Variable Reference Resistor Calibration for Resistive Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Resistive memory devices face operational reliability issues due to variations in process, voltage, and temperature, which affect the accuracy of data reading from variable resistance elements.
Innovation Solution
A memory device with a variable reference resistor and a calibration method that adjusts resistance based on temperature and process variations using calibration codes, allowing for arithmetic operations to generate an adjustment code for compensating these variations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a fixed reference resistor is used for reading data from memory cells, then the device structure is simple, but the data reading accuracy deteriorates due to process, voltage, and temperature variations
Solution Approach 1:
The reference resistor is transformed from a fixed value to a variable value that can be dynamically adjusted based on calibration codes. The reference resistor includes multiple resistive elements that can be selectively connected to ground through switches, allowing the reference resistance to vary according to process, voltage, and temperature conditions, thereby maintaining accurate data reading across different operating conditions
Solution Approach 2:
The reference resistance parameter is made changeable through calibration codes stored in non-volatile memory. The calibration codes represent different resistance values that compensate for PVT variations, allowing the reference resistor to adapt its resistance parameter to maintain measurement accuracy without requiring a completely complex restructuring of the device
2Reliability
If calibration circuits are added to compensate for variations, then the operational reliability is improved, but the device area and power consumption increase
Solution Approach 1:
The calibration functionality is merged with the reference resistor structure itself. The reference resistor includes multiple resistive elements and switches that are integrated into the same circuit block, eliminating the need for separate calibration circuits. The calibration codes are stored in on-chip non-volatile memory, further reducing external components. This merging approach improves reliability through variation compensation while minimizing additional area overhead
Solution Approach 2:
The reference resistor structure serves multiple functions: it provides the reference resistance for data reading and simultaneously acts as a variable element that can be adjusted through calibration codes to compensate for variations. The switches and resistive elements are designed to serve both the normal read operation and the calibration adjustment function, reducing the need for dedicated calibration-only components
3Measurement precision
If multiple calibration codes are stored and processed, then the compensation accuracy is improved, but the processing time and complexity increase
Solution Approach 1:
The calibration codes are pre-calculated and stored in non-volatile memory before the device operates. The calibration process involves reading these pre-stored codes and selecting the appropriate resistance configuration, rather than performing complex real-time calculations. This preliminary preparation of calibration data significantly reduces the processing time required during actual operation while maintaining high compensation accuracy
Data Source
AI summary
A memory device includes a cell array including a memory cell that includes a variable resistance element, a reference resistor configured to provide a resistance varying according to an adjustment code, and a read circuit configured to read data that is stored in the memory cell, based on a resistance of the variable resistance element and the resistance of the reference resistor. The memory device further includes a reference adjustment circuit configured to obtain a first calibration code corresponding to a temperature variation, and a second calibration code corresponding to a process variation, and perform an arithmetic operation on the obtained first calibration code and the obtained second calibration code, to obtain the adjustment code.