CN112399110A

Movatterモバイル変換

Info

Publication number: CN112399110A
Application number: CN201910752366.7A
Authority: CN
Inventors: 徐江涛; 陈全民; 查万斌
Original assignee: Tianjin University Marine Technology Research Institute
Current assignee: Tianjin University Marine Technology Research Institute
Priority date: 2019-08-15
Filing date: 2019-08-15
Publication date: 2021-02-23

Abstract

Description

Logarithmic photoresponse large-dynamic-range pixel structure

Technical Field

The invention belongs to the field of large-dynamic-range image sensors, and particularly relates to a logarithmic light response large-dynamic-range pixel structure.

Background

The dynamic range of the CMOS image sensor is usually 60-70 dB, and the difference between the dynamic range and the dynamic range of more than 100dB of human eyes is very large, so that the common sensor cannot truly restore the scene seen by the human eyes. In the application fields of automobile images, security monitoring and the like with complex ambient light, a large dynamic range sensor with the power of more than 100dB is also needed to acquire the bright and dark details in the scene. In order to make up for the deficiency of the CMOS image sensor, the large dynamic range CMOS image sensor is specially optimized for the dynamic range index so as to meet the requirement of a specific occasion. In the large dynamic range imaging scheme, in order to improve the imaging sensitivity under low light intensity, the Conversion Gain (CG) of a charge-voltage conversion (FD) node needs to be improved, but because of the limitation of the output voltage swing of a source follower in a pixel, the full-well capacity is also reduced when the CG is higher, so the design scheme of only adopting high CG is not enough to meet the requirements of people on the dynamic range.

In the traditional scheme, a dynamic range expansion technology based on multiple exposures is commonly used, and a large dynamic range is realized by fusing signals captured by multiple exposures, but the signal-to-noise ratio of the method is relatively low; another common method is to use two different pixels in the same pixel array, one with high CG and the other with low CG, and then fuse the different pixel signals to obtain the final signal, but this method will cause redundancy of pixels while considering optical uniformity and other issues; still another approach is to use transistors operating in the sub-threshold region to achieve logarithmic optical response, but with problems of noise and process mismatch at low light intensities.

Disclosure of Invention

Aiming at the problems in the prior art, the invention provides a logarithmic photoresponse large-dynamic-range pixel structure. Compared with the traditional 4T pixel structure, the pixel structure is additionally provided with a logarithmic response diode (LD), the logarithmic corresponding diode is used as a diode in reverse connection under low light intensity, a depletion region of the logarithmic response diode is connected with a depletion region of a Photodiode (PD) to be used as a pixel photosensitive region, partial photo-generated charges can be transferred to delay the saturation time of the PD under high light intensity, and the expansion of a dynamic range is realized.

A logarithmic photoresponse large dynamic range pixel structure is specifically as follows:

as shown in fig. 1, includes a Photodiode (PD), a reset transistor (RST), a Source Follower (SF), a select transistor (SEL), and a log-response diode (LD); the cathode of the photodiode is connected to the charge-voltage conversion node; the positive electrode is grounded, the drain electrode of the reset transistor is connected with a power supply (VDD), the source electrode is connected with a charge-voltage conversion node (FD), and the grid electrode is connected with a reset signal; the drain electrode of the source follower is connected with a power supply, the source electrode of the source follower is connected with the drain electrode of the selection transistor, and the grid electrode of the source follower is connected with the charge-voltage conversion node; the drain electrode of the selection transistor is connected with the source electrode of the source electrode follower, the grid electrode of the selection transistor is connected with a row selection signal, and the source electrode of the selection transistor is used as an output end;

the photodiode is a photosensitive area of the pixel, and converts received photons into charges through photoelectric conversion so as to determine the condition of received optical signals; the reset transistor is used for resetting the photodiode before exposure begins; the source follower SF is configured to convert a charge signal of the charge-voltage conversion node FD into a voltage signal, and the selection transistor is configured to control output of a pixel signal; the logarithmic response diode is reversely connected with a common diode under low light intensity, the depletion region works as a photosensitive region, and partial photo-generated charges are transferred under high light intensity, so that premature saturation of a PD region is avoided.

The principle of a logarithmic photoresponse large dynamic range pixel structure is as follows:

the charge-voltage conversion node is reset before the start of the exposure, i.e. the integration time. Potential V of charge-voltage conversion node after reset_FDThe PN junction of the log-response diode is in a reverse bias state above the anode of the log-response diode, and the depletion region of the log-response diode is connected with the depletion region of the photodiode. When in exposure, the depletion region of the photodiode and the depletion region of the logarithmic response diode are jointly used as photosensitive regions;

as the integration time proceeds or as the light intensity increases, the number of photo-generated electrons in the photodiode increases, and thus the N-type region potential of the photodiode gradually decreases, when the pixel operates in the linear response mode. That is, the potential of the charge-voltage conversion node changes linearly as the light intensity increases linearly;

when the potential of the N-type region of the photodiode is reduced to be less than the positive voltage V of the log-response diode_LAnd meanwhile, the working mode is automatically switched from the linear response mode to the logarithmic response mode. Namely, it isWhen the light intensity is linearly increased, the electric potential of the charge-voltage conversion node is changed in a logarithmic mode, so that the detectable maximum light intensity is effectively improved, and the dynamic range is expanded.

Compared with a pixel structure of a multi-exposure image sensor, the pixel structure with logarithmic light response and large dynamic range can realize higher signal-to-noise ratio and improve image quality; compared with a pixel structure which adopts a transistor working in a subthreshold region to realize logarithmic response, the pixel structure can realize lower noise under low light intensity and higher image quality; compared with a pixel structure with multiple Conversion Gains (CG), the pixel structure is simple in structure, fewer in transistor number and high in filling factor.

Drawings

FIG. 1 is a schematic circuit diagram of a pixel structure;

fig. 2 is a cross-sectional view of a pixel structure.

Detailed Description

The technical scheme of the invention is further clearly and completely described below by combining the attached drawings in the invention:

FIG. 1 is a schematic diagram of a pixel structure with a large dynamic range and logarithmic light response under high light intensity, which includes a Photodiode (PD), a reset transistor (RST), a Source Follower (SF), a selection transistor (SEL), a logarithmic response diode (LD), and a V_LA reference voltage that is the onset of a logarithmic optical response;

the photodiode is used for absorbing photons and generating photo-generated charges, the cathode of the photodiode is connected to the charge-voltage conversion node, and the anode of the photodiode is connected with the Ground (GND);

the reset transistor is used for resetting the photodiode before exposure, the drain electrode of the reset transistor is connected with a power supply (VDD), the source electrode of the reset transistor is connected to the charge-voltage conversion node, and the grid electrode of the reset transistor is connected with a reset signal;

the source follower SF is used for converting a charge signal of the charge-voltage conversion node into a voltage signal and outputting the voltage signal, the source electrode of the source follower SF is connected with the drain electrode of the selection transistor, the drain electrode of the source follower SF is connected with a power supply, and the grid electrode of the source follower SF is connected with the charge-voltage conversion node;

the selection transistor is used for controlling the output of a voltage signal, when the selection transistor is started, the voltage signal is transmitted to a subsequent circuit, the grid electrode of the selection transistor is connected with a selection signal, the drain electrode of the selection transistor is connected with the source electrode of the source electrode follower, and the source electrode of the selection transistor is used as an output end;

at low light intensity, the cathode voltage of the photodiode is higher than the reference voltage V_LAnd the PN junction of the logarithmic photoresponse diode is reversely biased, the depletion region of the logarithmic photoresponse diode is connected with the depletion region of the photodiode to be used as a photosensitive region, and the FD node potential linearly changes along with the linear increase of the light intensity. Under strong light, the voltage of the cathode of the photodiode is reduced to less than V under the action of photo-generated charges_LAt the moment, the PN junction of the logarithmic response diode is forward biased, the logarithmic response diode is in a conducting state, part of photo-generated charges can be transferred, and the FD node potential is in logarithmic change along with the linear increase of the light intensity. The cathode of the log-responsive diode is connected to the charge-voltage conversion node, and the anode is connected to a reference voltage V_L。

Fig. 2 is a schematic cross-sectional view of a logarithmic photoresponse large dynamic range pixel structure provided by the present invention, wherein an N-type implantation is performed on a P-type substrate (P-sub) as an N-type region of a photodiode, i.e., a cathode of the photodiode, and the P-substrate is an anode of the photodiode. In the left position of the N-type injection region of the photodiode, a P + injection is carried out, which forms the anode of the log-response diode and the reference voltage V_LAnd the negative electrode of the logarithmic response diode and the negative electrode of the photodiode share the same N-type region. At the right side position in the N-type injection region of the photodiode, N + injection is performed once as the position of the FD node, and the FD node is connected to the gate of the source follower SF.

The logarithmic photoresponse large-dynamic-range pixel structure provided by the invention can respond to a reference voltage V_LThe size of the pixel part is controlled, so that the pixel part enters a logarithmic light response mode under certain light intensity, the detectable maximum light intensity of the pixel part is adjusted, the dynamic range is expanded, and the requirements of different application environments are met.

Claims

1. A logarithmic photoresponsive large dynamic range pixel structure, characterized by: the device comprises a Photodiode (PD), a reset transistor (RST), a Source Follower (SF), a selection transistor (SEL) and a logarithmic response diode (LD); the cathode of the photodiode is connected to the charge-voltage conversion node; the positive electrode is grounded, the drain electrode of the reset transistor is connected with a power supply (VDD), the source electrode is connected with a charge-voltage conversion node (FD), and the grid electrode is connected with a reset signal; the drain electrode of the source follower is connected with a power supply, the source electrode of the source follower is connected with the drain electrode of the selection transistor, and the grid electrode of the source follower is connected with the charge-voltage conversion node; the drain electrode of the selection transistor is connected with the source electrode of the source electrode follower, the grid electrode of the selection transistor is connected with a row selection signal, and the source electrode of the selection transistor is used as an output end;

2. A log-photoresponsive high dynamic range pixel structure as claimed in claim 1, wherein:

resetting the charge-voltage conversion node before the start of exposure, i.e. the start of the integration time; potential V of charge-voltage conversion node after reset_FDThe PN junction of the logarithmic response diode is in a reverse bias state higher than the anode of the logarithmic response diode, and the depletion region of the logarithmic response diode is connected with the depletion region of the photodiode; when in exposure, the depletion region of the photodiode and the depletion region of the logarithmic response diode are jointly used as photosensitive regions;

as the integration time progresses or as the light intensity increases, the number of photo-generated electrons in the photodiode increases, and thus the potential of the N-type region of the photodiode gradually decreases, at which time the pixel operates in a linear response mode; that is, the potential of the charge-voltage conversion node changes linearly as the light intensity increases linearly;

when it is photoelectricThe potential of the N-type region of the polar tube is reduced to be less than the positive electrode voltage V of the logarithmic response diode_LWhen the light intensity is linearly increased, the potential of the charge-voltage conversion node changes in a logarithmic mode, so that the detectable maximum light intensity is effectively improved, and the dynamic range is expanded.