Abstract:
A method of fabricating BiCMOS devices, and the resultant BiCMOS devices are disclosed. According to the present invention, over-etching to the substrate (2) on the deposited polysilicon emitter (25) is prevented by providing additional oxide (22) beneath a polysilicon layer as an etch stop. Despite inclusion of an oxide to define an endpoint during patterning of an emitter, fabrication complexity is reduced by avoiding additional SAT (self-aligned transistor) masking and oxidation steps.
Abstract:
Ultra-small equal-width lines and spaces are generated on an integrated circuit wafer using multiple exposures and phase-shifting at the wafer level. In particular, an integrated circuit wafer is coated with a layer of photoresist and then masked using a mask defining a pattern of multiple feature lines arranged at a regular line pitch. The layer of photoresist is then underexposed so as to partially bleach portions of the layer of the photoresist in accordance with the pattern. Next, the mask and the integrated circuit wafer are positionally translated relative to one another by a predetermined fraction of the line pitch, and the layer of photoresist is then again underexposed. Developing the photoresist layer creates a stepped profile. The layer of photoresist is then blanket exposed, the stepped profile causing exposure in the vicinity of steps to be retarded. The layer of photoresist is then developed, producing thin lines of photoresist separated by substantially equal spaces of no photoresist.
Abstract:
Test jig apparatus for testing an integrated circuit chip that suppresses build-up and subsequent discharge of electrical charge on the test jig apparatus or on the chip. The test jig apparatus includes a base (33) of selected material having a top surface (34) of the same general shape and dimensions as the chip to be tested. Preferably, the entire top surface of the base is electrically grounded. The base has two or more side surfaces (37, 39) with side surface planes that are approximately perpendicular to a plane defining the top surface of the base. Each side surface accepts a side plate (47, 49, 51, 53), made of a selected material such as ULTIM, than can be attached to or removed from the base. The side plane material resists electrical charge buildup and subsequent discharge so that the chip being tested is not subjected to electrical discharge from this source. In another embodiment, the side plates are replaced by plates (67, 69) mounted on the top surface of the base.
Abstract:
A computer-aided design method and apparatus for compacting semiconductor circuit layouts to meet a specified set of design rules begins by fracturing a specified circuit layout into a set of trapezoids and storing the resulting cells in a database identifying the boundaries of each cell, and the cell adjacent each boundary. The method additionally includes a procedure for minimizing wire lengths in the compacted layout. Nonempty cells are identified as being of specific materials, and empty spaces between cells are represented. For each cell boundary, the database also stores data representing the boundary edge's end points. Neighboring cells on the same and related layers of the circuit layout share edges in the database. When a point on an edge of a cell is moved, the edge of each neighboring cell that shares that point is also moved. To adjust a circuit layout, the cells in the layout are processed in a sorted order. For each cell, width and spacing design rules are applied to first and second edges of the cell, which may result in movement of the cell and/or adjustment of the cell's width. An adjacent cell adjustment process conforms cells shared points with the adjusted cell. Once the compaction procedure is carried out in a positive x-direction, the coordinates are reversed and it is carried out again in a second pass. The compaction method is performed once for x-direction compaction and once for y-direction compaction. The compaction process is computationally efficient because each cell is linked by the database to its adjacent cells, virtually eliminating the need to search through cells not needed for each design rule check, and because movement of one cell's edges automatically adjusts adjacent cells via their shared edges. The wire minimization embodiment of the invention includes determining whether the second edges have been moved in the second pass, and if so applying the width rules to the first edges to minimize the length of wire represented by the width of the cell having the adjusted edges.
Abstract:
A spacer tray is disclosed. This spacer tray finds use in handling or shipping semiconductor packages having surface projections which extend the height of the semiconductor package beyond normal semiconductor height parameters. The spacer tray is generally used in conjunction with a standard carrier tray which holds the semiconductor base (i.e., the semiconductor body and leads). The spacer tray comprises a tray body having at least one bore, the bore being sized to house the semiconductor surface projection(s) while excluding the semiconductor base. Methods of use are also given.
Abstract:
A digital filter for a phase-locked loop operates to compare the bit values of the data represented by the incoming pulse stream (10) with patterns in that pulse stream known to produce bit shifting (either early or late). The bit shift caused by physical interaction of bits encoded on computer diskettes or the like always is predictable, in accordance with the pattern of previous bits, the current bit and the next bit. This information is processed by a logic circuit (18, 21) to predict which pulses in the incoming stream of data pulses are shifted. A signal is produced each time a predicted shifted pulse is determined; and this signal is utilized (in 21) in conjunction with the output of a phase difference counter (20) in the digital phase-locked loop to permit the phase of the controlled oscillator (14) to be adjusted at each unshifted bit in a normal manner, and compensated for adjustment in a modified manner as a result of the prediction of the shifted pulses.
Abstract:
For a structure with an overlapping gate region, a first insulator layer (35) is placed on a substrate. A source/drain polysilicon layer (37) is placed on the insulator layer. The source/drain polysilicon layer is doped with atoms of a first conductivity type. A second insulator layer (38) is placed on the source/drain polysilicon layer. A gap (39) is etched in the second insulator layer and the source/drain polysilicon layer to expose a portion of the first insulator layer. The exposed portion of the first insulator layer and an additional amount (40) of the first insulator layer under the second insulator is etched so as to enlarge the gap and to undercut a portion of the source/drain polysilicon layer. Two polysilicon filler regions (42, 43) are formed which fill a portion of the gap including the undercut area under the source/drain polysilicon layer. A gate polysilicon region (45) is formed in the gap and extends over the source/drain polysilicon layer. The gate polysilicon region is separated from the source/drain polysilicon layer and the polysilicon filler regions by a dielectric region (44). Source/drain regions (47, 48, 49, 50) are formed by atoms in the source/drain polysilicon layer diffusing through the polysilicon filler regions into the substrate.
Abstract:
For passivation of an integrated circuit device, a film of silicon dioxide (32) is deposited over the integrated circuit device. A film of silicon nitride (33) is deposited over the film of silicon dioxide. The film of silicon nitride and the film of silicon dioxide are etched using a single passivation mask to expose the bond pads of the integrated circuit device. Spacer regions (35) of silicon nitride are placed over edges of the film of silicon dioxide exposed by the etching. The spacer regions may be placed by depositing a second film of silicon nitride (34) over the film of silicon nitride. This second film of silicon nitride covers the metal bond pads and the exposed edges of the film of silicon dioxide. An anisotropic etchback of the second film of silicon nitride is performed to expose the metal bond pads while leaving spacer regions which cover the edges of the film of silicon dioxide.
Abstract:
A method for making an anti-fuse structure characterized by the steps of forming a conductive base layer (10); forming an anti-fuse layer (20) over the base layer; patterning the anti-fuse layer to form an anti-fuse island (24a, 24b); forming an insulating layer (26) over the anti-fuse island; forming a via hole (28a, 28b, 28c) through the insulating layer to the anti-fuse island; forming a conductive connection layer (32) over the insulating layer and within the via hole; and patterning the conductive connection layer to form a conductive contact (30a, 30b, 30c) to the anti-fuse island. Preferably, the anti-fuse island comprises amorphous silicon which can optionally be covered with a thin layer of a titanium-tungsten alloy.
Abstract:
After a semiconductor die (20) is placed onto a leadframe (22) and electrically connected to the die (20), the die (20) and the ends of the leads (26) adjacent to the die (20) are encased in a packaged body (30). The exposed ends (26c, 26b) of the leads (26) are trimmed so that the leads (26) are of desired lengths for leadforming, or for connection to substrates or sockets. The ends of the leads (26) are enclosed in a protective body (60), so that when the package is tested and handled, the protective body (60) reduces indesirable bending of the leads (26). By trimming the leads (26) before forming the protective body (60), the leadframe (22) used need not be larger than those normally used.