摘要:
A Bi.CMOS semiconductor memory device is provided which includes an arrangement to simultaneously select a plurality of memory cells, followed by using a 3 bit Z addressing arrangement to determine a read or write operation for the simultaneously selected memory cells. To speed up the word line selection, a static selection type operation is used with the word line selecting voltage being greater than signal amplitude of the data lines during the write operation. Also, to speed up the read operation, separate common I/O lines are provided for the read and write operations. Read signals are transmitted as current signals, and then converted to voltage signals for improving reading speed. Also, improved arrangements are provided for resistance structure, logic circuitry, input circuitry, fuse cutting circuitry, drive circuitry, power circuitry, electrostatic protection circuitry, layout structure and testing methods for the semiconductor device.
摘要:
A Bi.CMOS semiconductor memory device is provided which includes an arrangement to simultaneously select a plurality of memory cells, followed by using a 3 bit Z addressing arrangement to determine a read or write operation for the simultaneously selected memory cells. To speed up the word line selection, a static selection type operation is used with the word line selecting voltage being greater than signal amplitude of the data lines during the write operation. Also, to speed up the read operation, separate common I/O lines are provided for the read and write operations. Read signals are transmitted as current signals, and then converted to voltage signals for improving reading speed. Also, improved arrangements are provided for resistance structure, logic circuitry, input circuitry, fuse cutting circuitry, drive circuitry, power circuitry, electrostatic protection circuitry, layout structure and testing methods for the semiconductor device.
摘要:
A Bi.CMOS semiconductor memory device is provided which includes an arrangement to simultaneously select a plurality of memory cells, followed by using a 3 bit Z addressing arrangement to determine a read or write operation for the simultaneously selected memory cells. To speed up the word line selection, a static selection type operation is used with the word line selecting voltage being greater than signal amplitude of the data lines during the write operation. Also, to speed up the read operation, separate common I/O lines are provided for the read and write operations. Read signals are transmitted as curent signals, and then converted to voltage signals for improving reading speed. Also, improved arrangements are provided for resistance structure, logic circuitry, input circuitry, fuse cutting circuitry, drive circuitry, power circuitry, electrostatic protection circuitry, layout structure and testing methods for the semiconductor device.
摘要:
In order to eliminate erroneous reading of data by preventing noise which might otherwise be transmitted at the data reading time through parasitic capacitance in the data lines to other data lines, switches (Qt1 and Qt1′) are interposed between a sense amplifier (SA) for amplifying the potential of a data line (DL) and the data line, and the sense amplifier is fed with an operating voltage after the potential of the data line is transmitted to the sense amplifier, and the switch is turned off.
摘要:
In order to eliminate erroneous reading of data by preventing noise which might otherwise be transmitted at the data read time through parasitic capacitance in the data lines to other data lines, switches (Qt1 and Qt1′) are interposed between a sense amplifier (SA) for amplifying the potential of a data line (DL) and the data line, and the sense amplifier is fed with an operating voltage after the potential of the data line is transmitted to the sense amplifier, and the switch is turned off.
摘要:
There is provided a method of programming a non-volatile memory which can solve the problem of the data write system of the existing flash memory that a load capacitance of bit lines becomes large, the time required by the bit lines to reach the predetermined potential becomes longer, thereby the time required for data write operation becomes longer and power consumption also becomes large because the more the memory capacitance of memory array increases, the longer the length of bit lines becomes and the more the number of bit lines increases. In the non-volatile memory of the invention comprising the AND type memory array in which a plurality of memory cells are connected in parallel between the local bit lines and local drain lines, the local drain lines are precharged by supplying thereto a comparatively higher voltage from the common drain line side (opposite side of the main bit lines), the main bit lines are selectively precharged by applying thereto the voltage of 0V or a comparatively small voltage depending on the write data and thereafter a drain current is applied only to the selected memory cells to which data is written by applying the write voltage to the word lines in order to implant the hot electrons to the floating gate.
摘要:
Data are generated based on additional write data input to data latch circuits (DLR and DLL) and data read from memory cells (MC) to program non-volatile memory cells in a write state into the same write state and to program non-volatile memory cells in an erase state into a write state indicated by the additional write data. The generated data are latched in the data latch circuits to perform a logical synthesis process for additional writing. Even after the additional write operation, the logically synthesized data remain in the data latch circuits, and the latched data can be reused against abnormality in writing. This eliminates the need for receiving write data again from the outside when the additional write operation is to be retried.
摘要:
A control of a flash memory includes control for supplying a pulse-shaped voltage to each of non-volatile memory cells until a threshold voltage of the non-volatile memory cell having a first threshold voltage is changed to a second threshold voltage. The control involves a first write mode (coarse write) in which the amount of change in threshold voltage of each non-volatile memory cell, which is varied each time the pulse-shaped voltage is applied, is relatively rendered high, and a second write mode (high-accuracy write) in which the amount of change in threshold voltage thereof is relatively rendered low. As compared with the high-accuracy mode, the number of pulses required to change the threshold voltage of each memory cell is smaller than that in the coarse write mode. Therefore, the number of verify operations at the time that the coarse write mode is used, is small and hence the entire write operation can be speeded up.
摘要:
Externally supplied program data is latched into data latch circuits DLL and DLR. A judgment is made as to whether or not the latched program data corresponds to any threshold value of multi-levels every time each of plural programing operations is carried out. The program control information corresponding to the judgment result is latched into a sense latch circuit SL. Based upon the latched program control information, the programing operation for setting threshold voltages having multi-levels to a memory cell is carried out in a stepwise manner. Even when the programing operation is ended, the externally supplied program data is left in the data latch circuit. Even when the programing operation of the memory cell is retried due to the overprograming condition, the program data is no longer required to be again received from the external device.
摘要:
Data are generated based on additional write data input to data latch circuits (DLR and DLL) and data read from memory cells (MC) to program non-volatile memory cells in a write state into the same write state and to program non-volatile memory cells in an erase state into a write state indicated by the additional write data. The generated data are latched in the data latch circuits to perform a logical synthesis process for additional writing. Even after the additional write operation, the logically synthesized data remain in the data latch circuits, and the latched data can be reused against abnormality in writing. This eliminates the need for receiving write data again from the outside when the additional write operation is to be retried.