Research On Key Technology Of Indirect ToF Sensor Performance Optimization

In the era of the Internet of Things,the demand for 3D technology for depth sensors is increasing,especially in the fields of autonomous driving,virtual reality,augmented reality,robot vision,factory automation,artificial intelligence,and others.In the past ten years,three-dimensional imaging and ranging have become the most important and innovative research fields.The most popular way is the Time of Flight(To F)technology.To F technology is generally divided into two types according to different ranging principles: time difference ranging and phase ranging,that is,direct and indirect.The indirect To F depth sensor has better resistance to ambient light,is insensitive to dark current and noise,can achieve high frame rate and high resolution,and has a high detection resolution from a few centimeters to tens of meters,which can reach centimeters.Level or even millimeter level.Therefore,the chip design of the indirect To F depth sensor is a significant technical node for the further development of threedimensional imaging,and has great significance for the development of domestic threedimensional imaging sensors.This article will conduct research on indirect To F image sensors.Because the principle of the three-dimensional image sensor and the twodimensional image sensor are different,so they have different pixel structure,the evaluation of pixel performance and its readout circuit,For example,the threedimensional image sensor needs to maintain a good charge transfer efficiency under the high transfer frequency in the transfer gate,while the ordinary image sensor does not require this.Therefore,this paper will carry out targeted optimization of the 3D image sensor pixels suitable for the I-To F(Indirect Time of Flight)method to obtain better depth performance and obtain some optimized conclusions.By properly setting the pixel simulation,the following two conclusions can be drawn:1)Design the area of the photosensitive area of the pixel reasonably to obtain a suitable fill factor and demodulation contrast.Under the condition that the photosensitive area is 558)2 in the simulation,higher depth performance can be obtained.2)Design the width of the transmission gate reasonably to improve the potential distribution under the gate,and then improve the pixel Performance.In the case of a transmission gate width of 3um,better depth performance can be obtained in a higher modulation frequency.This article also analyzes the commonly used CDS circuits and It proposes a CDS circuit with a subtraction function for analysis and design,which simplifies the subsequent steps of processing data and speeds up data reading.The minimum relative error after optimization obtained through simulation can be controlled below 0.35%.Finally,a logic control unit for pixels and readout circuits suitable for the design is designed,and the correctness of the design is proved by mixed simulation.At the end of this article,the basic functions of the sensor after tape-out are tested.At the same time,analyze the grayscale performance of the pixels,the effect of different sensitive areas on the depth performance,and the impact of different transmission gate widths on the depth performance,which can be compared with pixel simulation results.The test shows that the comparison results have a very large correlation.After summarizing these test results,lading a solid foundation for subsequent pixel optimization and sensor performance optimization.