ToF Sensing Readout Circuit for Low-Power Depth Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Time-correlated single photon counting (TCSPC) methodology in Time-of-flight (ToF) sensors is power hungry and requires high data throughput, which is not suitable for miniaturization and low power consumption applications.
Innovation Solution
A sensing device with a SPAD array and a readout circuit that includes time-to-digital converters (TDCs), a delay locked loop, and a multiplexer, which performs pre-processing operations such as averaging or summing raw data frames to reduce data throughput and power consumption while maintaining depth measurement accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If TCSPC methodology is used in ToF sensor, then measurement precision is improved, but use of energy increases and data throughput increases
Solution Approach 1:
The patent divides the sensor array into multiple blocks, with each block having its own TDC and processing circuitry. This segmentation allows independent processing of data from different blocks, enabling the system to process fewer frames overall while maintaining measurement precision. The segmented architecture reduces total power consumption by avoiding redundant processing across the entire array for each frame.
Solution Approach 2:
The patent implements a TDC that is pre-configured with multiple latch groups that can be selectively activated. The system performs preliminary setup of the TDC structure with all potential latch groups ready, but only activates the necessary subset for each measurement task. This preliminary configuration allows rapid switching between different processing modes without full reinitialization, reducing power consumption while maintaining precision.
2Measurement precision
If TCSPC methodology is used in ToF sensor, then measurement precision is improved, but data throughput increases
Solution Approach 1:
By segmenting the sensor array into blocks with independent TDCs, the patent enables parallel processing of different blocks. This segmentation allows the system to achieve high measurement precision through dedicated per-block processing while reducing overall data throughput requirements through frame averaging across blocks, eliminating the need to process every single pixel frame at full resolution.
Solution Approach 2:
The patent introduces a block-level intermediate processing stage between individual pixel detection and final depth calculation. The TDC operates at the block level rather than individual pixel level, acting as an intermediary that aggregates and processes data from multiple pixels. This intermediary approach maintains measurement precision while significantly reducing the data throughput burden on subsequent processing stages.
3Volume of moving object
If miniaturization is pursued, then device size is reduced, but power consumption and data throughput requirements become more critical
Solution Approach 1:
The patent merges the TDC functionality directly into the sensor array structure, with TDCs integrated at the block level rather than as separate external components. This merging eliminates the need for separate data transfer interfaces and reduces overall device volume. The integrated architecture also reduces power consumption by eliminating redundant data movement and processing stages that would be required in a distributed configuration.
Solution Approach 2:
The patent divides the sensor into multiple small blocks, each with integrated TDC processing. This segmentation enables miniaturization by distributing processing functionality throughout the array rather than requiring a large centralized processing unit. Each block can be independently sized and optimized, allowing the overall device to be miniaturized while maintaining adequate processing capability for low power operation.
4Volume of moving object
If miniaturization is pursued, then device size is reduced, but data throughput requirements become more critical
Solution Approach 1:
The patent merges data processing functionality directly into the sensor array structure, combining detection and processing in a single integrated unit. This merging eliminates the need for separate high-speed data transfer interfaces between detection and processing components, reducing data throughput requirements. The integrated architecture processes data in-place, minimizing data movement and enabling miniaturization without compromising processing capability.
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 proposed solution achieves low power consumption and reduced data throughput while maintaining the accuracy of depth measurements, making it suitable for miniaturized ToF sensors.
Implementation Method 1
Each of the plurality of sensing pixels includes a single-photon avalanche diode that detects the reflected light from the object and generates an electrical signal when a photon is detected
Implementation Method 2
a time-to-digital converter coupled to the plurality of sensing pixels and configured to convert a time-of-flight value to a digital output
Data Source
AI summary
A sensing device that is configured to determine a depth result based on time-of-flight value is introduced. The sensing device includes a delay locked loop circuit, a plurality of time-to-digital converters, a multiplexer and a digital integrator. The delay locked loop circuit is configured to output a plurality of delay clock signals through output terminals of the delay locked loop circuit. The plurality of time-to-digital converters include a plurality of latches. The multiplexer is configured to select a sub-group of m latches among the latches of the plurality of time-to-digital converters to be connected to the output terminals of the delay locked loop circuit according to a control signal. The digital integrator is coupled to the plurality of time-to-digital converters and is configured to integrate digital outputs generated by the time-to-digital converters in each of n cycles to generate n raw data frames, wherein m and n are natural numbers, and the n raw data frames are used to generate the depth result.