Sony SWIR Sensor Designers Interview

Sony publishes an interview with its SWIR InGaAs stacked sensor designers. Few quotes:

"...the miniaturization of SWIR image sensors had been hindered by the use of bump connection. We already knew that this problem could possibly be solved using the Cu-Cu connection*2, which is the stacking technology Sony had developed for years for image sensors. This technology would make it possible to align pixels at a micro-pitch. Meanwhile, conventional image sensors use silicon as a photoelectric conversion layer, but this material does not absorb the SWIR range of light. So, we needed to use indium gallium arsenide (InGaAs) as the photodiode material that can absorb the SWIR spectrum and convert the light energy into electric signals. This material was never used in Sony’s image sensors before, but however, Another division of Sony had the compound semiconductor technology to produce InGaAs.

Another point was that conventional SWIR image sensors had many defects. White patches would appear in the dark image due to the quality issues particular to InGaAs. We had the technology to make compound semiconductors based on our years of expertise in developing laser technology, so we aimed to leverage it in creating defect-free, high-quality products.

As it was the Group’s first-ever SWIR image sensor to be developed using InGaAs, we thoroughly reviewed challenges to be addressed, which resulted in more than 300 in the early stages of the development. As we proceeded further into the development, we found more issues to deal with. There turned out to be so many challenges awaiting."

Yole on Mobile CIS Market

Yole Developpement publishes "Smartphone flagship battle from sensors to modules to image quality - Webcast"

China seems to become an image sensor land: 3 out of 10 largest semiconductor companies are designing image sensors, while SMIC, the 4th company, manufactures them:

Sony Reports 9% YoY CIS Sales Drop

Sony reports lower mobile image sensor sales in the last quarter, leaves full-year forecast unchanged:

New Theory of RTS Noise

LAAS (Laboratoire d'analyse et d'architecture des systèmes), ST, ISAE-SUPAERO, CEA, and CNR-IOM publish a paper "Clusters of Defects as a Possible Origin of Random Telegraph Signal in Imager Devices: a DFT based Study" by Antoine Jay, Anne Hémeryck, Fuccio Cristiano, Denis Rideau, Pierre-Louis Julliard, Vincent Goiffon, Alexandre Le Roch, Nicolas Richard, Layla Martin-Samos, and Stefano de Gironcoli, presented at International Conference on Simulation of Semiconductor Processes and Devices (SISPAD) in Sept. 2021.

"The origin of the random telegraph signal (RTS) observed in semiconductors-based electronic devices is still subject to debates. In this work, by means of atomistic simulations, typical clusters of defects as could be obtained after irradiation or implantation are studied as a possible cause for RTS. It is shown that:

(i) a cluster of defects is highly metastable,

(ii) it introduces several electronic states in the band gap,

(iii) it has an electronic cross section much higher than the one of point defects.

These three points can simultaneously explain why an electron- hole generation rate can switch with time, while respecting the experimental measurement."

The new theory is said to be able to explain all of the following RTN observations simultaneously:

IC Insights: CMOS Sensor Market to Grow by Only 7% in 2021

IC Insights is disappointed by the slow CMOS sensor market growth this year:

"The total optoelectronics market is expected to be held back by sales growth of just 7% in CMOS image sensors this year.  CMOS image sensor leader Sony blames softer growth conditions in 2021 on trade frictions between the U.S. and China and a “deterioration of product mix.”  CMOS image sensor sales have also been impacted by market fluctuations in some end-use applications, and shortages of ICs and other components used in digital-imaging systems."

Last 4 Days of Free Download of Single-Slope Column-Parallel ADC Book!

Only 4 days remain for free download of Now Publishers book "Welcome to the World of Single-Slope Column-Level Analog-to-Digital Converters for CMOS Image Sensors" by Albert Theuwissen and Guy Meynants.

Complimentary downloads of this book will be available till 1st of November 2021. After that date, you can receive the alert member discount price of $40 (includes non-trackable shipping) by quoting the Promotion Code: 137791. Please note that the discount price applies only to purchases of print copies by individuals paying in advance by credit card. The discounted price will not be honored for institutions or booksellers.

Chinese Startup RuisiZhixin Develops 8MP Combined Event-Driven and Regular Sensor

36kr.com, EET-China (Google automatic translation): Chinese startup Beijing Ruisi Zhixin Technology (English name AlpsenTek) presents hybrid vision sensors combining event-driven sensor and regular frame-based one in a single pixel array:

"The ALPIX-Pilatus (ALPIX-P for short) released by Ruisizhixin this time is developed based on "Hybrid Vision" and is the world's first chip that integrates traditional image sensor technology and bionic event camera technology in the same pixel . ALPIX-P can quickly switch between the traditional image sensor mode and the bionic event camera mode. In image mode, ALPIX-P is a global exposure sensor with a maximum frame rate of 120 frames, which is completely consistent with existing traditional image sensors and is fully compatible with existing mature vision systems. In the bionic event camera mode, ALPIX-P has the characteristics of ultra-high frame rate, high effective information ratio, and large dynamic range.

At present, the chip has begun to send samples. It also verified that the "Hybrid Vision" technology is feasible in mass production, cost, and performance .

Based on the design of ALPIX-P, Ruisi Zhixin has developed two ALPIX series sensor chip products, namely ALPIX-Titlis (abbreviated as ALPIX-T) fused low-power bionic vision sensor and ALPIX-Eiger (abbreviated as ALPIX-E). ) Fusion high-end bionic vision sensor.

The team has thoroughly studied the technical route of the bionic event visual sensor. In addition, the core founders all graduated from Cambridge University, ETH Zurich, Zhejiang University and other prestigious universities, and have many years of working experience in NXP, ARM, Freescale, Intel, Magic Leap and other companies. At the beginning of the establishment, the core team members were only about 6 people. Now the team has expanded to 50 people, of which R&D personnel accounted for 85% .

Recalling the financing history of RISZ, the company has now completed two rounds of financing. At the end of 2019, the company completed an angel round of financing of tens of millions of RMB. The investors were Zhongke Chuangxing and Lenovo Venture Capital . At the end of 2020, Ruisi Zhixin completed a Pre-A round of financing of nearly 100 million yuan, which was jointly led by Hikvision and Yaotu Capital."

The company explains the advantages of its hybrid sensors:

"Compared with traditional CIS, the bionic vision sensor has the characteristics of fast speed (>5000 frames/s), low power consumption (tens of mW), small data volume and large dynamic range (>120dB), which can successfully resolve the problem areas currently faced by computer vision as described earlier.

However, during actual operation, many scenarios require both event flow signals for rapid prediction, and traditional images for feature extraction.

AlpsenTek realized that the integration of bionic vision sensor chip and high-end image sensor chip technology may be a breakthrough.

This two-in-one function has the characteristics of low power consumption, high speed, high data efficiency and high dynamic range of bionic vision sensors. Compared with similar competing products, ALPIX effectively improves the signal-to-noise ratio of the chip, reduces noise, improves dark light performance, and meets the high performance needs demanded from customers.

Early on, many industrial giants in China recognized the potential of sensor chip technology. When AlpsenTek emerged with its unique technical strength, it also attracted the attention of giants in the industrial chain.

Recently, AlpsenTek secured USD15.5 million in a Pre-A round of financing. Investors include: Hikvision, Glory Ventures, Fargo Capital, iFlytek Venture Capital, Sunny Optical, Allwinner Technology, and Cowin Capital, in addition to previous shareholders Lenovo Capital and Casstar, who continue to invest."

"A product’s appearance is not designed, rather it is driven by the demand of the users,” the company's CEO Deng Jian says, “New technology can't change life, but new products have the power to do this.”

The company's patent applications EP3731516 and WO2020216867 show its pixel structure:

Old Presentations on Event-Driven Sensors

Iaria publishes two Laurent Fesquet's (University Grenoble Alpes) presentations on event-driven sensors: 2016 "Low-Power Event-driven Image Sensor Architectures" by Laurent Fesquet, Amani Darwish, Gilles Sicard and 2018 "Sensing and Sampling for Low-Power Applications" by Laurent Fesquet alone. Few slides are given below:

CIS Production BIST

ST and University of Montpellier publish a paper "A Fast and Low Cost Embedded Test Solution for CMOS Image Sensors" by Julia Lefèvre, Philippe Debaud, Patrick Girard, and Arnaud Virazel presented at 2021 IEEE International Test Conference (ITC).

"This paper presents a novel test solution directly embedded inside CMOS Image Sensors (CIS) to sort out PASS and FAIL dies during production test. The solution aims at reducing test time, which can represent up to 30% of the final product cost. By simplifying the way optical tests are usually applied with an ATE, the proposed Built-In Self-Test (BIST) solution overcomes the drawbacks of long test time and huge amount of test data storage. We experimented our solution by considering that roughly half of the tests usually performed with an ATE can be embedded and applied using the proposed fast and low cost BIST engine. Results obtained on more than 2,400 sensors have shown that our solution reduces test time by about 30% without impacting the defect coverage. The area cost of our solution is about 1% of the digital part of the sensor, i.e., approximately 0.25% of the total sensor area. The proposed embedded CIS test solution outperforms existing solutions in terms of area overhead and test time saving, thus encouraging its future implementation in an industrial production flow."

Smartsens Progresses with IPO at Valuation of $4.4B

EE-Ofweek (Google automatic translation): Smartsens plans to go public and has been accepted by the SSE Science and Technology Innovation Board on June 28. The sponsor is China Securities. It is proposed to publicly issue no more than 49.1 million shares, not less than 10%, and plan to raise 2.82 billion yuan for the equipment and system construction project of the R&D center, the image sensor chip test project, CMOS sensor chip upgrade and industrialization project and supplementary working capital (790 million yuan).

These numbers put Smarsens valuation at 28.2 billion yuan, or $4.4 billion.

Smartsens CEO and founder Xu Chen (Richard Xu) is the controlling shareholder and actual controller of the company, and the CTO Mo Yaowu is the second largest shareholder. Among them, Xu Chen directly holds 15.23% of the company’s shares, with 47.32% of the voting rights, and Mo Yaowu directly holds 6.66% of the company’s shares, with 4.14% of the voting rights, and the two together Holds 21.89% of the company’s shares, with a total of 51.46% of the company’s voting rights and these voting rights are owned or controlled by Xu Chen.

Revenue has grown by leaps and bounds, but compared with peers, the scale is still small, and the gross profit margin is lower than the peer average.

Smartsens products are used in many high-tech applications such as security monitoring, machine vision, and intelligent vehicle electronics. From 2018 to 2020 and from January to March 2021, the company's operating income was 325 million yuan, 678 million yuan, 1.527 billion yuan and 541 million yuan, respectively, with a year-on-year increase of 92% and 124.89% in 2019 and 2020; net profits were 166 million yuan, 242 million yuan, 121 million yuan and 69 million yuan, respectively, with a total loss of 408 million yuan in 2018 and 2019 Yuan, starting to turn losses into profits in 2020.

From 2018 to 2020 and from January to March 2021, the company's FSI series sales were 56 million units, 86 million units, 139 million units, and 49 million units, with revenues of 284 million yuan, 480 million yuan, 880 million yuan and 288 million yuan, accounting for 87.45%, 70.61%, 57.66% and 53.3% of revenue, respectively, and their proportions continued to decline. ; The company’s sales of BSI-RS products were 2.0483 million, 9.243,900, 27,624,400, and 11.11 million, respectively, with revenues of 0.36 billion yuan, 149 million yuan, 473 million yuan and 157 million yuan, the proportion of revenue increased from 10.99% in 2018 to 30.97% in 2020.

EPIC Meeting on Low-Light Imaging

EPIC Technology Meeting on Low Light Cameras Technology and Applications features Valeo, Ibeo, Leica, and Thales presentations. Few slides are below:

Sigmaintell: CIS Prices Go Down

Science and Technology Materials quotes Sigmaintell China CIS market price tracker predicting 5-7% price decline in Q4 2021:

"At the beginning of 21Q4, the CIS inventory level continues to rise, but there is no significant increase in demand. The specific analysis is as follows: on the supply side, affected by the decline in demand, the supply of low- and medium- pixel capacity is reduced, and the inventory is prioritized. The new capacity of high pixel products continues to increase in 21Q4. On the demand side, the demand for smartphones has been reduced, and the demand for CIS stocking has been continuously reduced. In summary , the overall pixel supply-demand relationship in October shows an oversupply situation. According to the latest survey data from Sigmaintell in October, the prices of low pixel products have begun to loosen up, and the prices of 48M+ pixel products have continued to fall."

Quantum Dot Sensor Company SWIR Vision Raises $5M in A-Round

Optics.org, EINPressWire: SWIR Vision Systems, a Durham, North Carolina, company developing colloidal quantum dots (CQDs) image sensor has raised $5M in a round A of funding. SWIR Vision Systems has been founded in 2018 as a spin-off from the Research Triangle Institute. The company's “Acuros” family of thermoelectrically cooled CMOS-based CQD sensors reaches 2.1MP resolution.

“While InGaAs SWIR cameras are generally constrained to 640×512 or 1280×1024 pixel formats, our pioneering CQD sensor technology enables 2.1 megapixel, full-HD resolution, the first commercially available SWIR product of its kind.”

The new eSWIR version of the product family operates across an extended wavelength range, between 350 nm and 2000 nm, with a pixel pitch of 15 µm. 

The company's Vimeo video compares regular RGB imaging with SWIR in foggy environment.

China Domestic Market Dynamics

Sohu: Sunrise Big Data publishes its analysis of unit volume CIS market in Q1 2021. Galaxycore quickly captures a market share from Sony, Samsung, and Omnivision, as compared with Q4 2020:

"According to Sunrise Big Data statistics, global smartphone CMOS image sensor shipments in Q1 2021 will be 1.413 billion, of which Galaxycore will ship about 447 million, accounting for 31.6% , ranking first.

According to the observation, Galaxycore has also become a sensor head supplier in the security, automotive and other sub-sectors. In the first half of 2021 , Galaxycore CIS non-mobile phone business revenue has exceeded 100 million US dollars , exceeding the full year of 2020."

ToF Basics by Terabee

Terabee publishes a few webinars on ToF technology basics and applications:

TI Proprietary V3Link Competes with MIPI A-PHY

TI presentation "V3Link Industrial SerDes" unveils the company's approach to compete with MIPI A-PHY and Auto-Serdes standards:

Smartsens Improves its AI Sensor Series

CoreIntelligence (Google translation): SmartSens keeps a high pace of incremental improvements and launches three new 4MP image sensors in AI series (AI stands for Advanced Imaging) for security applications-SC400AI, SC401AI, and SC433:

"The performance has been significantly improved compared with the previous generation products. It is equipped with Smartsens' innovative SFCPixel patented technology, which can achieve excellent night vision full-color imaging effects.

In addition, thanks to the PixGain dual-pixel conversion gain technology, SC401AI improves the image quality while further expanding the applicability of terminal products. As a product of the same specification as SC401AI, SC400AI has a frame rate of up to 60fps and can support 30fps dynamic line overlap HDR (Staggered HDR) image output, which provides customers with more choices while imaging performance is upgraded.

Compared with the previous generation products, the full well electrons of SC400AI and SC401AI have increased by 48.6%, the dynamic range has increased by 3dB, and the QE (quantum efficiency) in the 520nm band has increased by 42%."