Multi-Column Shared Readout Architecture for Imager Throughput
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current semiconductor imager readout architectures face limitations in high-speed and large-format imaging applications due to slow readout speed, high power consumption, increased noise, and parasitic capacitance, making them unsuitable for high-accuracy and high-resolution imaging.
Innovation Solution
A multi-column shared readout architecture is introduced, where columns are grouped and each group has dedicated readout channels with sample-and-hold circuits, programmable gain amplifiers, and analog-to-digital converters, along with digital offset correction and gain calibration, allowing for serial readout within groups but parallel processing across groups, reducing noise and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If serial readout scheme is used with single gain amplifier and ADC, then device complexity is reduced, but readout speed and throughput are limited
Solution Approach 1:
The pixel array columns are divided into multiple groups, with each group having its own dedicated readout channel including sample-and-hold circuit, gain amplifier, and ADC. This segmentation allows parallel processing of multiple column groups simultaneously, dramatically increasing readout throughput while keeping each individual channel relatively simple
Solution Approach 2:
The patent transitions from a single-dimensional serial readout architecture to a multi-dimensional parallel architecture by adding the dimension of multiple simultaneous readout channels. Multiple columns are read out in parallel across different channels, effectively moving from sequential processing to concurrent processing in multiple dimensions
2Productivity
If high data throughput rate is achieved in large-format imagers, then imaging speed is improved, but power consumption becomes unacceptably high
Solution Approach 1:
The readout function is segmented across multiple independent channels, allowing the total data throughput to be distributed across parallel paths. Each channel operates at a lower individual data rate while collectively achieving high system throughput, reducing the power burden on any single circuit element
Solution Approach 2:
Multiple column groups are merged into shared readout channels that process data from multiple sources simultaneously. This merging allows efficient utilization of readout resources and reduces redundant circuitry, lowering overall power consumption while maintaining high throughput
3Measurement precision
If switched-capacitor circuits are used in readout circuitry, then precision is improved, but chip size becomes unacceptably large
Solution Approach 1:
The precision-critical switched-capacitor circuits are segmented and distributed across multiple readout channels rather than requiring one large centralized precision circuit. This allows the use of smaller, more compact switched-capacitor circuits in each channel while collectively achieving the required overall precision
Solution Approach 2:
The patent moves from a single large precision readout circuit to multiple smaller precision circuits operating in parallel across different channels. This dimensional transition from one large circuit to many small circuits reduces the area required for each precision element while maintaining overall system precision
4Device complexity
If serial readout is used, then device complexity is reduced, but readout noise increases
Solution Approach 1:
The readout function is segmented into multiple parallel channels, each with its own sample-and-hold circuit and amplification stage. This segmentation allows each channel to process a subset of columns with dedicated low-noise circuitry, reducing the cumulative noise impact compared to a single serial channel handling all columns
Solution Approach 2:
Sample-and-hold circuits are implemented at the input of each readout channel to capture and hold the pixel signals before further processing. This preliminary action stabilizes the signals early in the readout path, reducing noise accumulation and interference during subsequent amplification and conversion stages
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 enhances throughput, reduces readout noise, achieves higher light sensitivity, and provides better gain linearity and finer gain settings, enabling efficient scaling for reduced pixel pitch while maintaining improved image quality and reduced power consumption.
Implementation Method 1
a photosensor, for example, a photogate, photoconductor, or a photodiode overlying a substrate for accumulating photo-generated charge
Data Source
AI summary
An imager and method of operating an imager employing multi-column shared readout circuitry. Columns of a pixel array are organized into groups, each group having a respective multi-column shared readout circuit. The columns of each group are readout serially but in parallel with the columns of other groups. Each multi-column shared readout circuit may comprise a black level correction clamp, a multi-column analog gain amplifier, a analog-to-digital converter, a digital offset correction block, and a digital gain calibration block. A single-column analog gain amplifier may amplify an analog pixel signal value of each column prior to processing by a respective multi-column shared readout circuit.