When mobile phones came to the era of one-inch soles, the development speed of imaging gradually slowed down, and it was no longer feasible to rely on the simple and crude method of directly increasing the size and area of the sensor, and consumers began to consider whether to sacrifice the size and appearance of mobile phones for the sake of imaging capabilities. How to continuously optimize the quality of small sensors without changing the volume of the sensor has become the primary problem for sensor manufacturers.
Samsung ISOCELL: Dual Vertical Transfer Gate (D-VTG)
The growing demand for CMOS on mobile devices has prompted Samsung Electronics to increase the number of pixels from 108 million to 200 million. However, due to the limitations of the optical module structure and the size of the image sensor, a smaller pixel size is required for high resolution, which means that the size of the photodiode must be further reduced. To meet this challenge, Samsung has continuously improved its ISOCELL pixel technology based on the FDTI structure, successfully achieving a pixel size of 0.64μm, and has continuously reduced the DTI size to optimize plasma-assisted doping (PLAD) to maintain a full-well capacity of 6,000e-.
D-VTG technology is another of Samsung's process breakthroughs in small pixels, which enables dual vertical transmission gates (D-VTGs) with two VTGs per pixel and applies them to the S5KHP2 image sensor. Thanks to the new Double Vertical Transmission Gate (D-VTG) technology, it not only improves the controllability of the transmission gate voltage, but also greatly improves the transmission capacity compared with the single-tube transmission gate VTG, and optimizes the parameters such as gap, depth and taper slope of the VTG, maximizing the electronic transmission efficiency.
The ISOCELL HP2 is Samsung's first image sensor with D-VTG technology and sufficient full-well capacity to achieve a small size and high pixels. Each pixel absorbs 33% more electrons per pixel than the previous generation of 200 million pixel image sensors (up to 10,000e FWC per pixel), which effectively improves color performance. As the full-well capacity increases, each pixel is able to utilize more electrons, further improving the ability to line colors in bright light conditions.
(S23 Ultra VS S22 Ultra VS iPhone 14 Pro,感 /1.3英 ,图源 主@宇 )
In terms of sample performance, the S23 Ultra with the D-VTG technology sensor S5KHP2 can greatly reduce the problem of overexposure in bright light performance, while more accurately reproducing the color of the object itself.
Sony Exmor T: Dual-layer transistor pixel structure (DLT)
Sony has developed an innovative dual-layer transistor pixel structure that separates photodiodes and pixel transistors into different substrate layers for stacking. This structure results in a significant increase in the amount of saturated signal, increasing the dynamic range and reducing noise. With this structure, imaging characteristics can be significantly improved even at smaller pixel sizes.
Specifically, Sony has made two main changes:
Separate the photodiode from the pixel transistor to create a stack structure
Sony has made use of SSS's stacking technology to create a stacked structure that separates the layers of photodiodes, which are used to be distributed on pixel chips to convert light into electrical signals, and pixel transistor layers, which control signals, onto different substrates, thereby creating a stacked structure and creating a larger circuit space.
The increase in photodiode capacity and amplification transistor size
By separating the photodiode from the pixel transistor layer and building a stack structure, the capacity of the photodiode can be expanded and the amount of saturated signal can be increased by about 2 times. The result is a successful expansion of the dynamic range, which represents the range of light and dark differences that can be imaged.
In addition, pixel transistors such as reset transistors (RSTs), selective transistors (SELs), and amplification transistors (AMPs) other than transmission control transistors (TRGs) are distributed on substrate layers without photodiodes, thereby increasing the size of the amplification transistors. This improvement significantly reduces the noise that tends to occur when shooting dark scenes such as night scenes.
In the sample given by Sony, it is not difficult to see that the dual-layer transistor pixel structure has two advantages
1. Reduce overexposure and expand dynamic range
2. Reduce noise in low light
·豪威TheiaCel:横向溢出积分电容器(LOFIC)
The basic principle of transverse overflow integral capacitance technology is to place a high-density capacitor in each photodiode of the mobile phone image sensor to collect photoelectrons that might otherwise be spilled due to saturation. In this way, when the number of photoelectrons converted by the photodiode exceeds the maximum amount that it can carry, the excess photoelectrons will flow into the adjacent capacitors instead of being "overexposed" due to overflow, which makes the highlight information in the shooting scene better retained by the sensor, and the output is closer to the real light and shadow scene.
Before LOFIC, large and small capacitors were not interoperable, but now the two can be used together, so it is possible to expand the fwc and realize single-pixel DCG HDR at the same time.
(Comparison of Honor Magic6 Pro and Honor Magic6 Ultimate,
Pro mode, shutter lock 1/125 sec., rest auto)
AS YOU CAN SEE, IN THE SAME SCENE, THE ISO OF THE OV50K WITH LOFIC IS MUCH LOWER THAN THAT OF THE OV50H WITHOUT LOFIC, WHICH MEANS THAT MORE DETAIL CAN BE PRESERVED AND NOISE CAN BE REDUCED.
Summary
Samsung, Sony, and OmniVision have all come up with innovative technology solutions in the field of mobile imaging.
Samsung's dual vertical transmission gate technology achieves the goal of high FWC even in small pixels, and increases dynamic range while high resolution.
Sony uses a double-layer transistor pixel structure: the FWC is improved, the dynamic range is increased, and the noise performance in low light is also reduced.
OmniVision's transverse overflow integral capacitor technology has greatly improved the FWC and improved dynamic range.
These three have different technical directions, but they all focus on improving the dynamic range, but what needs to be known is that as consumers, in addition to understanding that dynamic range is only a part of the sensor's performance, we also need to consider the impact of other factors on the shooting effect, such as resolution, high sensitivity and noise control, shooting speed and response speed, etc., and finally we have to experience it more clearly to understand our shooting needs and preferences.
END