Abstract:
Non-volatile memory device organised with memory cells that are arranged by row and by column, comprising at least a sector of matrix cells (100), row decoders (D) and column decoders suitable to decode address signals and to activate respectively said rows or said columns, at least a sector of redundancy cells (110) such that it is possible to substitute a row of said sector of matrix cells with a row of said sector of redundancy cells. Said non-volatile memory device comprises a local column decoder (L) for said matrix sector (100) and a local column decoder (L) for said redundancy sector (110). The local column decoders (L) are controlled by external signals so that said row of said redundancy sector (110) is activated simultaneously with said row of said matrix sector (100).
Abstract:
The multilevel memory (50) stores words formed by a plurality of binary subwords in a plurality of cells (63a-63p), each cell having has a respective threshold value. The cells are arranged on cell rows and columns, are grouped into sectors (56) divided into sector blocks (57), and are selected via a global row decoder (51), a global column decoder (54), and a plurality of local row decoders (58), which simultaneously supply a ramp voltage (V R ) to a biasing terminal of the selected cells. Threshold reading comparators (72a, 72b) are connected to the selected cells, and generate threshold attainment signals when the ramp voltage reaches the threshold value of the selected cells; switches (65a-65d), are arranged between the global word lines (52) and local word lines (59a-59d, 60a-60d), opening of the switches is individually controlled by the threshold attainment signals, thereby the local word lines are maintained at a the threshold voltage of the respective selected cell, after opening of the switches.
Abstract:
Non-volatile semiconductor memory device comprising an address buffer block (10), a matrix of memory cells (11) and an output buffer block (12), said address buffer block (10) receiving input signals (8) external to the memory device, that in a first operating mode are controlled by devices outside to the memory device, and transmitting signals to said matrix of memory cells (11), adapted to decode said received signals and to transmit in turn output decoded signals through said output buffer block (12). A command block (13;17,18,19) is provided with, activatable through an external control signal (16) that, once activated, puts the memory device in a second operating mode in which said command block (13;17,18,19) receives at least a part of said signals in output of said matrix of memory cells (11) and, after having processed them, transmits internal address signals (15) to said address buffer block (10), so to have a feedback inside said memory device capable of making the same able to execute a succession of instructions memorized in said matrix of memory cells (11) autonomously.
Abstract:
The invention relates to an ESD protection network for a CMOS circuit structure integrated in a semiconductor substrate (2) and comprising discrete circuit blocks formed in respective substrate portions which are electrically isolated from one another and independently powered from at least one primary voltage supply (Vcc) having a respective primary ground (GND), and from at least one secondary voltage supply (Vcc_IO) having a respective secondary ground (GND_IO). This network comprises essentially:
a first ESD protection element (15) for an input stage of the circuit structure; a second ESD protection element (5) for an output stage of the circuit structure, the first (15) and second (5) protection elements having an input/output terminal (20) of the integrated circuit structure in common; at least one ESD protection element (B0) between the primary supply (Vcc) and the primary ground (GND); at least one ESD protection element (B) between the secondary supply (Vcc_IO) and the secondary ground (GND_IO).
Abstract:
The invention relates to a method and a circuit for generating a pulse synchronization signal (ATD) for timing the memory cell read phase in semiconductor integrated electronic memory devices. The pulse signal (ATD) is generated upon detection of a change in logic state of at least one of a plurality of address input terminals of the memory cells. The method consists of duplicating the ATD signal into at least one pair of signals (ATD1,ATD2) and propagating such signals through separate parallel timing chains (6,9) at the ends of which the ATD signal is reinstated, the chains (6,9) being alternately active.
Abstract:
A charge pump circuit comprises at least one pump stage. Said pump stage includes a capacitor (C1,C2,C3;C'1,C'2,C'3) having a first plate and a second plate. The pump stage further includes a first circuital node (N1,N3,N5;N'2,N'4,N'6) connected to the first plate, a voltage of the first circuital node is forced to a first forced voltage (GND) during a forcing phase of the charge pump operation, and a second circuital node (N2,N4,N6;N'1,N'3,N'5) connected to the second plate, a voltage of the second circuital node is forced to a second forced voltage (Vdd) during the forcing phase. The voltages of the first and the second circuital nodes are free of changing with respect to the first and to the second forced voltage, respectively, except during said forcing phase. The pump stage still further includes a first forcing circuit (110,130,150;420,440,460) associated to the first circuital node, the first forcing circuit being activable for forcing the voltage of the first circuital node to the first forced voltage during the forcing phase, and a second forcing circuit (120,140,160;410,430,450) associated to the second circuital node, the second forcing circuit being activable for forcing the voltage of the second circuital node to the second forced voltage during the forcing phase.
Abstract:
The invention relates to a nonvolatile electronic memory device (1) being monolithically integrated on a semiconductor substrate and of the Flash EEPROM type having a NAND architecture and comprising at least one memory matrix (11) organised in rows and columns of memory cells. Advantageously, the matrix comprises at least one portion (3) having smaller data storage capacity and faster access speed than the other portion (2).