Mitchell S. Cohen

Passive laser-fiber alignment by index method

A method for packaging a laser-fiber module was explored in which the critical laser-fiber alignment is carried out not with the usual active means, i.e., with the laser activated, but by a passive method based on the registration principles of photolithography. The method relies on an index scheme in which fiducial marks are lithographically placed on the laser chip and on a fiber carrier. At 850 nm, using the index technique with cleaved multimode fibers, it was possible to achieve the same laser-fiber coupling efficiency as attained by active alignment; with cleaved single-mode fibers about 80% of the active-alignment coupling efficiency was achieved.<<ETX>>

Packaging of high-density fiber/laser modules using passive alignment techniques

A novel method for packaging a laser-fiber module is explored in which the critical laser-fiber alignment is carried out not with the usual active means, i.e., with the laser activated, but by a passive method based on the registration principles of photolithography. The novel method relies on an index scheme in which fiducial marks are lithographically placed on the laser chip and on a fiber carrier. At 850 nm, using the index technique with cleaved multimode fibers, it has been possible to achieve the same laser-fiber coupling efficiency as attained by active alignment; with cleaved single-mode fibers about 80% of the active-alignment coupling efficiency was achieved. Details of the novel index-alignment method are discussed, and a brief description of an improved computer-controlled version of the apparatus is given.<<ETX>>

Low-cost fabrication of optical subassemblies

The optical subassembly is a major contributor to the cost of a fiberoptic computer data link. A technology for low-cost fabrication of optical subassemblies is described, with emphasis on the transmitter subassembly. A factor of three cost reduction is achieved by limiting the parts count to only three: a laser or receiver chip packaged in a TO-can, a plastic housing, and a plastic aspheric lens; and by employment of a fast, automated active-alignment and subsequent fixing technique. Key enabling features include the rise of precision injection molding of specially chosen plastics, an aspheric lens design which permits wide positional variations in the axial direction, and curing of a fast setting epoxy through the use of RF power. A tool was constructed which produced subassemblies at high yield having satisfactory performance.

A novel low-cost small-form-factor transceiver module

A key trend in the optoelectronic data-communication industry is the move towards lower cost and smaller transceivers. This trend has driven transceiver modules to the Small-Form-Factor (SFF) dimensions, requiring SFF optical connectors, resulting in a cost lower than their larger predecessors (for example GBICs or 1/spl times/9 transceivers based on the larger duplex SC optical connector). The design and implementation of a low-cost SFF fiber optic transceiver based on the SFF LC fiber optic connector is discussed. Since most of the transceiver cost is associated with the optical subassemblies (OSAs), low cost, yet smaller OSAs are required. In addition, the entire transceiver assembly has been optimized for cost reduction by using direct-chip-attach electronics, and surface-mount components all mounted on a single card that snaps into a plastic-molded retainer. These SFF LC transceivers are well suited for Gigabit Ethernet, Fibre Channel, and 1394b applications, along with other applications requiring low cost data transfer at data rates of 1.25 Gb/s to 2.125 Gb/s, or higher.

Packaging aspects of the Jitney parallel optical interconnect

The Jitney parallel optical interconnect is a prototype 20-channel wide, low cost data link, designed for operation at speeds up to one Gbyte/s over distances approaching 100 meters. The package is based on (1) an inexpensive overmolded leadframe, (2) passive optical alignment, and (3) plastic molded parts. The packaging challenges in the fabrication of the components, their assembly, and in the achievement of the performance goals are described.

Lithographically fabricated fiber guides for optical subassemblies

A new optoelectronic packaging technology is presented which permits highly accurate fiber-to-chip alignment at low cost. Instead of the more common methods employed in fabricating optical subassemblies, which use leadframe, precision plastic-molding, or silicon-optical-bench technology, here fiber guides are fabricated in a photoresist by use of standard photolithographic procedures. By this means the fiber guides are directly created on an entire wafer of either VCSELs or receivers, resulting in structures with very tight dimensional tolerances fabricated at very low cost. After dicing, a fiber is interfaced with a chip under computer control using a very simple semiautomatic tool to insert the fiber into the fiber guide. This new technology may be used in the fabrication of a wide variety of single or multi-channel optoelectronic transceivers.

Influence of NiFe quality on bubble devices

It is anticipated that as the period of bubble devices decreases, undesired effects due to loss of permeability of the NiFe overlay will become increasingly important, even for half‐disk elements. The failure mechanism of ’’I‐bar crossing ’’ in H‐I tracks, which leads to high values of the minimum drive field Hxy(min), was studied for both 8 and 20 μm bubble periods as a means of gaining insight into such effects. Direct Ferrofluid observations, along with measurements of the parameter Hbb1 which characterizes the coercivity of an individual H bar, confirmed that the I‐bar crossing phenomenon originated in high coercivity, and hence high remanence and low permeability of the NiFe. This interpretation was supported by a study of the effects on a sample of a series of annealing treatments at increasing temperatures, a process known to increase NiFe coercivity. The annealing studies demonstrated that for a given NiFe thickness Hbbl was better correlated with Hxy(min) than was the sheet‐film coercivity Hc, sh...

Making negatives and plates for printing by electroerosion: II. larger-scale fabrication and testing

The principles of producing direct negatives and direct plates by electroerosion writing were given in Part 1 of the present series of papers. Part II is concerned with larger-scale fabrication techniques, characterization methods, and the results of tests, together with their interpretation. Major problems which were encountered are also presented with their solutions. sheet material for the direct negative/direct plate (DNP) was made by 1) Coating polyester rolls with an underlayer. The underlayer coating fluid contained silica, a cellulosic binder, a saturated polyester dispersant, and an isocyanate cross-linker. The coating fluid was carefully milled to control silica particle size before coating. 2) Curing the rolls either at ambient or elevated temperature. 3) Calendering the underlayer. 4) Vacuum-depositing an aluminum film. 5) Coating with an overlayer containing graphite and a binder. Maintenance of good control of material characteristics was found essential for acceptable functional performance. Among the parameters requiring control were underlayer thickness and surface roughness, aluminum thickness, and overlayer thickness, as well as the composition of the various components. After fabrication, functional testing of the DNP material was carried out on the IBM 4250 printer in order to study the major performance problems of scratching, writing failure, head wear, gouging, and head fouling. Scratching could be suppressed by decreasing the surface roughness and increasing the thickness of the overlayer. Writing efficiency could be improved by increasing the roughness, decreasing the overlayer thickness, decreasing the aluminum thickness, and increasing the pulse length. Head wear could be suppressed by calendering and reducing the roughness. Gouging and excessive head wear could be suppressed by adequate milling, dispersing and filtering of the underlayer coating fluid, and calendering and curing of the coated web, while ensuring good underlayer-substrate adhesion. Fouling was reduced by decreasing the overlayer thickness and reducing the writing pulse length.

Making negatives and plates for printing by electroerosion: III. use of the direct negative and direct plate

Issues related to the practical usage in the pressroom of both the direct negative and direct plate are discussed. Two concerns associated with the usage of the direct negative during platemaking are treated: 1) the effect of light transmission through defects (voids) in the aluminum film, and 2) the effect of light transmission directly through the aluminum. These concerns may be addressed respectively by careful fabrication of the material and careful control of the exposure conditions during platemaking. Two aspects of the usage of the direct plate are considered: 1) the need for a simple prepress “activation” treatment, and 2) press-life limitations caused by wear of the direct plate on the press. Passing a direct plate through an activator solution immediately prior to mounting it on the press prevented scumming in the image area and blinding in the background (nonimage) area. The life on the press of the direct plate is limited by the wearing away of the aluminum from the background areas, thereby causing scumming. Changes in the structure of the direct plate which could prolong press life were investigated.

Making negatives and plates for printing by electroerosion: I. physical principles

Electroerosion printing involves removal of the aluminum overlayer from selected areas of a black-coated paper. “Direct negatives” as well as “direct plates” for use in offset lithographic printing may also be generated by electroerosion if a clear polymer sheet is used as the substrate instead of paper, and the black base layer is omitted. If such a substrate is metallized and written by electroerosion, the desired direct negative is created in principle since the metal stops transmitted light and the polyester does not. The direct plate is simultaneously created in principle since the aluminum is hydrophilic and the polyester is hydrophobic. Practical realization of these concepts required studies of the physical principles of the processes involved, which led to techniques for avoidance of mechanical scratching of the aluminum film during writing. For this purpose a mechanically hard underlayer was applied to the substrate under the aluminum, while a very thin lubricating overlayer having some electrical conductivity was applied over the aluminum. The underlayer consisted of silica particles in an organic binder, while the overlayer consisted of graphite particles in a binder. Although scratching is less for smooth than for rough underlayers, rough underlayers were preferred because they offered better writing reproducibility. In particular, debris created during writing was scoured away from the styli in rough-underlayer samples. For writing, a two-phase driver was used, in which the first phase provided a high current for Joule heating with consequent breaking of direct local aluminum-stylus contacts, while the second phase provided an arc which removed the remainder of the aluminum under the stylus.