of the many advantages and features available from representative embodiments and are presented only to assist in understanding the invention. It should be understood that they are not to be considered limitations on the invention as defined by the claims, or limitations on equivalents to the

claims. For instance, some of these advantages are mutually contradictory, in that they cannot be simultaneously present in a single embodiment. Similarly, some advantages are applicable to one aspect of the invention, and inapplicable to others. Thus, this summary of features and advantages should not be considered dispositive in determining equivalence. Additional features and advantages of the invention will become apparent in the following description, from the drawings, and from the claims.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 shows, in perspective view, a mating portion of a connector, a module onto which the mating portion will be mounted, and a flexible spacer, all relative to a common set of X, Y, and Z axes; FIG. 2 shows one example flexible spacer; FIGS. 3A-3C collectively show one example of a process in accordance with the teachings of the invention; FIG. 4 shows example features of a chip array to be aligned with the features on an optical coupler ; FIG. 5 shows example features on an optical coupler to be aligned with the features on the chip array of FIG. 4; FIG. 6 shows an example photo mask standard for aligning the features of FIG. 4 with the features of FIG. 5; FIG. 7 shows the photo mask of FIG. 6 relative to the images from FIGS. 4 and 5; FIG. 8 illustrate the steps involved in the alignment process; FIG. 9 is a photograph of an alignment apparatus constructed for operation in accordance with an aspect of the invention; and FIGS. 10A and 10B show, in generic form, alternative arrangements constructed for operation in accordance with an aspect of the invention.

DETAILED DESCRIPTION In overview, we use a flexible spacer layer between the mating portion and the component on which the mating portion will be mounted. This allows the easy positioning of the mating portion in all three (X, Y and Z) dimensions. The positioning in the X and Y direction being facilitated by the initial moveability of the spacer within the X-Y plane, and the positioning in the Z direction being facilitated by the flexibilty/compressibility of the spacer layer.

In one variant, the spacer layer is porous to provide a matrix into which an epoxy or other bonding agent can be applied, once the mating portion is aligned in the X, Y and Z directions, and then cured to permanently affix the mating portion to its mounting site, for example a location on an opto-electronic module. In another variant, the spacer layer is non- porous and the epoxy or other bonding agent is flowed around the spacer and hardened or cured to hold it in place. This permits any coupling elements inside the connector and behind the mating portion to be placed with far less accuracy, since only the final alignment involving the two pieces and the spacer need be precise.

As a result, because of the range of movement afforded by the flexible spacer, the accuracy of the connector or any alignment elements on either the connector or the module need not be precise. As long as the accuracy is such that the two pieces can be aligned relative to each other, for example, the connector can be aligned relative to the module, within the range of movement in the X, Y and Z direction the two pieces can be accurately aligned and attached. In addition, this approach allows for adjustments to also easily be made in roll, pitch and/or yaw as necessary as part of the process. This addresses a problem inherent in the prior art, in that if the pieces to be mated were individually or when combined collectively out of specification due to variances in roll, pitch or yaw, the time and labor cost of correcting for such errors typically exceeded the cost of the parts such that they were either scrapped or salvaged through trial and error attempts with other parts on an ad hoc basis.

Thus, our arrangement simplifies the manufacturing of product by only requiring one high precision step, this alignment step. In addition, our approach reduces the need to purchase parts made to extremely high accuracy (thereby reducing costs), reduces the likelihood that slightly out-of-specification parts will be rendered unusable, and/or reduces the costs associated with attempts to use slightly out-of-specification parts.

Referring now to the figures, FIG. 1 shows a mating portion 1-100 of a connector, a module 1-102 onto which the mating portion will be mounted (which can include any of an optical receiver, an optical transmitter, or an optical transceiver), and a flexible spacer 1-104, all relative to a common set of X, Y, and Z axes. As shown, the flexible spacer has a thickness 1-106 that provides an ample range of motion in the X, Y directions through shearing flex, motion in the Z direction based upon compressibility in the Z direction,-and for roll, pitch and or yaw pivoting, due to the ability to twist or compress the spacer rotationally about the X, Y and/or Z axes as required to achieve the desired alignment.

Depending upon the particular implementation, the mating portion 1-100 will typically be constructed and shaped on a portion 1-108 to accept and mate with a

commercially available optical connector, such as connectors of the commonly available commercial ST, LC, MT-RJ, MTPO, MPO, MPX and SMC styles, to name a few examples.

FIG. 2 shows one example flexible spacer 2-200 having a square hole 2-200 to allow for part of the connector and/or light signals to pass and thereby allow for optical access between the connector and module 1-102 (not shown in FIG. 2).

One material suitable for use as the flexible spacer is commercially available from Minnesota Manufacturing and Mining Co. (3M) under the name 3M Silicone Rubber Tape Model S8020T-031A. An epoxy usable as the bonding agent in conjunction with the 3m flexible spacer material is the EP65HT-1 epoxy commercially available from Master Bond Inc., 154 Hobart St. , Hackensack, NJ 07601.

FIGS. 3A-3C collectively show one example of a process in accordance with the teachings of the invention that used two pieces of reduced tolerance and a flexible spacer as shown in connection with FIG. 2, using a module 3-102 onto which the mating piece 3-100 will be mounted. FIG. 3A shows each of the pieces 3-100,3-102, 3-104 after the mating piece and module have been aligned in the X, Y, roll, pitch and yaw directions.

FIG. 3B shows the pieces of FIG. 3A 3-100,3-102, 3-104 after they have been brought together following alignment in all but the Z direction. FIG. 3C shows the pieces 3- 100,3-102, 3-104 of FIG. 3B after alignment in the Z direction and bonding together to form an integrated and aligned unit 3-300. There are many ways for aligning the two pieces.

For example, in the case of a fiber optic connection, a connector coupled to a constant light source can be connected to the mating piece and photodetectors in the module can be enabled and the incident light level detected by the photodetectors is monitored. The two pieces are moved relative to each other in the X, Y, Z, roll, pitch and yaw directions until the incident light detected by the photodetectors is maximized.

An alternative method, uses a high precision jig that holds each of the two pieces. A partially transparent alignment standard having alignment marks common to both pieces is placed between the two pieces. A camera is brought between one piece and the standard so that the other piece can be viewed through the standard. The other piece is then aligned relative to the standard. Next, the camera is moved to between the other piece and the standard and the first piece is aligned relative to the standard. Once this is done, the pieces are aligned in all but the Z direction. The standard is then removed and the pieces are brought together and aligned solely in the Z direction using, for example, the light method described above.

An another alternative method for alignment is shown in commonly assigned United States Provisional Patent Application Serial No. 60/302,246 filed June 29,2001 and United States Patent Application filed June 27,2002 both entitled"Simple Deterministic Alignment Method for Array Based Optical Component Packaging", the description of which is contained below in the section of the same title.

We turn now to a specific detailed example in accordance with the teachings of the invention with reference to FIGS. 3A-3C. For simplicity, the actual alignment method used will not be discussed in detail in this example, since the use of different methods of determining exactly when the pieces are in alignment will not typically materially alter the process as described. Rather, this simple explanation will provide the necessary details for an understanding of the invention as defined by the claims annexed hereto.

First, as noted above, the module 3-102 and the mating piece 3-100 are each held in precision holders that allow for precision adjustment in at least the X, Y and Z direction, and, in some more commercially suitable variants, optionally in each of the 6 degrees of movement (X, Y, Z, roll, pitch and yaw).

The two pieces are then aligned in the X and Y directions and, if necessary and/or desired and/or available, for rotation in the XY plane (i. e. about the Z axis).

Then the mating piece 3-100 is moved in the Z direction until the mating piece 3-100 is in contact with a flexible spacer, for example the flexible spacer of FIG. 2, that is, in turn, in contact with the module 3-102.

At this point, the mating piece is moved in the Z direction, and the spacers compressed, until the connector is the exact height required to optimize the coupling of light through that particular connector piece is obtained. If possible and desired or necessary, the mating piece 3-100 is also moved in rotation about one or both of the X and Y axes.

Depending upon the particular case, the pitch and/or yaw adjustment, if made, can be made before the Z alignment is complete or after it is complete. Of course, the Z, pitch and/or yaw alignment can be done concurrently, although such an approach may not be desirable in particular instances.

Advantageously, this apparatus and approach allows for minor manufacturing variations between mating pieces and/or modules in a manufacturing run to be accounted for.

Once the Z direction distance, and if used, rotation in the XZ, XY and YZ planes, is established, an epoxy or other bonding agent is flowed around, or, in the case of a porous spacer, into the spacer. The epoxy or bonding agent is then allowed to harden or is cured to hold the pieces in place permanently.

Once this is done, the alignment between joined pieces will be fixed, absent destructive or detrimental circumstances not relevant to understanding the invention.

As a result, by using the principles of the invention, the mating piece and the optical module can be aligned relative to each other in the X, Y and Z directions and in roll, pitch and yaw, to an accuracy that is greater, in one or more of the roll, pitch and yaw directions than could be achieved if the mating piece and the optical module were directly abutting each other without the flexible spacer in between.

It should also be understood that, in different variants, the alignment in the Z direction can precede the alignment in the X and/or Y direction. Similarly, adjustment of alignment in roll, pitch and/or yaw can occur before, during or after any of the above, the only requirement being that the flexible spacer must be interposed between the two pieces prior to final alignment in the Z direction (i. e. along the Z axis).

Moreover, in other variants, a physical feature located on one or both of the two pieces to be mated can be used to establish one or both of the X and Y direction alignment.

In such variants, alignment using the spacer need only be performed in the Z direction and the other non aligned directions. For example, if a physical feature establishes accurate alignment in the X direction, alignment in the Y and Z directions, and if desired and possible, in roll pitch and/or yaw need be performed. The analogous circumstances apply for the other non-Z alignments.

Simple Deterministic Method For Array Based Optical Component Packaging People currently align connectors to modules but typically do it via either an active optical alignment scheme (where they emit light into or from individual devices) or use very small numbers of devices where an accurate pick & place machine can get integration alignment. For example, using one laser where there is no concern regarding rotational alignment.

The processes typically used for alignment of connectors requires the individual devices be illuminated and then the fiber (s) are scanned across the optical device with the output light from the end of the fiber monitored for the intensity of light output. This process is repeated and the fiber light output is continuously monitored as fibers are moved in several dimensions to allow accurate alignment. An example of this technique is described in, for example, IBM Micro News, Volume 6, Number 3, Third Quarter 2000.

Such techniques are costly, since requiring illuminating devices necessitates the use of significant capital equipment to power up each device, to monitor the output powers, etc.

Moreover, because the techniques are active device techniques, they run the risk of damaging the devices.

We have devised a passive technique for aligning a connector containing an array of optical fibers with an optical module containing an array of optical devices prior to attachment. Furthermore, these techniques can be used, but are not limited to, for the following alignments : aligning an array of optical fibers with another array of optical fibers; aligning an array of optical fibers with an optical chip; and aligning a micro-lens with an optical chip. These techniques are not limited to any particular optical devices, the devices could be lasers, cameras, detectors, modulators, micro-electronic mechanical systems (MEMS) or other devices.

In the optical device fields, alignment of connector pieces onto modules is crucial to proper operation.

We have devised a simple, passive deterministic method toward alignment of components for array based transmitter, receiver or transceiver packaging.

Our approach uses an element having features common to each of the devices to be aligned as a central standard. Each of the devices can then be passively aligned to the standard which, in turn, causes the pieces to be in alignment relative to each other. As a result, the pieces will be aligned relative to each other when they are brought together.

The packaging alignment techniques allows the devices to be optimally coupled with an optical coupler.

FIG. 4 shows example features, in this case lasers and detectors, that need to be aligned with the features on an optical coupler (shown in FIG. 2). Shown in FIG. 4 is a chip array 4-100. chip array 4-100 has a laser array 4-110 and a detector array 4-120. FIG. 5 illustrates an exemplary optical coupler 5-200. Optical coupler 5-200 has feature array 5-210 that match with the laser array 5-110, and feature array 5-220 that match with detector array 5-120.

As shown in FIG. 6, the approach creates a photo mask consisting of a properly aligned superposition of the two sets of features to be aligned. An example photo mask 6- 300 for the features of FIGS. 4 and 5 is shown in FIG. 6. Namely, set 6-310 having feature array 6-210 superimposed with the laser array 6-110, and set 6-320 having feature array 6- 220 superimposed with detector array 6-120.

FIG. 7 shows the photo mask 7-300 placed between the two images from FIGS. 4 and 5, namely chip array 7-100 and optical coupler 7-200 where it can be readily seen that both items features are contained on the photo mask in proper alignment.

This photo mask 7-300 thus serves as the alignment standard. By comparing the photo mask 7-300 with the actual devices and the optical coupler the offset between each device and the corresponding optical couplers can be visually determined at once with high accuracy. Therefore, without actually turning devices on, the loss figures of the optical path can be determined. In addition, alignment using the mask may be accomplished utilizing one or both sides of the photo mask 7-300.

In overview, the actual alignment process proceeds as follows using high accuracy, low back-lash XYZ stages, with continuing reference to FIG. 8.

Shown in FIG. 8 is an exemplary method of alignment. It is understood that these individual steps may be done in any order and/or may delete or add steps of the method depending on the implementation desired Shown in block 8-500 is the alignment process of a connector assembly or optical coupler chip 8-504 with a photo mask 8-502 utilizing a camera 8-501. Shown in block 8-510 is the alignment process of an optical device chip with a photo mask 8-502 utilizing camera 8-501. These alignments may be done individually or at the same time depending on the implementation. The camera 8-501 may be replaced with a laser, microlens, or any other device that can be used in the assistance of aligning the optical device chip 8-511 with the fixed mask 8-502.

In one example, the device chip 8-511 is held on one end of a stage and a coupler or connector assembly 8-504 to which the devices are to be aligned is held in the other end of the stage. The mask 8-502 is placed in between in a position that will remain fixed throughout the alignment process or, in certain variants, can be removed and replaced with high accuracy.

The device chip 8-511 is then viewed through the mask using a zooming viewing scope or camera or any other device that can be used in alignment procedures. The device chip 8-511 is then moved out of the way using the XYZ stage, so that it can be accurately replaced later. After that the optical coupler chip 8-504 having a fiber array 8-503 is aligned in a similar manner with the fixed mask 8-502 using a zooming viewing scope from the other side. Again, the mask can be utilized on one or both sides. Depending on the implementation, the order of alignment may be reversed as shown in FIG. 8. Shown in FIG.

8 is the optical coupler or connector assembly 8-504 being aligned with mask 8-502 prior to alignment of device chip 8-511. Again, either one side or both sides of the mask can be used in this alignment process as well as alignment at the same time or individually, as shown in block 8-500 and block 8-510. Shown in block 8-500 is mask feature 8-505 and optical coupler feature 8-506 that are brought into an alignment position 8-507. Shown in block 8-

510 is mask feature 8-505 and optical device chip feature 8-508 that are brought into an alignment position 8-509. Block 8-520 illustrates alignment of the optical device chip 8-511 with optical coupler chip 8-504. Again, any optical component may utilize this technique of alignment. For example, the optical components given in this description as examples, specifically aligning a connector containing an array of optical fibers with an optical module containing an array of optical devices prior to attachment, may readily be replaced with aligning an array of optical fibers with another array of optical fibers; aligning an array of optical fibers with an optical chip; and aligning a micro-lens with an optical chip.

The photo mask 8-502 is then removed as shown in block 8-530 and, optionally, depending upon the separation distance between the two, the device chip 8-511 is moved away from the optical coupler 8-504.

The device is then axially adjusted relative to the optical coupler position to optimize the coupling efficiency. This adjustment and coupling is shown in block 8-540.

Advantageously, it should thus be recognized that the whole process is simple and deterministic.

Moreover, by using a simple deterministic approach, transceiver packaging cost and complexity is reduced.

In particular, the approach proceeds as follows.

A filter mask, which contains features, which resemble both the optical fiber array and the optical device array, is created or, if previously created, attached to the center of the XYZ stage. In the example in FIGS. 4-6 above, the optical device array contains both lasers and detectors (though it could contain a myriad of other devices), which have different sizes and orientations. The fiber array has yet a third size and shape in this example. Thus, the filter mask has all three of those features on it to act as an absolute positioning standard to which all of the pieced can be aligned.

The mask has a series of elements, which correspond to the elements on both the optical fiber array (also known as the optical coupler) and to the optical chip array (i. e. the laser and detector arrays).

Once the mask standard is positioned, alignment can begin.

The optical chip array and optical fiber array/connector assembly are mounted on a high precision, reproducible, low-backlash stage. The optical chip array is then moved away to accommodate a camera or, if there is enough space, the camera is merely interposed between the optical chip array and the standard.

The optical fiber array is then brought close to the mask and the camera is used to look through the mask at the optical fiber array.

The fiber array/connector assembly is moved around in a plane parallel to the mask, as well as for roll, pitch and yaw, until the fiber elements align to the corresponding elements on the filter mask as observed using the camera. Once alignment is achieved, the position of the fiber array/connector assembly is noted.

The fiber array/connector assembly is then moved aside on a high precision, reproducible, low-backlash stage (so that later it can be repositioned to its previously noted position above). The optical chip assembly is then positioned near the filter mask. As was done for the optical fiber array, a camera is then brought in and used to look through the mask at the optical device chip. The optical device chip assembly is then moved around in a plane parallel to the mask, as well as for roll, pitch and yaw, until the optical elements on the chip align to the corresponding elements on the mask standard as observed using the camera.

Once alignment is achieved, the position of the optical chip assembly is optionally noted.

It should be appreciated that, although the alignment was described in a particular order, the chip array could have been aligned first. Alternatively, the first component could be aligned to the standard before the second component is even mounted.

In any case, once the two have each been aligned relative to the mask standard, the camera is moved aside and the fiber array/connector assembly is repositioned to its aligned location. At this point, the optical device assembly and the fiber array/connector assembly are aligned accurately in'X','Y', and Rotational dimensions as well as in tilt.

Next, the filter element is moved aside from the central region between the fiber array/connector assembly and the optical chip assembly.

The two aligned pieces are then brought together in the'Z'dimension until they are in contact and secured together.

FIG. 9 is a photograph of an assembly station 9-600 used for the alignment process described herein. Through use of this station to align various pieces including fiber bundles, optical chips and connector assemblies, we have achieved an accuracy of, as low as, 20 nanometers of tolerance.

In other alternative variants, the same approach can be used with a single camera so long as the camera can be accurately and reproducibly be moved from one position to another.

In still other alternative variants, another device, such as a laser, a photodetector (detector) a non-coherent light source, etc. can be used in place of the camera as the device used to check alignment between a given component and the standard, such as the photo mask.

FIG. 10A shows, in generic form, such an arrangement. One of the components 10- 700 to be aligned is mounted on a moveable high precision stage 10-702. Another of the components 10-704 to be aligned is also mounted on a moveable high precision stage 10-706.

The standard 10-708 is located between the two components 10-700,10-704. As shown in FIG. 10A, the arrangement is constructed so that at least one of the components to be aligned is moved out of the way so that the device 710-10 that is used to check alignment of the other component can be moved in its place. Alternatively, for example as shown in FIG. 1 OB, if spacing permits, the device 10-710 can be interposed between the standard 10-708 and a component to be aligned. The alignment then proceeds as described herein, first for that component and then for the other component.

It is to be understood that these techniques are not limited to alignment of any particular optical devices or combinations thereof, the devices could be lasers, cameras, detectors, modulators, micro-electronic mechanical systems (MEMS) or other devices.

In summary, by using a passive deterministic approach to alignment advantages not present in the prior art can be achieved. For example, by not illuminating the individual devices, we can perform alignment 1) more quickly, and 2) with lower cost of capital equipment for each assembly station.

By making a purely passive system, the cost of capital equipment is minimal (essentially the cost of the translation stages and camera (s), lasers or other optical devices used in the alignment process. In addition, setup and insertion of the module components to prepare for alignment can also occur much more rapidly when module components are passively aligned than when they are actively aligned, thereby reducing labor costs.

It should be understood that the above description is only representative of illustrative embodiments. For the convenience of the reader, the above description has focused on a representative sample of all possible embodiments, a sample that teaches the principles of the invention. The description has not attempted to exhaustively enumerate all possible variations. That alternate embodiments may not have been presented for a specific portion of the invention, or that further undescribed alternate embodiments may be available for a portion, is not to be considered a disclaimer of those alternate embodiments. One of ordinary skill will appreciate that many of those undescribed embodiments incorporate the same principles of the invention and others are equivalent.