Trench Capacitor Stacked Image Sensor Noise Removal

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing CMOS image sensors face challenges in miniaturization and high image quality due to the need for noise removal in pixel columns, which requires significant transfer capacity and occupies valuable space, limiting the compatibility between miniaturization and image quality.

Innovation Solution

The image sensor incorporates a trench-structured capacitor with a dielectric film and electrode, connected to a column reading circuit and scanning circuit, which amplifies and resets imaging signals, allowing for noise removal without the need for a sampling condenser, thereby reducing the occupied area and enhancing miniaturization while maintaining high image quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a noise removing portion with transfer capacity is provided in each pixel column, then reset noise can be removed, but the occupied area increases, limiting miniaturization

Engineering Contradiction:
Improvenoise removal capabilityVSAvoidoccupied area of noise removing portion
Core Design Contradiction:
ReliabilityVSArea of stationary object

Solution Approach 1:

The patent moves the noise removing portion from the pixel chip to a separate circuit chip, transitioning from a two-dimensional planar arrangement to a three-dimensional stacked configuration. This dimensional change allows the noise removing portion to be physically separated from the pixel array, reducing the occupied area on the pixel chip while maintaining noise removal functionality through vertical stacking and bump electrode connections.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent divides the image sensor into two separate chips: a pixel chip containing the pixel array and a circuit chip containing the noise removing portion and column reading circuits. This segmentation separates the noise removal function from the pixel array, allowing each component to be optimized independently and reducing the overall occupied area through selective placement of functional elements.

Inventive Principle:
Principle #1Segmentation

2Reliability

If transfer capacity is increased to improve noise removal, then image quality improves, but the capacity occupies valuable space

Engineering Contradiction:
Improveimage qualityVSAvoidvaluable space in pixel column
Core Design Contradiction:
ReliabilityVSArea of stationary object

Solution Approach 1:

By relocating the transfer capacity to a separate circuit chip in the vertical dimension, the patent eliminates the space constraint on the pixel chip. The transfer capacity can be made larger for better noise removal performance without competing for valuable horizontal space in the pixel column, as it now resides in the third dimension on the circuit chip.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Area of stationary object

If miniaturization is pursued, then device size decreases, but noise removal capability is compromised

Engineering Contradiction:
Improvedevice sizeVSAvoidnoise removal capability
Core Design Contradiction:
Area of stationary objectVSReliability

Solution Approach 1:

The stacked configuration allows the device to be miniaturized in the horizontal plane while maintaining or even enhancing noise removal capability through the vertical dimension. The circuit chip can provide sufficient transfer capacity for effective noise removal without increasing the footprint area, as the additional capacity is achieved through vertical stacking rather than horizontal expansion.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

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 configuration enables further miniaturization of the image sensor while maintaining high image quality by effectively removing noise from the imaging signal, allowing for more compact designs without compromising image fidelity.

Implementation Method 1

each capacitor being provided at a respective one of the first transfer lines and having a trench structure including: a first diffusion layer connected to the first transfer line; a trench formed in the first diffusion layer; a dielectric film formed inside the trench; and an electrode provided inside the dielectric film

Methodology Applied
Scientific EffectCapacitance: Capacitance

Implementation Method 2

each pixel being configured to receive external light, generate an imaging signal in response to a light receiving amount and output the generated imaging signal

Methodology Applied
Scientific EffectPhotoelectric Effect: Photoelectric Effect

Data Source

PatentUS10153311B2Image sensor
Publication Date: 2018.12.11 OLYMPUS CORPORATION(JP)
  • US10153311B2 patent drawing
  • US10153311B2 patent drawing
  • US10153311B2 patent drawing

AI summary

An image sensor includes: a pixel chip provided with a plurality of pixels, a plurality of first transfer lines, and a plurality of capacitors; a circuit chip provided with a plurality of column reading circuits, a plurality of column scanning circuits, a second transfer line, and a constant current source; and a connection portion stacked and provided between the pixel chip and the circuit chip and configured to connect a capacitor, which is arranged in the pixel chip and has a trench structure, and a first transistor arranged in the circuit chip to each other via an electrode. The capacitor is configured to form a transfer capacity removing a noise included in an imaging signal and connect the pixel chip and the circuit chip to each other via the electrode and the connection portion.