Noise Research Of Vertically-charge-transferring Pixel Sensor(VPS)and Its System

CCD and CMOS are the two mainstream image sensors,and their imaging performance has been increasing with the research of their internal noise sources and the innovation of noise suppression methods.It can be seen that noise analysis of the image sensor is necessary.As a new type of image sensor,Vertically-charge-transferring Pixel Sensor(VPS)has a problem of excessive noise,which greatly limits the image quality,so ameliorating noise performance is imminent.At present,no researchers have conducted systematic noise analysis on the VPS,and it is even more difficult to suppress noise.Therefore,noise analysis of VPS is extremely important,which is conducive to the development of a new generation of image sensors.First,this paper analyses the noise source at the device level about VPS,based on the working principle of device and connection method of array,a series of methods such as software simulation,theoretical calculation and experiment determined that the VPS has 49%diffusion crosstalk,33.2 electrons dark current,108?V readout noise and terrible FPN.Finally,based on the working principle of the VPS system readout circuit,the noise source existing in the readout circuit is analyzed.The reset noise from the read capacitor and the quantization noise of the ADC are determined by theoretical calculation.The influence of each control signal on the readout noise in the readout circuit is verified by statistical experimental methods,the optimal pulse duration and interval time between each control signal are determined,and the frame rate is improved under the premise of reducing noise.Through the system analysis of VPS noise,the key factors affecting the quality of imaging are dark current,crosstalk,reset noise,quantization noise and the FPN.Among those,the reset noise and the quantization noise can be optimized by increasing the readout capacitance and reducing the stepping of the ramp power supply respectively,and the FPN can be eliminated by an algorithm such as the zero-field correction,however,the dark current and the crosstalk are both the most serious part of noise and the most difficult to suppress,which greatly restricts the improvement of image quality.Based on the in-depth study of VPS noise,the scheme of suppressing dark current and crosstalk is explored,which provides a theoretical basis for further suppressing noise and enhancing the imaging quality,and indicates the direction of research improvement.