Other

Standard Microsystems Corporation

Hughes Aircraft Company

Integrated Technology Corporation

Zytrex Corporation

Frank M Wanlass

An alpha particle striking the cell of a DRAM bit can destroy stored charge, resulting in a single bit soft error. A DRAM architecture is described that circumvents this problem by storing every DRAM bit redundantly in two cells. If a stored charge is represented by a logic 1, then when reading a DRAM bit, if either of it's cells is storing charge then the bit is a logic 1. Only if both cells of a bit have no charge is the bit a logic 0.

Soft error immune dynamic random access memory

This invention is a processing method for forming flash memory MOS transistors. The method uses chemical mechanical polishing to self align both a floating gate and an overlying control gate to the MOS channel region in both the width and length directions, thereby improving layout density. The method enables the capacitance between the control gate and the floating gate to be much larger than the capacitance between the floating gate and the channel; this reduces programing voltages. The method does not require any Shallow Trench Isolation (STI), and does not require Local Oxidation of Silicon (LOCOS), thereby resulting in little damage to the silicon.

Method of making damascene flash memory transistor

This invention is a damascene processing method for forming ultra short channel MOS transistors, where the channel length is not determined by photolithography. The method uses chemical mechanical polishing to self align an MOS transistor gate electrode to the MOS channel region in both the width and length directions. The method enables metal interconnect lines to make borderless connections to the MOS gate electrodes directly over channel regions, and allows borderless connections to be made to the MOS source and drain regions, thereby improving layout density of small transistors. The method uses metal for first level interconnect lines rather than polysilicon. The method enables the interconnect lines to be patterned on a planar surface, which facilitates the etching of very narrow and closely spaced lines. The method does not require any Shallow Trench Isolation (STI), and does not require Local Oxidation of Silicon (LOCOS), thereby resulting in little damage to the silicon substrate. The method also prevents plasma damage of very thin gate dielectrics during processing.

Ultra short channel damascene MOS transistors

This invention is a damascene processing method for forming ultra short channel MOS transistors, where the channel length is not determined by photolithograpy. The method uses chemical mechanical polishing to self align an MOS transistor gate electrode to the MOS channel region in both the width and length directions. The method enables metal interconnect lines to make borderless connections to the MOS gate electrodes directly over channel regions, and allows borderless connections to be made to the MOS source and drain regions, thereby improving layout density of small transistors. The method uses metal for first level interconnect lines rather than polysilicon. The method enables the interconnect lines to be patterned on a planar surface, which facilitates the etching of very narrow and closely spaced lines. The method does not require any Shallow Trench Isolation (STI), and does not require Local Oxidation of Silicon (LOCOS), thereby resulting in little damage to the silicon substrate. The method also prevents plasma damage of very thin gate dielectrics during processing.

Method of making damascene completely self aligned ultra short channel

This invention is a processing method for forming MOS transistors. The method uses chemical mechanical polishing to self align an MOS transistor gate electrode to the channel region in both the width and length directions. The method enables metal interconnect lines to make borderless connections to MOS gate electrodes directly over channel regions, and allows borderless connections to be made to source and drain regions, thereby improving layout density of small transistors. The method enables interconnect lines to be patterned on a planar surface, which facilitates the etching of very narrow and closely spaced lines. The method does not require any Shallow Trench Isolation (STI), and does not require Local Oxidation of Silicon (LOCOS), thereby resulting in little damage to the silicon. The method also prevents plasma damage of very thin gate dielectrics during processing.

Damascene formation of borderless contact MOS transistors

This invention is a processing method for electrically isolating CMOS transistors. The method involves implanting a channel stop dopant into field regions between transistor active regions, self aligning relatively thick silicon dioxide over these field regions and providing thin oxide in the active regions that are self aligned to the field regions. The method does not require any shallow trench isolation (STI), and does not require Local Oxidation of Silicon (LOCOS), thereby resulting in little damage to the silicon.

Damascene isolation of CMOS transistors

The present invention is a technique for producing silicon-on-insulator MOS transistors by damascene patterning of source-drain regions in a thin film of amorphous silicon deposited on a layer of oxide grown on a silicon wafer, where the oxide has previously been etched with a pattern of trenches. In addition, the technique provides for the amorphous layer to contact the underlying silicon substrate through multiple small oxide openings; where these openings have been previously filled with amorphous silicon, planarized and annealed at high temperature to form single crystal silicon; and where subsequent transistor channel regions will align to these filled openings. After patterning, the wafer is annealed in a second high temperature cycle, where the regions of amorphous silicon in contact with the single crystal silicon in the openings will convert into single crystal silicon suitable for transistor channel regions.

Double damascene pattering of silcon-on-insulator transistors

The present invention is a technique for producing planar silicon on insulator MOS transistors, where the channel regions are created in an underlying single crystal silicon wafer, and where the source-drain extension regions are created by damascene patterning a thin film of amorphous silicon deposited on a layer of oxide deposited on the silicon wafer.

Damascene pattering of SOI MOS transistors

The present invention is a technique for producing silicon-on-insulator MOS transistors by damascene patterning of source-drain regions in a thin film of amorphous silicon deposited on a layer of oxide grown on a silicon wafer, where the oxide has previously been etched with a pattern of trenches. In addition, the technique provides for the amorphous layer to contact the underlying silicon substrate through multiple small oxide openings, where subsequent transistor channel regions will align to these openings. After patterning, the wafer is annealed in a high temperature cycle, where the regions of amorphous silicon in contact with the silicon substrate will grow into single crystal silicon suitable for transistor channel regions.

Damascene method for source drain definition of silicon on insulator MOS

A back biasing technique is provided for increasing the field inversion voltage between adjacent MOS transistors and for reducing parasitic capacitances in an integrated circuit. The use of a charge pump is avoided by connecting the body portions of the MOS transistors to ground and the sources of the MOS transistors to the anode of a diode, the cathode of which are connected to a reference voltage such as to ground. In this manner, the sources are back biased relative to the material in which they are formed by a diode forward voltage drop. This technique is particularly applicable to CMOS circuits operating from a 3.3 volt supply, with p-well doping densities in excess of 1.times.10.sup.17 atoms/cm.sup.3.

Electrical isolation in integrated circuits

A CMOS device having many output channels at least one of which channel includes a first pair of pull-up/pull-down of transistors between a 'noisy Vcc and noisy ground', a second pair of pull-up/pull-down of transistors between a 'quiet Vcc and quiet ground', and logic to switch the transistor pairs such that initial switching of an output is powered by the noisy Vcc and ground, and maintenance of an output state is powered by the quiet Vcc and ground. Quiescent channels are decoupled from active channels and will hold their assigned output levels.

Apparatus and method to prevent the disturbance of a quiescent output

A CMOS integrated circuit (IC) device embodiment of the present invention comprises an internal logic circuit operating with traditional 3.3 volt or five volt internal logic levels, an output buffer to convert the internal logic levels to external logic levels of 0.3 volts and an input buffer to convert the 0.3 volt external logic levels to the internal logic levels. In a CMOS IC device having numerous external output loads including relatively high capacitive values that are driven at very high clock rates, the restricted voltage swings of the 0.3 volt external logic levels permit unusually large numbers of devices to be driven without exceeding a predetermined power dissipation limit of the CMOS IC device. The low external logic levels further permit electrostatic discharge (ESD) protection to be included on all signal inputs and outputs of the CMOS IC device. The ESD protection comprises a pair of opposite polarity silicon PN junction diodes in parallel and connected between each signal line and a ground reference.

Very low voltage inter-chip CMOS logic signaling for large numbers of

A lower power, noise rejecting TTL-to-CMOS input buffer, without the use of a current consuming voltage reference, has the characteristic of recognizing a logic LOW as less than 0.8 volts and a logic HIGH as greater than 2.0 volts for DC TTL signals while drawing only leakage current from its Vcc power supply, and simultaneously possesses the characteristic of rejecting high-amplitude Vin noise. For an input signal rapidly rising from zero to three volts, the buffer output switches at an input signal level of approximately 2.5 volts; and for the input signal rapidly falling from 3 to zero volts, the buffer output switches at an input signal level of approximately 1.4 volts.

Low power noise rejecting TTL to CMOS input buffer

A method of controlling the ill-effects of ground bounce in a CMOS device, according to the present invention, comprises increasing the impedance between (1) the output line of a quiescent channel that is already at a low state, and (2) the local ground within the CMOS device; the increased impedance occurring when a ground bounce condition caused by an adjacent channel within the CMOS device would otherwise cause the output of the quiescent channel to be dragged high.

Apparatus and method to prevent the unsettling of a quiescent, low

A TTL to CMOS buffer circuit includes a relatively high-speed first inverter path and a second relatively low-speed inverter path connected to the first path and effective to control the operation of the first path inverter so that short-duration positive- or negative-going noise pulses of amplitude up to 2.4 volts do not incorrectly affect the output level.

Growing community of inventors

Sunnyvale, CA, United States of America

Frank M Wanlass