CMOS Image Sensor Overflow-Charge Readout for Wider Dynamic Range
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
CMOS image sensors face limitations in achieving broader dynamic range and higher frame rates due to charge overflow from photodiodes not being utilized in real time, leading to constraints on noise reduction and effective pixel region expansion, while also being cost-inefficient with ADC and memory components distributed across each pixel.
Innovation Solution
A solid-state imaging device with pixels that include a photoelectric conversion element, a transfer element, an output buffer, and a comparator performing first and second comparison processing to handle overflow and accumulated charges separately, allowing for real-time utilization of charges and efficient noise reduction, while optimizing pixel density and cost-effectiveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If an ADC including a comparator and memory part is arranged in each pixel to enable global shutter function, then the global shutter capability is improved, but the area occupied by each pixel increases and the effective pixel region cannot be expanded to the maximum limit
Solution Approach 1:
The pixel array is divided into a first pixel region and a second pixel region. The first pixel region contains pixels with ADCs for high-light conditions, while the second pixel region contains pixels without ADCs for low-light conditions. This segmentation allows each region to be optimized for its specific function, maximizing the overall effective pixel region while maintaining global shutter capability where needed.
Solution Approach 2:
Different pixel regions are assigned different functional qualities based on local requirements. The first pixel region (with ADCs) is optimized for handling overflow charges in high-light conditions, while the second pixel region (without ADCs) is optimized for maximum sensitivity in low-light conditions. This local quality differentiation resolves the contradiction by providing appropriate functionality only where needed.
2Object-affected harmful factors
If the transistor size is enlarged to reduce flicker noise, then the noise reduction is improved, but the area occupied by each pixel increases
Solution Approach 1:
The pixel array is segmented into regions with different transistor sizes optimized for their specific functions. The first pixel region uses larger transistors to reduce flicker noise for high-light conditions, while the second pixel region uses smaller transistors to maximize pixel density for low-light conditions. This segmentation allows noise reduction without uniformly increasing all pixel areas.
Solution Approach 2:
The transistor size parameter is changed differently across different pixel regions based on their specific requirements. By adjusting the transistor size parameter locally rather than uniformly, the invention achieves noise reduction where needed while maintaining small pixel areas in regions where sensitivity is more critical.
3Object-affected harmful factors
If a capacity is added to the output of the comparator to lower the bandwidth and achieve noise filtering effect, then the noise filtering is improved, but the inversion delay of the comparator increases and the frame rate cannot be raised
Solution Approach 1:
The pixel array is divided into first and second pixel regions with different processing characteristics. The first pixel region is optimized for high-speed processing with larger bandwidth to maintain high frame rates, while the second pixel region can use lower bandwidth for maximum noise filtering. This segmentation allows the system to achieve high frame rates overall while still providing noise filtering where applicable.
4Ease of operation
If ADCs are arranged in each pixel to achieve digital signal processing, then the signal processing capability is improved, but the manufacturing cost increases
Solution Approach 1:
The pixel array is segmented such that only the first pixel region contains ADCs for digital signal processing, while the second pixel region uses simpler analog processing. This segmentation reduces the total number of ADCs needed, thereby lowering manufacturing costs while still providing digital processing capability for high-light conditions where it is most beneficial.
Solution Approach 2:
Digital signal processing capability is provided locally in the first pixel region where it is most needed for handling overflow charges in high-light conditions, rather than uniformly across all pixels. This local quality approach reduces overall manufacturing cost while maintaining signal processing capability where it provides the most value.
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
The solution enables broader dynamic range, higher frame rates, reduced noise, and expanded effective pixel regions while minimizing costs and maintaining area efficiency, allowing for real-time charge handling and efficient signal processing.
Implementation Method 1
a photoelectric conversion element which accumulates a charge generated by photoelectric conversion in an integration period
Data Source
AI summary
An AD conversion part has a comparator for performing comparison processing comparing a voltage signal read out by a photoelectric converting and reading part and a reference voltage and outputting a digitalized comparison result signal, the comparator, under the control by a reading part, performs first comparison processing for outputting a digitalized first comparison result signal with respect to a voltage signal corresponding to an overflow charge overflowing from a photodiode PD1 to a floating diffusion FD1 in an integration period and second comparison processing for outputting a digitalized second comparison result signal with respect to a voltage signal corresponding to an accumulated charge of the photodiode PD1 transferred to the floating diffusion FD1 in a transfer period after the integration period. Due to this, it becomes possible to substantially realize a broader dynamic range and higher frame rate.