Abstract: Resistor string DAC's are known to utilize lots of area and slow in data conversion due to the large utilization of switches. The problem becomes worse when differential outputs are required in the conversion process. This invention describes a N-bit DAC architecture utilizing a substantially lower number of switches through a unique placement of tap-points in the resistor string and decode logic. Differential outputs share the same set of switches through 2 levels of decoding.

Abstract: A microstructure relay is provided, having a body that includes upper and lower portions. The lower portion is formed from a substrate, and the upper portion is formed on the substrate to avoid bonding of the lower portion to the upper portion. A support member is fixed to the body at a first end of the support member to form a cantilever, wherein an upper surface of the support member and a lower surface of the upper portion of the body form a cavity. A first contact region is located on the upper surface at a second end of the support member. The first contact region comprises a first contact, wherein pivoting the support member toward the lower surface causes the first contact to be electrically coupled to a counter contact.

Abstract: It is often desirable to be able to perform an array of micro-chemical reactions simultaneously but with each reaction proceeding at a different temperature and/or for a different time. A classic example is the polymerase chain reaction associated with DNA analysis. In the present invention, this is achieved by means of an apparatus made up of a chip of plastic, or similar low cost material, containing an array of reaction chambers. After all chambers have been filled with reagents, the chip is pressed up against a substrate, typically a printed circuit board, there being a set of temperature balancing blocks between the chip and the substrate. Individually controlled heaters and sensors located between the blocks and the substrate allow each chamber to follow its own individual thermal protocol while being well thermally isolated from all other chambers and the substrate. The latter rests on a large heat sink to avoid temperature drift over time. A process for manufacturing the apparatus is also disclosed.

Abstract: A new method of fabricating solder bumps in the manufacture of an integrated circuit device has been achieved. Contact pads are provided overlying a semiconductor substrate. A passivation layer is provided overlying the contact pads. The passivation layer has openings that expose a top surface of the contact pads. A sacrificial layer is deposited overlying the passivation layer and the exposed top surface of the contact pads. The sacrificial layer is not wettable to solder. Under bump metallurgy (UBM) caps may be formed either by deposition and patterning of a UBM layer stack or by selective electroless deposition of a material such as nickel and gold. An aperture mask is formed overlying the sacrificial layer. The aperture mask has openings that expose a part of the sacrificial layer overlying the contact pads. A solder layer is printed into the openings in the aperture mask. The solder layer is reflowed to form solder bumps overlying the contact pads. The aperture mask is stripped away.

Abstract: Resistor string DAC's are known to utilize lots of area and slow in data conversion due to the large utilization of switches. The problem becomes worse when differential outputs are required in the conversion process. This invention describes a N-bit DAC architecture utilizing a substantially lower number of switches through a unique placement of tap-points in the resistor string and decode logic. Differential outputs share the same set of switches through 2 levels of decoding.

Abstract: A method for use in a real-time video encoder based on two processing means to optimise the use of computing power of the processing means. This is achieved by dynamically load-balancing between two processing means. The load pattern is determined with measures for the progress of motion estimation in one processing means and the idle time in the other. Adaptive adjustment of load balancing is fulfilled via making a decision on whether the DCT/IDCT/Quantization/Dequantization sub-process is skipped for a macroblock. The invention is not constrained to particular motion estimation techniques.

Abstract: A method of forming a via-first type dual damascene structure in the absence of an etch stop layer and without via-edge erosion or via-bottom punch-through is described. The invention uses two organic films deposited within via hole prior to trench etching. A via hole over a lower level metal line is first etched in the dielectric film. Two, preferably organic, bottom anti-reflective coating (BARC) films, first one being the conformal type to coat the surfaces and the walls of the via and the second one being the planarizing type to at least partially fill the via, are then deposited. Using a mask aligned to via hole, a wiring trench of desired depth is etched in the top portion of the dielectric film. During trench etching, the conformal BARC-1 film protects the via-edges from eroding and the planarizing BARC-2 film prevents punch-through of the via-bottom. Desired metal such as aluminum or copper are deposited within said dual damascene pattern.

Abstract: A method of forming interconnect structures in a semiconductor device, comprising the following steps. A semiconductor structure is provided. In the first embodiment, at least one metal line is formed over the semiconductor structure. A silicon-rich carbide barrier layer is formed over the metal line and semiconductor structure. Finally, a dielectric layer, that may be fluorinated, is formed over the silicon-rich carbide layer. In the second embodiment, at least one fluorinated dielectric layer, that may be fluorinated, is formed over the semiconductor structure. The dielectric layer is patterned to form an opening therein. A silicon-rich carbide barrier layer is formed within the opening. A metallization layer is deposited over the structure, filling the silicon-rich carbide barrier layer lined opening. Finally, the metallization layer may be planarized to form a planarized metal structure within the silicon-rich carbide barrier layer lined opening.

Abstract: A relaxation current controlled oscillator (CCO) is provided by forming an integrator out of a transconductance amplifier and a capacitor. The output of the integrator is fed to comparators which in turn feed a bistable circuit. The outputs of the bistable circuit control either the polarity of the input signals to the transconductance amplifier or the polarity of the input signals to the comparators. Switches, controlled by the bistable circuit, in turn control the polarity of the input signals. The feedback path created by the transconductance amplifier, comparators, flip-flops, and switches produces continuous oscillations. A DC current input adjusts the gm of the transconductance amplifier allowing the oscillation frequency of the CCO to be adjusted. Several embodiments of CCOs are described which are fully compatible with PLLs with automatic time-constant or bandwidth tuning of a gm-C filter.

Abstract: Method and apparatus for fluid-based cooling of heat-generating devices are disclosed. A heat-generating device is mounted on a carrier. The heat-generating device is spatially displaced from the surface of the carrier, thereby forming a channel. The heat-generating device and the carrier are enclosed in an enclosure having an inlet and an outlet. A substantially electrically non-conductive cooling fluid for introduction into the enclosure and into the channel and expulsion from the enclosure and for extracting heat from and thereby cooling the heat-generating device and the carrier.

Abstract: Formation of micro-fluidic systems is normally achieved using a multi-wafer fabrication procedure. The present invention teaches how a complete micro-fluidic system can be implemented on a single chip. The invention uses only dry etch processes to form micro-chambers. In particular, it makes use of deep reactive ion etching whereby multiple trenches of differing depths may be formed simultaneously. Buried micro-chambers are formed by isotropically increasing trench widths using an etchant that does not attack the mask so the trenches grow wider beneath the surface until they merge. Deposition of a dielectric layer over the trenches allows some trenches to be sealed and some to be left open. Micro-pumps are formed by including in the micro-chamber roof a layer that is used to change chamber volume either through electrostatically induced motion or through thermal mismatch as a result of its being heated.

Abstract: A process is described for trimming photoresist patterns during the fabrication of integrated circuits for semiconductor devices and MEMS devices. A combination of a low temperature (<20° C.), high density oxygen and argon plasma and intense UV radiation is used to simultaneously trim and harden a photoresist linewidth in an ICP chamber. As an alternative, a UV hardening step can be performed in a flood exposure tool prior to the ICP plasma etch. Another option is to perform the argon plasma treatment first to harden the resist and then in a second step apply an oxygen plasma to trim the photoresist. Vertical and horizontal etch rates are decreased in a controllable manner which is useful for producing gate lengths in MOS transistors of less than 100 nm. The process can also be used to controllably increase a space width in a photoresist feature.

Abstract: A chip scale package assembly comprises an integrated circuit die wire bonded to a carrier for mounting to a printed circuit board. The carrier comprises top and bottom ground planes thermally and electrically bonded together by a number of grounded thermal vias. The top ground plane completely surrounds the wire bond signal connections made with the die, enhancing signal integrity. The top ground plane covers the die mounting area, providing grounding and heat spreading for the die. The thermal vias are also positioned in the mounting area, and thermally couple the die to the bottom-side ground plane. The bottom ground plane is positioned within a central area around which the signal connections from the top-side are arranged. Ground pads with attached solder balls are positioned within the bottom ground plane and conduct heat transferred from the die into a primary circuit board on which the carrier is mounted.

Abstract: The present invention provides a circuit and a method for high speed prescaler circuits which utilize pull-down transistors in the critical feedback path. This invention contains a high speed CMOS dual modulus prescaler circuit made up of data or D-flip flops connected serially where the flip-flop positive output Q of stage N is connected to the D-input of the N+1 flip-flop stage. It is also made up of a pull-down field effect transistor. The invention has a clock input which has a frequency known as a circuit input frequency, Fin. The output of this prescaler circuit has an output frequency, Fout. The frequency division which results from this prescaler circuit is a divide by [2 to the power (n+2)] minus 1 if a mode signal equals 1 as opposed to a divide by [2 to the power (n+2)] counter, which results when the mode signal is low.

Abstract: A feedback image rejection downconversion system is described, which can be used in low IF receivers with good performance and completely integrated. In the forward path of the system, quadrature mixers and complex filters are used for frequency downconversion and separation of the RF signal from the image signal. In the feedback path, a correlator, a gain mismatch estimator and two VGAs have been used to detect, estimate and compensate the amplitude and phase mismatch between the forward I and Q path signals. The whole system is self-tuned and can operate in both closed and open loop mode. A very high and robust image rejection ratio (over 60 dB) has been achieved.

Abstract: A silicon condenser microphone is described. The silicon condenser microphone of the present invention comprises a perforated backplate comprising a portion of a single crystal silicon substrate, a support structure formed on the single crystal silicon substrate, and a floating silicon diaphragm supported at its edge by the support structure and lying parallel to the perforated backplate and separated from the perforated backplate by an air gap.

Abstract: In the prior art LDMOSFET devices capable of handling high power have been made by locating the source contact on the bottom surface of the device, allowing for good heat sinking, with connection to the source region being made through a sinker. However, this structure has poor high frequency characteristics. Also in the prior art, good high frequency performance has been achieved by introducing a dielectric layer immediately below the source/drain regions (SOI) but this structure has poor power handling capabilities. The present invention achieves both good high frequency behavior as well as good power capability in the same device. Instead of inserting a dielectric layer over the entire cross-section of the device, the dielectric layer is limited to being below the heavily doped section of the drain with a small amount of overlap into the lightly doped section. The structure is described in detail together with a process for manufacturing it.

Abstract: An accelerometer design is described. It operates by measuring a change in capacitance when one plate is fixed and one is mobile (free to accelerate). Unlike prior art designs where such changes are caused by variations in the plate separation distance, in the design of the present invention the plate separation distance is fixed, it being the effective plate area that changes with acceleration. A key feature is that the basic unit is a pair of capacitors. The fixed plates in each case are at the same relative height but the mobile plates are offset relative to the fixed plates, one mobile plate somewhat higher than its fixed plate with the other mobile plate being somewhat lower. Then, when the mobile plates move (in the same direction), one capacitor increases in value while the other decreases by the same amount. This differential design renders the device insensitive to sources of systematic error such as temperature changes. A process for manufacturing the design is described.

Abstract: A novel silicon RF LDMOSFET structure based on the use of a stacked LDD, is disclosed. The LDD has been modified from a single layer of N type material to a stack of three layers. These are upper and lower N type layers with a P type layer between them. The upper N type layer is heavily doped to reduce the on-resistance of the device, while the lower N type layer is lightly doped to reduce the output capacitance, thereby improving the high frequency performance. The middle P layer is heavily doped which allows it to bring about pinch-off of the two N layers, thereby raising the device's breakdown voltage. A process for manufacturing the device, as well as experimental data concerning its performance are also given.

Abstract: A temperature compensating biasing circuit is constructed by first determining a piecewise function substantially describing a required bias current with respect to temperature. Reference signals are created such that each reference signal describes an amount of contributing currents that, when summed together, generate a master biasing current. The biasing current generator is further constructed to create a thermal signal indicating an operating temperature. Each of the reference signals is compared to the thermal signal. The biasing current generator then identifies which of the contributing currents or portions of the contributing currents are being included to generate the master biasing current. The identified contributing currents and the portions of the contributing currents are then summed to form the master biasing current. The master biasing current may be mirrored to form bias currents that have the temperature compensation bias function.