Buffered Direct Injection Pixel Autozero Circuit
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Infrared imagers face challenges with decreasing pixel size, leading to reduced well capacitor to pixel area ratio, which affects Signal-to-Noise Ratio (SNR), and existing in-pixel ADC circuits face issues with active power consumption, non-uniformity, and amplification errors due to space constraints and varying reverse bias across pixels.
Innovation Solution
A buffered direct injection pixel circuit with a common source amplifier and nulling switches is introduced, which includes an integration capacitor, injection transistor, and nulling capacitors to maintain constant reverse bias and improve SNR by controlling the gate voltage of the injection transistor using an amplifier, and performing auto-nulling operations to correct bias errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If amplifier is added to maintain constant reverse bias, then bias stability is improved, but power consumption increases
Solution Approach 1:
The amplifier operates in a periodic manner with distinct phases: during the integration period, the amplifier actively maintains constant reverse bias; during the readout period, the amplifier can be disabled or put in low-power mode since charge transfer is complete. This periodic operation reduces average power consumption while maintaining bias stability when needed
Solution Approach 2:
The amplifier's operating parameters (gain, bandwidth, bias current) are optimized for the specific application. The nulling capacitor and nulling switches dynamically adjust the amplifier's feedback path to correct bias errors only when necessary, reducing power consumption during periods when perfect bias stability is not critical
2Measurement precision
If nulling switches and capacitors are added to correct bias errors, then non-uniformity is reduced, but device complexity increases
Solution Approach 1:
The nulling capacitor is pre-charged during a nulling period before the integration phase begins. This preliminary action stores the bias error correction value, which is then automatically applied during integration without requiring continuous active correction, reducing complexity compared to continuous feedback systems while maintaining pixel uniformity
Solution Approach 2:
The nulling capacitor serves multiple functions: it stores bias error correction values, acts as a feedback element during nulling operations, and can be shared across multiple pixels in an array through column-parallel operation. This multi-functionality reduces the per-pixel complexity while achieving uniformity correction across the entire imager
Data Source
AI summary
A buffered direct injection pixel can be operated such that it is automatically zeroed. The operation includes: during a normal operating mode, controlling a gate voltage of an injection transistor with the output of an amplifier to control a bias of photo-current source, an inverting input of the amplifier being connected to input of the injection transistor through a nulling capacitor; during a nulling operation, closing a first switch to connect the nulling capacitor directly to an output of the amplifier; during the nulling operation, closing a second switch to directly couple the input of the injection transistor to a bias voltage causing the nulling capacitor to store a difference between an output of the amplifier and the bias voltage; and after the nulling operation, providing the voltage stored on the nulling capacitor to the inverting input by opening the first and second switches.