摘要:
In non-volatile memory devices, a write typically consists of an alternating set of pulse and verify operations. At the end of a pulse, the device must be biased properly for an accurate verify, after which the device is re-biased for the next pulse. In some cases a non-volatile memory is programmed by an alternating set of pulses, but, for at least some pulses without any intervening verify operations. After a one pulse, but before biasing the memory for the next pulse without an intervening very, the source and bit line levels can be left to float.
摘要:
In non-volatile memory devices, a write typically consists of an alternating set of pulse and verify operations. At the end of a pulse, the device must be biased properly for an accurate verify, after which the device is re-biased for the next pulse. The intervals between the pulse and verify phases are considered. For the interval after a pulse, but before establishing the verify conditions, the source, bit line, and, optionally, the well levels can be equalized and then regulated at a desired DC level. After a verify phase, but before applying the biasing the memory for the next pulse, the source and bit line levels can be equalized to a DC level.
摘要:
In non-volatile memory devices, a write typically consists of an alternating set of pulse and verify operations. At the end of a pulse, the device must be biased properly for an accurate verify, after which the device is re-biased for the next pulse. In some cases a non-volatile memory is programmed by an alternating set of pulses, but, for at least some pulses without any intervening verify operations. After a one pulse, but before biasing the memory for the next pulse without an intervening very, the source and bit line levels can be left to float.
摘要:
In non-volatile memory devices, a write typically consists of an alternating set of pulse and verify operations. At the end of a pulse, the device must be biased properly for an accurate verify, after which the device is re-biased for the next pulse. The intervals between the pulse and verify phases are considered. For the interval after a pulse, but before establishing the verify conditions, the source, bit line, and, optionally, the well levels can be equalized and then regulated at a desired DC level. After a verify phase, but before applying the biasing the memory for the next pulse, the source and bit line levels can be equalized to a DC level.
摘要:
In non-volatile memory devices, a write typically consists of an alternating set of pulse and verify operations. At the end of a pulse, the device must be biased properly for an accurate verify, after which the device is re-biased for the next pulse. The intervals between the pulse and verify phases are considered. For the interval after a pulse, but before establishing the verify conditions, the source, bit line, and, optionally, the well levels can be equalized and then regulated at a desired DC level. After a verify phase, but before applying the biasing the memory for the next pulse, the source and bit line levels can be equalized to a DC level.
摘要:
In non-volatile memory devices, a write typically consists of an alternating set of pulse and verify operations. At the end of a pulse, the device must be biased properly for an accurate verify, after which the device is re-biased for the next pulse. The intervals between the pulse and verify phases are considered. For the interval after a pulse, but before establishing the verify conditions, the source, bit line, and, optionally, the well levels can be equalized and then regulated at a desired DC level. After a verify phase, but before applying the biasing the memory for the next pulse, the source and bit line levels can be equalized to a DC level.
摘要:
Sensing circuits for sensing a conduction current of a memory cell among a group of non-volatile memory cells being sensed in parallel and providing the result thereof to a data bus are presented. A precharge circuit is coupled to a node for charging the node to an initial voltage. An intermediate circuit is also coupled to the node and connectable to the memory cell, whereby current from the precharge circuit can be supplied to the memory cell. The circuit also includes a comparator circuit to perform a determination the conduction current by a rate of discharge at the node; a data latch coupled to the comparator circuit to hold the result of said determination; and a transfer gate coupled to the data latch to supply a result latched therein to the data bus independently of the node. This arrangement improves sensing performance and can help to eliminate noise on the analog sensing path during sensing and reduce switching current.
摘要:
Sensing circuits for sensing a conduction current of a memory cell among a group of non-volatile memory cells being sensed in parallel and providing the result thereof to a data bus are presented. A precharge circuit is coupled to a node for charging the node to an initial voltage. An intermediate circuit is also coupled to the node and connectable to the memory cell, whereby current from the precharge circuit can be supplied to the memory cell. The circuit also includes a comparator circuit to perform a determination the conduction current by a rate of discharge at the node; a data latch coupled to the comparator circuit to hold the result of said determination; and a transfer gate coupled to the data latch to supply a result latched therein to the data bus independently of the node. This arrangement improves sensing performance and can help to eliminate noise on the analog sensing path during sensing and reduce switching current.
摘要:
In a sensing method, accuracy of sensing operations, such as read or verify, in a memory device is improved by avoiding fluctuations in a sense amp supply voltage which can occur when different sense amps are strobed at different times. First and second sets of sense amps perform a sensing operation on respective storage elements, such as in an all bit line configuration. The first set of sense amps is strobed at a first time point. In response, a sensed analog level is converted to digital data. The A/D conversion relies on the sense amp supply voltage being accurate. To avoid a fluctuation in the sense amp supply voltage, a bypass path allows the storage elements associated with the first set of sense amps to continue to draw power from the sense amp supply voltage. The second set of sense amps is strobed at a later, second time point.
摘要翻译:在感测方法中,通过避免当在不同时间选通不同的感测放大器时可能发生的感测放大器电源电压的波动,改进了存储器件中的读取或验证等感测操作的精度。 第一和第二组感测放大器对各个存储元件执行感测操作,例如以全位线配置。 第一组感测放大器在第一时间点选。 作为响应,将感测的模拟电平转换为数字数据。 A / D转换依赖于正确的读出放大器电源电压。 为了避免感测放大器电源电压的波动,旁路通路允许与第一组感测放大器相关联的存储元件继续从感测放大器电源电压获取功率。 第二组感测放大器在稍后的第二时间点被选通。
摘要:
A NAND string in which bit line-to-bit line noise is discharged prior to sensing a programming condition of a selected non-volatile storage element in the NAND string. A source voltage is applied which boosts the voltage in conductive NAND strings. The voltage boost results in capacitive coupling of noise to neighboring NAND strings. A current pull down device is used to discharge each NAND string prior to performing sensing. After each NAND string is coupled to a discharge path for a predetermined amount of time, bit lines of the NAND string are coupled to voltage sense components for sensing the programming condition of the selected non-volatile storage elements based on a potential of the bit lines. The selected non-volatile storage elements may have a negative threshold voltage. Further, a word line associated with the selected non-volatile storage elements may be set at ground.