Self-Check Calibration of Memory Cell Programming Using Distribution Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Memory devices, such as NAND flash, experience changes in responsiveness over time due to aging and usage, making it difficult to select optimal bias signals for programming, which affects write and erase performance.
Innovation Solution
Modifying program parameters like voltage amplitude, pulse width, and step size based on changes in threshold voltage distributions of memory cells, allowing for dynamic adjustment during the programming process to maintain optimal conditions as the device ages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fixed bias signals are used for programming memory cells, then the programming process is simple to implement, but the programming reliability deteriorates over time due to aging and usage changes
Solution Approach 1:
The patent applies dynamics by transitioning from fixed bias signals to dynamically adjustable bias signals. The system continuously monitors the distribution of read parameters (threshold voltages) of memory cells and adjusts programming bias signals in real-time based on observed changes in the memory cell population distribution. This dynamic adaptation allows the programming process to respond to aging and usage-induced changes, maintaining programming reliability throughout the device lifespan.
Solution Approach 2:
The patent implements feedback by measuring the distribution of read parameters (threshold voltages) of memory cells during operation and using this information to adjust subsequent programming bias signals. The system creates a closed-loop control where the output (programming results) is monitored and fed back to modify the input (bias signals), ensuring that programming operations remain effective despite device aging and parameter drift over time.
2Productivity
If traditional programming methods are used, then the device structure remains simple, but the programming speed decreases over time as memory cells age
Solution Approach 1:
The system dynamically adjusts programming parameters including voltage amplitude, pulse width, and step size based on real-time monitoring of memory cell distribution characteristics. This dynamic parameter adjustment optimizes programming speed at different stages of device operation, preventing performance degradation that would otherwise occur as memory cells age and their electrical characteristics change.
3Adaptability or versatility
If bias signals are not adjusted, then the programming process is straightforward, but the adaptability to aging memory cells is poor
Solution Approach 1:
The system employs feedback mechanisms to automatically detect changes in memory cell population distribution and adjust programming bias signals accordingly. By continuously monitoring threshold voltage distributions and other read parameters, the system adapts to aging effects without requiring manual intervention or complex external calibration equipment, thereby improving adaptability while keeping the adjustment process automated and manageable.
Solution Approach 2:
The memory device performs self-calibration by monitoring its own parameter distributions and automatically adjusting its programming bias signals. This self-service capability allows the device to adapt to its own aging and degradation without requiring external intervention, maintaining programming effectiveness throughout its operational life while avoiding the complexity of external calibration systems.
Data Source
AI summary
Apparatus and methods determine a program verify (PV) induced reading parameter distribution. A measured post-PV reading parameter distribution can be compared with an expected post-PV reading parameter distribution. For example, de-convolution can be applied to identify the PV induced reading parameter distribution. Based on the PV-induced reading parameter distribution, adjustments can be made to one or more parameters of the PV process.