Abstract:
An autobaud dectection mechanism, installed as part of the communication control software of a serial communication device's microcontroller, enables the device to automatically determine the baud rate employed by a remote digital data communications device. Starting with a default baud rate, corresponding to the highest available baud rate, the mechanism steps through successively lower baud rates in the course of a search for the baud rate at which the remote device is transmitting. When an incoming call is received, respective bits of the received data are examined for the presence of transmission errors. If a transmission error is detected, the baud rate is stepped to the next lowest baud rate. If no error is detected, the received data bits are compared with a reference character. If the two compared data bit patterns match, the controller locks the baud rate at that baud setting for the remainder of the call. When the call is terminated, the baud rate reverts to its default setting. If the bit pattern comparison does not indicates a match, the baud rate is stepped to the next lowest baud rate. When the autobaud detection routine has stepped to a new baud rate entry, it initiates a prescribed time-out as a safeguard against the receiver becoming "out-of-sync" with a transmitter. Once, the lowest baud rate has been reached, the routine steps to the default setting and starts again.
Abstract:
A semiconductor chip having a cellular topography and a method of packaging a cellular semiconductor chip includes plural interdigitated metal gate runners that overlie and contact selected gate electrodes on the chip surface, each of the gate runners having an integral widened area to enable a package carried gate electrode contact foil to be bonded thereto. The gate runner widened areas are relatively small and have little impact on chip active area. The plural gate runners have portions that underlie a package-carried power electrode contact foil and that are separated therefrom by a nonbondable, insulating layer. The gate runners may be deposited on the chip in the same step and from the same material as the power electrode. The portion of the power electrode on the chip surface that underlies the package-carried gate electrode contact foil is separated therefrom and available for use as active area of the chip. Package lid-to-chip alignment tolerances may be relaxed as they are not dictated by alignment of the lid-carried gate contact foil with the gate electrode on the chip.
Abstract:
A semiconductor-on-insulator structure incorporating a layer of diamond material and method for preparing such. The structure comprises a layer containing diamond material and having a first surface. A layer of silicon nitride is formed on the first surface and a layer of semiconductor material is positioned over the silicon nitride layer. In one embodiment of the method there is provided a removable deposition surface. A layer of crystalline diamond material is formed on the deposition surface. A first surface of the diamond material is separated from the deposition surface. The structure is useful for formation of integrated circutis thereon.
Abstract:
The gain of a parasitic lateral bipolar device in a MOS SOI structure is substantially reduced so as to increase the differential between the snap-back sustaining voltage and the maximum recommended power supply voltage. Prior to insulated gate structure definition, very lightly doped source and drain regions are implanted so that they extend down to the underlying insulator layer. By very lightly doped is meant that the source and drain regions have a doping concentration that is within an order of magnitude of the doping concentration of the well portion of the semiconductor layer. Because the very lightly doped and physically deep source and drain regions are implanted in the absence of gate structure the implanted energy can be increased above the level that is normally employed to bottom out heavily doped regions, so that the reduced doping density will extend completely through the well layer to the underlying oxide. After the very lightly doped regions have been implanted, the implant mask is stripped and an insulated gate structure is formed atop the channel surface portion of the well layer between the source and drain regions. Using the insulated gate structure as a mask, off-axis, high angle implants of the same conductivity type as the source and drain regions are carried out to a first depth that only partially penetrates the depth of the deep source and drain implants. Very shallow high impurity concentration ohmic contact regions are then formed in surface portions of the first and second regions, and ohmic contact layers are formed on the conductive gate layer and the high impurity concentration ohmic contact regions.
Abstract:
Warpage in a bonded wafer is limited by maintenance of a stress compensation layer on the backside of the bonded wafer during device fabrication processing. One embodiment applies a sacrificial polysilicon layer over a stress compensation silicon dioxide layer for bonded silicon wafers. The fabrication processing consumes the polysilicon layer but not the stress compensation silicon dioxide.
Abstract:
An EPROM cell comprises an MOS device including a floating gate electrode (44, 64) overlying, and ohmically insulated from, the channel region (26) of the MOS device, and a separate diode (18) including a p-n junction (54) having a substrate surface intercept (30). A floating gate electrode (64) overlies the diode p-n junction intercept (54) and is ohmically isolated therefrom by an intervening insulating layer. Writing of data into the floating gate electrode of the MOS device is achieved by causing a voltage breakdown across the diode p-n junction (54) and the flow of high energy electrons across the junction. A voltage is simultaneously applied to the diode gate electrode (64) thereby attracting some of the high energy electrons through the overlying insulating layer into the diode floating gate electrode. The diode gate electrode (64) is ohmically connected to the MOS floating gate electrode (44) on which some of the electrons are stored for affecting the turn-on, turn-off, characteristics of the MOS device.
Abstract:
A subscriber line circuit for a telephone network having current controlled switches and current sources. Methods of making and using the components are also disclosed.
Abstract:
A semiconductor protection circuit comprises a semiconductor substrate (2) of a first conductivity type; a first region (6) of the first conductivity type formed in the substrate at a surface thereof and having a relatively different degree of conductivity from the substrate; a region of a second conductivity type (8) formed in the first region (6) of the first conductivity type; and a second region (10) of the first conductivity type formed partly in each of the semiconductor substrate (2) and the first region (6) of the first conductivity type so as to bridge a junction therebetween.
Abstract:
Impulse noise suppression upstream of digital processing circuitry contains a sample and hold (S/H) mechanism (24) which samples the input signal and stores a plurality of sequential samples representative of the amplitude of the input signal at successive times. The contents of the S/H are compared with an input signal sample (22) to determine whether there are abnormal amplitude variations which potentially constitute impulse noise. In one embodiment the comparison is referenced to the average magnitude (fig. 6) of the input signal. In another embodiment a cascaded arrangement (fig. 5) of S/H circuits (54-1-N) sample and store a plurality of sequential samples values. The contents of the last S/H circuit (54-N) are compared with the contents of each of selected other S/H circuits of the cascaded chain. A potential noise impulse sample is prevented from being coupled to downstream processing circuitry. Otherwise it is coupled through a downstream lowpass filter for subsequent signal analysis.