Abstract: A method is provided for manufacturing electronic non-volatile memory devices on a semiconductor substrate including a matrix of memory cells having floating gate regions formed on respective active areas and an oxide layer separating the active areas. The method may include forming sidewalls of the floating gate regions that are slanted with respect to a surface of the semiconductor substrate, forming a trench in the oxide layer following the formation of the floating gate regions, and forming a plug of polycrystalline silicon in the trench. The slanted sidewalls of the floating gate regions provide a lead-in for the formation of upper layers.
Abstract: A method for testing a programmable, non-volatile memory including a matrix of memory cells is provided. A plurality of memory cells are programmed. The programmed memory cells are addressed in succession to identify a lowest of threshold voltage levels. The addressing for each memory location includes applying a selection voltage that is lower than the lowest threshold voltage level corresponding to a memory location currently being addressed. The bits are read from the programmed memory cells for the memory location currently being addressed. The reading is repeated while progressively changing the selection voltage supplied to the word line corresponding to the memory location currently being addressed until it is detected that at least one of the bits of the memory location currently being addressed has switched from a first logic level corresponding to a reading of a programmed memory cell to a second logic level corresponding to a reading of a non-programmed memory cell.
Type:
Application
Filed:
February 14, 2001
Publication date:
August 23, 2001
Applicant:
STMicroelectronics S.r. I.
Inventors:
Alessandro Camera, Paolo Sandri, Ignazio Bellomo, Albino Pidutti
Abstract: A device for detecting the application of a high voltage signal to an internal node of an integrated circuit includes a high-voltage divider circuit and a threshold detection circuit. The threshold detection circuit receives a signal given by the output of the divider circuit, and provides a threshold crossing detection signal at an output thereof based upon the signal crossing a threshold. The detection circuit is connected between the logic supply voltage and ground, and further includes a negative feedback loop. The negative feedback loop is connected to the output of the divider circuit to limit the voltage build-up of the high voltage signal at the output thereof after the crossing of the detection threshold by the signal.
Abstract: A manufacturing process providing a zener diode formed in an N-type well housing a first N-type conductive region and having a doping level higher than the well, and a second P-type conductive region arranged contiguous to the first conductive region. The first conductive region is connected, through a third N-type conductive region having the same doping level as the first conductive region, to a conductive material layer overlying the gate oxide layer to be protected. The third conductive region, the well, and the substrate form an N+/N/P diode that protects the gate oxide layer during manufacture of the integrated device from the deposition of the polycrystalline silicon layer that forms the gate regions of the MOS elements.
Abstract: A method including: forming doped regions on a monocrystalline substrate; growing an epitaxial layer; forming trenches in the epitaxial layer extending to the doped regions; anodizing the doped regions in an electro-galvanic cell to form porous silicon regions; oxidizing the porous silicon regions; removing the oxidized porous silicon regions to form a buried air gap; thermally oxidizing the substrate to grow an oxide region from the walls of the buried air gap and the trenches, until the buried air gap and the trenches themselves are filled.
Type:
Grant
Filed:
May 14, 1999
Date of Patent:
August 21, 2001
Assignee:
STMicroelectronics S.R.L.
Inventors:
Gabriele Barlocchi, Flavio Francesco Villa
Abstract: An amplifier stage having a first and a second transistor connected in series to each other between a first and a second reference potential line. The first transistor has a control terminal connected to an input of the amplifier stage through a first inductor, a first terminal connected to the second reference potential line through a second inductor, and a third terminal connected to a first terminal of the second transistor. The second transistor has a second terminal forming an output of the amplifier stage, and connected to the first reference potential line through a load resistor. To improve the noise figure, a matching capacitor is connected between the control terminal and the first terminal of the first transistor.
Type:
Grant
Filed:
December 21, 1999
Date of Patent:
August 21, 2001
Assignee:
STMicroelectronics S.r.l.
Inventors:
Giuseppe Palmisano, Giuseppe Ferla, Giovanni Girlando
Abstract: An integrated oscillator and associated methods are provided for providing clock signals. The integrated oscillator preferably includes a micro-mechanical oscillating circuit for providing an oscillating clock signal. The micro-mechanical oscillating circuit preferably includes a support layer, a fixed layer positioned on a support layer, remaining portions of a sacrificial layer positioned only on portions of the fixed layer, and an oscillating layer positioned on the remaining portions of the sacrificial layer, overlying the fixed layer in spaced relation therefrom, and extending lengthwise generally transverse to a predetermined direction for defining a released beam for oscillating at a predetermined frequency. The spaced relation is preferably formed by removal of unwanted portions of the sacrificial layer.
Type:
Grant
Filed:
October 5, 1999
Date of Patent:
August 21, 2001
Assignee:
STMicroelectronics, Inc.
Inventors:
Tsiu Chiu Chan, Melvin Joseph DeSilva, Syama Sundar Sunkara
Abstract: A memory counter circuit includes a plurality of mutually connected counter stages, an internal address bus interfaced with each one of the counter stages for sending an external address signal to each one of the counter stages, a circuit for loading the external address signal onto the internal address bus, and an enabling circuit for enabling a connection between the internal bus and each one of the counter stages. The enabling circuit may be driven by a true address latch enable signal. The memory counter circuit may further include a circuit for generating the true address latch enable signal starting from an external address latch signal and a fast address latch enable signal for driving the circuit for loading the external address signal onto the internal address bus. A signal generation circuit may also be included for generating clock signals for synchronizing each one of the counter stages. The synchronization signals are preferably not simultaneously active.
Abstract: In a EEPROM memory architecture organized into word columns that includes n memory cells per word column, there is, for each word of the column, one diffusion line to connect sources of the memory cells to a ground connection transistor using a source line. A word read access includes simultaneously selecting the word accessed in a read mode in a first group of memory cells, and an additional word in the second group of memory cells. Each column has n bit lines ranked 0 to n−1, each connected to the same ranked cells in the first group of memory cells.
Abstract: A method of forming source and drain regions for LV transistors that includes the steps of forming sacrificial spacers laterally to LV gate regions; forming LV source and drain regions in a self-aligned manner with the sacrificial spacers; removing the sacrificial spacers; forming HV gate regions of HV transistors; forming gate regions of selection transistors; forming control gate regions of memory transistors; simultaneously forming LDD regions self-aligned with the LV gate regions, HV source and drain regions self-aligned with the HV gate regions, source and drain regions self-aligned with the selection gate region and floating gate region; depositing a dielectric layer; covering the HV and memory areas with a protection silicide mask; anisotropically etching the dielectric layer, to form permanent spacers laterally to the LV gate regions; removing the protection silicide mask; and forming silicide regions on the LV source and drain regions and on the LV gate regions.
Type:
Grant
Filed:
December 21, 1999
Date of Patent:
August 14, 2001
Assignee:
STMicroelectronics S.r.l.
Inventors:
Matteo Patelmo, Bruno Vajana, Giovanna Dalla Libera, Carlo Cremonesi, Nadia Galbiati
Abstract: A method and device decode a compressed image, and in particular, an image compressed according to the MPEG standards, especially a bidirectional image. To perform two successive decodings of a bidirectional image, the address of the data packet containing the start-of-image identifier of the bidirectional image is tagged, and the temporal reference of this image is stored. After the first decoding, the stored address of the memory is again pointed to and a second decoding is performed after a new detection of the temporal reference of the image.
Abstract: A pulse generator circuit for non-volatile memories, is disclosed, including a circuit for determining the instant at which a pulse for incrementing a counter of the memory is generated and generating an increment pulse duration start signal; a circuit for determining the minimum amplitude of the increment pulse and generating an increment pulse duration end signal; a first logic circuit for enabling the generation of the increment pulse based upon the increment pulse duration start and end signals; and an increment pulse generation circuit for generating or suppressing the increment pulse of the counter of the memory, based upon the current condition of the memory.
Abstract: An integrated electronic device having a first charge pump, intended to drive a first line having a high capacitive load, and a second charge pump having a high current pumping capacity and intended to drive a second line, a controlled switch is interposed between the outputs of the two pumps, such as to connect the output of the high current capacity pump to the first line, to charge the first line quickly to the preset voltage, without the first charge pump being oversized. When the voltage present on the first line becomes greater than the voltage at the output of the second charge pump, owing to the current required by the second line, the switch is opened. A common phase generator which drives both the pumps is also provided.
Abstract: A pair of equivalent controlled impedance buffers are connected in a push-pull configuration to a transformer primary coil. A pair of equivalent pre-drivers are connected to the pair of buffers. Each pre-driver receives a driver input signal and outputs a buffer input signal and a proportional flyback compensation signal. Each buffer receives the buffer input signal generated from one of the pre-drivers for buffered output as a line driver signal to the primary coil which induces a flyback voltage effect in each buffer. Each buffer further receives the flyback compensation signal generated from the other one of the pre-drivers, with the buffer operating to cancel the flyback voltage effect induced in that buffer using the flyback compensation signal received from the other one of the pre-drivers. An adjustment circuit further outputs an adjustment signal for application to an adjustable current source.
Abstract: A Pfield operation defined according to the Montgomery method by Pfield(A, B)N=A*B*I mod N, where I is a determinable error, is implemented in a processor. The least significant word of the data elements A and N which are stored in elementary sub-registers are shifted twice. This eliminates delay cells in a processor used for executing the Pfield operation.
Abstract: A PET decoder for an ATM network has a modular architecture including a processing unit having various memories and a processing pipeline for constructing from a block of m data of a certain number of bits, a square matrix A based on a vector D of relative points over the Galois field. The processing pipeline also decomposes by triangular factorization the square matrix A and solves the subsystem of equations by simple substitution. The decoder also includes a control unit interfacing with the ATM network, a programmable parallel processor, a random access memory and the processing unit.
Type:
Grant
Filed:
September 2, 1998
Date of Patent:
August 14, 2001
Assignee:
STMicroelectronics S.r.l.
Inventors:
Sergio Mazzaglia, Francesco Italia, Mario Lavorgna
Abstract: A method is for self-test and correction of errors due to a loss charge for a flash memory including an array or matrix of cells (bits), organized in rows and columns, erasable and programmable by whole sectors in which the matrix is divided.
Type:
Grant
Filed:
December 16, 1998
Date of Patent:
August 14, 2001
Assignee:
STMicroelectronics S.r.l.
Inventors:
Paolo Cappelletti, Alfonso Maurelli, Marco Olivo
Abstract: The protection device for an interconnection line of an integrated circuit includes a charge flow-off device connected between the interconnection line to be protected and the substrate of the integrated circuit. The protection device also includes a dummy interconnection line ANT to activate the flow-off device. The protection device is active throughout the manufacture of the integrated circuit.
Abstract: A device for the regeneration of a clock signal uses a reference clock signal given by an internal oscillator to measure the number of reference clock pulses between the first two synchronization pulses sent by an external serial bus or USB at the beginning of each transaction. Thus a rough measurement N is obtained of the USB clock signal to be regenerated. The delay of each of these two synchronization pulses with respect to the previous pulse of the reference clock signal is measured. This delay is computed with respect to an internally defined time unit. On the basis of the measurement of these two delays, and the measurement of the number of reference clock periods, and knowing the measurement n of the period of the reference clock signal in the time unit, the period of the USB clock signal to be regenerated is computed with precision.
Abstract: A digital image color correction device and method employing fuzzy logic, for correcting a facial tone image portion of a digital video image is provided.