CMOS Non-Volatile Memory Cell Using Hot Carrier Injection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional non-volatile memory cells require high programming current, leading to high power consumption and limited simultaneous programming capabilities, with reliability issues due to high voltages used in Fowler-Nordheim tunneling.
Innovation Solution
The proposed solution involves generating hot carriers at both the source and drain junctions of transistors in CMOS-compatible memory cells, allowing for programming and erasure with minimal channel current, using hot electron or hot hole injection, and reducing the need for high voltages, thereby lowering power consumption and improving reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If drain avalanche hot carrier injection or Fowler-Nordheim tunneling is used to program conventional non-volatile memory cells, then the memory cells can be programmed and erased, but high programming current (several hundred microamps) is required leading to high power consumption
Solution Approach 1:
The patent changes the fundamental programming mechanism from high-current injection methods to hot carrier generation through velocity modulation. By modulating the velocity of charge carriers using RF signals rather than applying high programming currents, the power consumption is dramatically reduced while maintaining reliable programming capability. The invention transforms the programming approach from DC current-based to AC signal-based carrier velocity modulation.
Solution Approach 2:
The patent replaces the electrical field-based injection mechanism (Fowler-Nordheim tunneling or drain avalanche) with a mechanical oscillation-based approach. By using RF signals to modulate carrier velocity and generate hot carriers through dynamic heating, the system substitutes static high-voltage electrical fields with dynamic electromagnetic oscillations, achieving programming with minimal power consumption.
2Reliability
If high programming current is used to program conventional non-volatile memory cells, then programming can be achieved, but the number of memory cells that can be programmed simultaneously is limited
Solution Approach 1:
The patent segments the programming operation into independent parallel channels by using separate RF signal generators for each memory cell or group of cells. Each cell can be programmed independently with its own modulated RF signal, allowing simultaneous programming of multiple cells without current sharing constraints. This segmentation enables scalable parallel programming capability.
Solution Approach 2:
The patent creates a universal programming interface that can address and program any number of memory cells simultaneously using the same RF-based hot carrier generation mechanism. The system is designed to scale from single-cell to multi-cell programming without requiring different programming architectures, providing multi-functionality across different memory array sizes.
3Reliability
If Fowler-Nordheim tunneling is used for programming, then memory cells can be programmed, but reliability issues arise due to the use of high voltages
Solution Approach 1:
The patent replaces the high-voltage electrical field mechanism with RF-based carrier velocity modulation. Instead of applying high voltages to create tunneling fields, the system uses RF signals to dynamically heat and accelerate carriers, generating hot carriers through kinetic energy rather than static electrical fields. This eliminates high-voltage stress while maintaining programming capability.
Solution Approach 2:
The patent introduces RF signal modulation as an intermediary mechanism between the control signal and the memory cell programming. Rather than directly applying high voltage to the memory cell, the system first modulates RF signals to generate hot carriers, which then serve as the intermediary that transfers energy to program the cell without requiring high-voltage stress on the cell structure.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly reduces power consumption, enables simultaneous programming of a larger number of memory cells, and alleviates high-voltage reliability issues, allowing for faster and more reliable in-field updates.
Implementation Method 1
drain avalanche hot carrier injection can be used by applying a high voltage at the drain of a program transistor in a non-volatile memory cell. This generates a large electric field at the drain of the program transistor, which accelerates electrons onto the floating gate and programs the non-volatile memory cell
Implementation Method 2
Fowler-Nordheim tunneling can be used by applying a high voltage at a control gate of the program transistor in the non-volatile memory cell. This generates an electric field through a gate oxide separating the program transistor from the floating gate, allowing electrons to reach the floating gate
Data Source
AI summary
A memory cell includes a first transistor and a second transistor. Gates of the transistors are coupled together to form a floating gate. During programming, a first voltage is applied to a source and a drain of the first transistor, such as a negative voltage (like −5V). This causes electrons from the source and the drain of the first transistor to be injected onto the floating gate. During erasure, a second voltage could be applied to a source and a drain of the second transistor, such as a positive voltage (like +5V). This causes hot holes from the source and the drain of the second transistor to be injected onto the floating gate. The memory cell could also be erased by generating an electric field across a gate oxide of the second transistor, which allows electrons on the floating gate to tunnel through the gate oxide.