Image Sensors World        Go to the original article...

This entry covers the progression of products originally developed by Reticon through its acquisition by EG&G (retaining the Reticon name) then EG&G's acquisition of part of PerkinElmer and the PerkinElmer name (Reticon was still applied sporadically as a brand) and finally the spinoff of PerkinElmer's optoelectronics businesses as Excelitas. Throughout this process, many of the original Reticon products retained their original part numbers as can be seen in the data sheets. 

Four items of note:

1 - Documents issued under "EG&G Reticon" are in the Reticon folder since the Reticon identity was maintained.  There is not yet an EG&G folder but this will be added when the amorphous silicon flat-panel products developed by EG&G are covered. The EG&G Amorphous Silicon facility was in the same building as Reticon in Sunnyvale, California, but operated independently. That business was later sold by PerkinElmer to Varex, formerly Varian Medical Systems.

2 - Excelitas no longer sells any Reticon-originated products. These became difficult to source when the fab Reticon operated in its own building was closed in the late 90's. Moving from a 3 micron process to 180 nm was quite difficult, especially while the selection of fabs who could make CCDs was rapidly shrinking. Excelitas still sells products originally made by some of the other optoelectronics companies EG&G bought, like thermopile arrays.  Those are included in the Excelitas folder but their history will be told later.

3 - Reticon, up through the PerkinElmer days, made many custom devices including a 20,000 pixel long three-color TDI sensor that flew on the U2 and an ultraviolet CCD for i-line semiconductor mask inspection. If the specifications or evaluation reports for any of these turn up, they will be posted.

4 - Reticon started by making bucket-brigade devices to be used as delay lines and switched-capacitor filter chips. These aren't imagers but they are in the folder to provide some background.

Reticon & EG&G Reticon document archive

PerkinElmer document archive

Excelitas document archive

Return to the Documentation List

Go to the original article...

Image Sensors World        Go to the original article...

Link: https://issw2024.fbk.eu/cfp

International SPAD Sensor Workshop (ISSW 2024) will be organized by Fondazione Bruno Kessler - FBK.
When: June 4-6, 2024
Location: Trento, Italy

Call for Papers & Posters

The 2024 International SPAD Sensor Workshop (ISSW) is a biennial event focusing on Single-Photon Avalanche Diodes (SPAD), SPAD-based sensors and related applications. The workshop welcomes all researchers (including PhDs, postdocs, and early-career researchers), practitioners, and educators interested in these topics.
 
After two on-line editions, the fourth edition of the workshop will return to an in-person only format.
The event will take place in the city of Trento, in northern Italy, hosted at Fondazione Bruno Kessler, in a venue suited to encourage interaction and a shared experience among the attendees.

The workshop will follow a 1-day long introductory school on SPAD sensor technology, which will be held in the same venue as the workshop on June 3rd, 2024.
 
The workshop will include a mix of invited talks and, for the first time, peer-reviewed contributions.
Accepted works will be published on the International Image Sensor Society website (https://imagesensors.org/).

Submitted works may cover any of the aspects of SPAD technology, including device modelling, engineering and fabrication, SPAD characterization and measurements, pixel and sensor architectures and designs, and SPAD applications.
 
Topics
Papers on the following SPAD-related topics are solicited:
● CMOS/CMOS-compatible technologies
● SiPMs
● III-V, Ge-on-Si
● Modelling
● Quenching and front-end circuits
● Architectures
● Time-to-Digital Converters
● Smart histogramming techniques
● Applications of SPAD arrays, such as:
o Depth sensing / ToF / LiDAR
o Time-resolved imaging
o Low-light imaging
o High dynamic range imaging
o Biophotonics
o Computational imaging
o Quantum imaging
o Quantum RNG
o High energy physics
o Free space communication
● Emerging technologies & applications
 
Paper submission
Workshop proposals must be submitted online. A link will be soon made available.
 
Each submission should consist of a 100-word abstract, and a camera-ready manuscript of 2-to-3 pages (including figures), and include authors’ name(s) and affiliation, short bio & picture, mailing address of the presenter, telephone, and e-mail address of the presenter. A template will be provided soon.
The deadline for paper submission is 23:59 CET, Friday December 8th, 2023.
 
Papers will be considered on the basis of originality and quality. High quality papers on work in progress are also welcome. Papers will be reviewed confidentially by the Technical Program Committee.

Accepted papers will be made freely available for download from the International Image Sensor Society website. Please note that no major modifications are allowed.

Authors will be notified of the acceptance of their abstract & posters at the latest by Wednesday Jan 31st, 2024.
 
Poster submission
In addition to talks, we wish to offer all graduate students, post-docs, and early-career researchers an opportunity to present a poster on their research projects or other research relevant to the workshop topics .

If you wish to take up this opportunity, please submit a 1-page description (including figures) of the proposed research activity, along with authors’ name(s) and affiliation, mailing address, telephone, and e-mail address.

The deadline for poster submission is 23:59 CET, Friday December 8th, 2023.

Go to the original article...

Image Sensors World        Go to the original article...

 Rivian

Senior Camera Tuning & Image Quality Engineer	Palo Alto, California, USA	Link
Sr. Sensor Engineer, Autonomy	Palo Alto, California, USA	Link
Camera Imaging Software Engineer - Vehicle Software	Belgrade, Serbia	Link

 OmniVision

Analog Design and Characterization Engineer	Oslo, Norway	Link
Analog engineer: Senior staff engineer / Staff engineer	Shin-Yokohama, Japan	Link
SoC Design Engineer	Santa Clara, California, USA	Link
Sr. Image Camera Simulator Engineer	Santa Clara, California, USA	Link

Postdoc Positions

Postdoctoral Research Associate in Radiation Detector Characterization and Modeling (US citizen)	Los Alamos National Laboratory · Los Alamos, New Mexico, USA	Link
Postdoc: MDL- Microdevice And Sensor Systems	Jet Propulsion laboratory, Pasadena, California, USA	Link
Post-doctoral position in RoHS compliant colloidal quantum dot photodetectors and image sensors	ICFO The Institute of Photonic Sciences, Barcelona, Spain	Link

New Listings

Apple Camera Electrical Engineer

San Diego, California, USA

Link

 

 

Go to the original article...

Image Sensors World        Go to the original article...

XDEP - X-ray Detector Technologies for Physics - 5-6 Feb 2024 - Saint-Aubin, France - Website

IEEE International Solid-State Circuits Conference (ISSCC) - 18-22 Feb 2024 - San Francisco, California, USA - Website

SPIE Medical Imaging - 18-24- Feb 2024 - San Diego, California, USA - Website

innoLAE (Innovations in Large-Area Electronics) - 20-22 Feb 2024 - Cambridge, UK - Website

16th Terascale Detector Workshop - 21-23 Feb 2024 -  Mainz, Germany  - Website

MSS Detectors and Materials Conference (clearance may be required) - 26 Feb-1 Mar 2024 - Orlando, Florida, USA - Website

Return to Conference List Index  

Go to the original article...

Image Sensors World        Go to the original article...

Tektronix and SITe are posted together because one transformed into the other carrying along the existing fab, products and people. Tektronix (yes, the oscilloscope company) originally worked in CCDs to produce a delay line use to convert a fast data stream into a slower one for an oscilloscope designed to capture high-speed transients. Tek decided to use its CCD technology to make large sensors for astronomy and medicine. Making imaging devices was an odd business for them so they sold it to an investment company which, in October 1993, formed Scientific Imaging Technologies. SITe supplied big CCDs into the early 2000s.

Tektronix document archive

SITe document archive

Go to the original article...

Image Sensors World        Go to the original article...

A team from Military University of Technology (Poland) and Shanghai Institute of Technical Physics (China) have published a review article titled "Infrared avalanche photodiodes from bulk to 2D materials" in Light: Science & Applications journal.

Open access paper: https://www.nature.com/articles/s41377-023-01259-3

Abstract: Avalanche photodiodes (APDs) have drawn huge interest in recent years and have been extensively used in a range of fields including the most important one—optical communication systems due to their time responses and high sensitivities. This article shows the evolution and the recent development of AIIIBV, AIIBVI, and potential alternatives to formerly mentioned—“third wave” superlattices (SL) and two-dimensional (2D) materials infrared (IR) APDs. In the beginning, the APDs fundamental operating principle is demonstrated together with progress in architecture. It is shown that the APDs evolution has moved the device’s performance towards higher bandwidths, lower noise, and higher gain-bandwidth products. The material properties to reach both high gain and low excess noise for devices operating in different wavelength ranges were also considered showing the future progress and the research direction. More attention was paid to advances in AIIIBV APDs, such as AlInAsSb, which may be used in future optical communications, type-II superlattice (T2SLs, “Ga-based” and “Ga-free”), and 2D materials-based IR APDs. The latter—atomically thin 2D materials exhibit huge potential in APDs and could be considered as an alternative material to the well-known, sophisticated, and developed AIIIBV APD technologies to include single-photon detection mode. That is related to the fact that conventional bulk materials APDs’ performance is restricted by reasonably high dark currents. One approach to resolve that problem seems to be implementing low-dimensional materials and structures as the APDs’ active regions. The Schottky barrier and atomic level thicknesses lead to the 2D APD dark current significant suppression. What is more, APDs can operate within visible (VIS), near-infrared (NIR)/mid-wavelength infrared range (MWIR), with a responsivity ~80 A/W, external quantum efficiency ~24.8%, gain ~105 for MWIR [wavelength, λ = 4 μm, temperature, T = 10–180 K, Black Phosphorous (BP)/InSe APD]. It is believed that the 2D APD could prove themselves to be an alternative providing a viable method for device fabrication with simultaneous high-performance—sensitivity and low excess noise.

Fig. 1: Bulk to low-dimensional material, tactics to fabricate APDs and possible applications: FOC, FSO, LIDAR and QKDs.

Fig. 2: The APD’s operating principle. a Electron and hole multiplication mechanisms, schematic of multiplication mechanism for b k = 0 (αh = 0) and c k = 1 (αe = αh), where k = αh/αe – αe, αh represent electron and hole multiplication coefficients. d αe, αh ionization coefficients versus electric field for selected semiconductors used for APDs’ fabrication

Fig. 3: APDs. a p–n device, b SAM device, and c SAGCM device with electric field distribution. F(M) dependence on M for the selected k = αh/αe in APDs when: d electrons and e holes dominate in the avalanche mechanism. The multiplication path length probability distribution functions in the: f local and g non-local field “dead space” models

Fig. 4: InGaAs/InP SAM-APD. a device structure, b energy band profile, and electric field under normal reverse bias condition. AlxIn1–xAsySb1–y based SACM APD: c detector’s design with the E distribution within the detector, d measured and theoretically simulated gain, dark current, photocurrent versus reverse voltage for 90 μm diameter device at room temperature39. InAs planar avalanche photodiode: e a schematic design diagram, f comparison of the gain reached by 1550 nm wavelength laser132,133. The M normalized dark current for 100 μm radius planar APD was presented for 200 K

Fig. 5: F(M) versus M for. a Si, AlInAs, GaAs, Ge, InP [the solid lines present the F(M) for k within the range 0–1 (increment 0.1) calculated by the local field model24, typical F(M) are shown by shaded regions37 and b selected materials: 3.5 μm thick intrinsic InAs APDs (50 μm and 100 μm radius), 4.2 μm cut-off wavelengths HgCdTe and 2.2 μm InAlAs APDs

Fig. 6: Gain and k versus Hg1–xCdxTe bandgap energy. a the crossover between e-APD and h-APD. The crossover at Eg ≈ 0.65 eV corresponds to the λc = 1.9 μm for 300 K46. Hole-initiated avalanche HgCdTe photodiode: b detector profile, c energy band structure, d hole-initiated multiplication process energy band structure. The multiplication layer bandgap energy is adjusted to the resonance condition where the bandgap and the split-off valence band energy and the top of the heavy-hole valence band energy difference are equal. Electron-initiated avalanche HgCdTe photodiode: e diagram of electron-initiated avalanche process for HgCdTe-based high-density vertically integrated photodiode (HDVIP) structure (n-type central region and p-type material around), f electron avalanche mechanism, and g relative spectral response for 5.1 μm cut-off wavelength HgCdTe HDVIP at T = 80 K

 

Fig. 7: HgCdTe APDs performance. a the experimental gain versus bias for selected cut-off wavelengths for DRS electron-initiated APDs at 77 K together with extra measured data points taken at ∼77 K51 and LETI e-APDs at 80 K59, b constant F(M) ~ 1 versus M at 80 K for 4.3 μm cut-off wavelength APD135

Fig. 8: The device structure comparison between low-noise PMT and multi-quantum well APDs. a schematic presentation of a photomultiplier tube, b multi-quantum well p-i-n APD energy band sketch with marked intrinsic region (i), c energy band profiles of staircase APD under zero (top) and reverse (bottom) voltage. Multistep AlInAsSb staircase avalanche photodiode: d 3-step staircase APD device profile, e theoretically calculated by Monte Carlo method and measured gain of 1-, 2-, and 3-stairs APDs for 300 K70. MWIR SAM-APD structure with AlAsSb/GaSb superlattice: f device design profile, g energy band structure under reverse voltage, and h carriers impact multiplication coefficients versus reciprocal electric field at 200 K

Fig. 9: Low-dimensional solid avalanche photodetectors. a graphite/InSe Schottky avalanche detector - injection, ionization, collection electron transport mechanisms, b e-ph scattering dimensionality reduction affects electron acceleration process and gain versus electric field in 2D (red line) and 3D (blue line), c breakdown voltage (Vbd) and gain as a function of temperature—exhibits a negative temperature coefficient81. Nanoscale vertical InSe/BP heterostructures ballistic avalanche photodetector: d schematic of the graphene/BP/metal avalanche device83, e ballistic avalanche photodetector operating principle, f quasi-periodic current oscillations, g schematic of the graphene InSe/BP83, h Ids–Vds characteristics for selected temperatures (40 − 180 K), i avalanche breakdown threshold voltage (Vth) and gain versus temperature—showing a negative temperature coefficient. Pristine PN junction avalanche photodetector: j device structure, k as the number of layers increases, a positive/negative signal of SCM denotes hole/electron carries, l APD’s low temperature (~100 K) dark and photocurrent I–V curves

Fig. 10: An idea of laser-gated system connected with passive thermal imaging for enhanced distant identification. a operation principle [at t0—camera is closed—light pulse is emitted, at t1—target reflects light pulse, at t2—the camera is opened for a short period (∆t) matching the needed depth of view]; b typical images of wide FOV thermal and laser-gating systems

Go to the original article...

Image Sensors World        Go to the original article...

 Alphacore

Sr Design Engineer

Tempe, Arizona, USA

Link

Tesla Motors

Image Scientist, Camera Technology	Palo Alto, California, USA	Link
Supplier Industrialization Engineer, Camera Top Level Assembly	Palo Alto, California, USA	Link
Sr. Process Engineer, Vision Automation	Austin, Texas, USA	Link
Internship, Machine Vision, Cell Manufacturing (Spring/Summer 2024)	Palo Alto, California, USA	Link

TSMC

CMOS Image Sensor Analog Design Engineer	Taiwan	Link
CIS Technology Program-Process Integration Engineer	Taiwan	Link

ams OSRAM

Senior Engineering Manager	Valencia, Spain	Link
Digital Verification Engineer	Garching, Germany	Link

GlobalFoundries

MTS Silicon Photonics Integration Engineer	Malta, New York, USA	Link
Si Photonic Pipeline Engineer	Malta, New York, USA	Link

Sony

Imaging Field Applications Engineer	Weybridge, Surrey, UK	Link
Digital Physical Implementation Engineer	Trento, Trentino, Italy	Link

Go to the original article...

Image Sensors World        Go to the original article...

IS&T Electronic Imaging - 21-25 Jan 2024 - Burlingame, CA, USA - Website

IEEE Applied Sensing Conference - 22-24 Jan 2024 - Goa, India - Website

SPIE Photonics West - 27 Jan-1 Feb 2023 - San Francisco, CA, USA - Website

Return to Conference List Index

Go to the original article...

Image Sensors World        Go to the original article...

 IEEE International Electron Devices Meeting - 9-13 Dec 2023 - San Francisco, CA, USA - Website

 Return to Conference List index

Go to the original article...

Image Sensors World        Go to the original article...

Now for a third company with unique technology out of the sensor business - Foveon. For those of you unfamiliar with Foveon, they started out to build three-chip, prism-based color cameras with custom CMOS sensors.  When this market was discovered to be way too small, they acquired Dick Merrill and his inventions from National Semiconductor and developed sensors with stacked RGB photodiodes in which the silicon itself provided color separation.  These were intended for DSLRs, then point-and-shoot cameras and then mobile phones. Didn't happen so, eventually, their only photographic customer, Sigma Photo, bought the assets for about 10 cents on the dollar and moved the work to Japan. Full disclosure - I sold Foveon sensors in the non-photo markets for about 10 years. Aside from Sigma, the only enduring legacy is the F13 on the ESA ExoMARS rover (now scheduled for an October 2028 launch).

Link to the Foveon folder

Return to the Documentation List

Go to the original article...

Image Sensors World        Go to the original article...

Hadfield et al. recently published a review paper titled "Single-photon detection for long-range imaging and sensing" in Optica:

Abstract: Single-photon detectors with picosecond timing resolution have advanced rapidly in the past decade. This has spurred progress in time-correlated single-photon counting applications, from quantum optics to life sciences and remote sensing. A variety of advanced optoelectronic device architectures offer not only high-performance single-pixel devices but also the ability to scale up to detector arrays and extend single-photon sensitivity into the short-wave infrared and beyond. The advent of single-photon focal plane arrays is poised to revolutionize infrared imaging and sensing. In this mini-review, we set out performance metrics for single-photon detection, assess the requirements of single-photon light detection and ranging, and survey the state of the art and prospects for new developments across semiconductor and superconducting single-photon detection technologies. Our goal is to capture a snapshot of a rapidly developing landscape of photonic technology and forecast future trends and opportunities.

 

Fig. 1. Examples of imaging LIDAR configurations. (a) Flash LIDAR configuration using an array sensor and full-field illumination (a bistatic system is shown, with source and sensor separated). (b) Scanning LIDAR approach where the source is scanned and an individual sensor is used. (In this illustration, a bistatic configuration is shown; however, a monostatic scanning configuration is often used with a common transmit and receive axis).

 

Fig. 2. Single-photon LIDAR depth profiles taken at a range of greater than 600 m using a 100-channel Si SPAD detector system in scanning configuration. The operational wavelength is 532 nm. (a) Visible-band photograph of scene. (b) Reconstructed depth image of the city scene. (c) Detailed depth profile of the subsection of the scene within the red rectangle in (a). Further details in Z. Li et al. [60]. Figure reproduced with permission of Optica Publishing Group.

Fig. 3. Example of data fusion of a 3D image from a CMOS SPAD detector array and passive imagery of a scene at 150 m range. (a) Retrieved depth information from a SPAD detector array. (b) Intensity information from the SPAD overlaid on top of the retrieved depth information. (c) Intensity information from a color camera overlaid on top of the retrieved depth information [65]. Figure reproduced with permission of Springer Nature publishing.

Fig. 4. Solar irradiance versus wavelength at sea level (red) and in the upper atmosphere (blue). MODTRAN simulation [86]. The following spectral bands beyond the visible wavelength range are denoted by the shaded regions: near infrared (NIR), yellow; short-ware infrared (SWIR), cyan; mid-wave infrared (MWIR), red.

Fig. 5. Example of scanning SWIR single-photon LIDAR imaging. (a) Visible-band image of a residential building taken with an f=200mm camera lens. (b) Depth intensity plot of the building imaged with 32×32 scan points over a range of 8.8 km. (c) Depth plot of the building imaged with 32×32 scan points over a range of 8.8 km; side view of the target [89]. Figure reproduced with permission of Optica Publishing Group.

Fig. 6. Reconstruction results of a mountain scene over a range of 201.5 km using SWIR single-photon LIDAR [91]. (a) Visible-band imaged photograph. (b) Reconstructed depth result using algorithm by Lindell et al. [92] for data with signal-to-background ratio ∼0.04 and mean signal photon per pixel ∼3.58. (c) 3D profile of the reconstructed result. Figure reproduced with permission of Optica Publishing Group.

Fig. 7. Analysis of a scene with an actor holding a wooden plank across his chest and standing 1 m behind camouflage netting at a range of 230 m in daylight conditions. (a) Photograph of the scene, showing the actor holding a wooden plank behind the camouflage. (b), (c) Intensity and depth profiles of the target scene using all the collected single-photon LIDAR data. (d), (e) Intensity and depth profiles after time gating to exclude all data except those with a 0.6 m range around the target location. The pixel format used in the depth and intensity profiles is 80×160 [95]. Figure reproduced with permission of SPIE publishing.

Fig. 8. Schematic diagram of a SWIR single-photon 3D flash imaging experiment. The scene consists of two people walking behind a camouflage net at a stand-off distance of 320 m from the LIDAR system. An RGB camera was positioned a few meters from the 3D scene and used to acquire a reference video. The proposed algorithm is able to provide real-time 3D reconstructions using a graphics processing unit (GPU). As the LIDAR presents only 32×32 pixels, the point cloud was estimated in a higher resolution of 96×96 pixels. The acquired movie is shown in [101]. Figure reproduced with permission of Springer Nature publishing.

Fig. 9. Single-photon detector technologies for infrared single-photon LIDAR, with spectral coverage for each detector type indicated. (a) Schematic diagram cross section of a Si-based SPAD detector. The design is a homojunction. (b) Schematic diagram cross section of a Ge-on-Si structure, illustrating optical absorption in the Ge layer, and multiplication in the intrinsic Si layer. (c) Schematic diagram cross section of an InGaAs/InP SPAD detector; the absorption is in the narrow-gap InGaAs and the multiplication in the wider gap InP layer. In both (b) and (c), the charge sheet is used to alter the relative electric fields in the absorption and multiplication layers. (d) Schematic illustration of SNSPD architecture for near-unity efficiency at 1550 nm wavelength and optical micrograph of chip with single-pixel detector [109]; (d) reproduced with permission of Optica Publishing Group.

 

Link to paper (open access): https://opg.optica.org/optica/abstract.cfm?URI=optica-10-9-1124

Go to the original article...

Image Sensors World        Go to the original article...

The first article in the Sensors special issue for IISW2023 is now available:

https://www.mdpi.com/1424-8220/23/18/7959

Chao et al. from TSMC in a paper titled "Random Telegraph Noise Degradation Caused by Hot Carrier Injection in a 0.8 μm-Pitch 8.3Mpixel Stacked CMOS Image Sensor" write:

In this work, the degradation of the random telegraph noise (RTN) and the threshold voltage (Vt) shift of an 8.3Mpixel stacked CMOS image sensor (CIS) under hot carrier injection (HCI) stress are investigated. We report for the first time the significant statistical differences between these two device aging phenomena. The Vt shift is relatively uniform among all the devices and gradually evolves over time. By contrast, the RTN degradation is evidently abrupt and random in nature and only happens to a small percentage of devices. The generation of new RTN traps by HCI during times of stress is demonstrated both statistically and on the individual device level. An improved method is developed to identify RTN devices with degenerate amplitude histograms.

 

Figure 1. Simplified test chip architecture. The device under stress is the source follower (SF) NMOS in the 4 × 2-shared pixels on the top layer. The PD0–7 are the photodiodes, and the TG0–7 are the transfer gates in each 4 × 2-shared pixel. The total number of SF is 628 × 1648 = 1.03 M.

Figure 2. (a) The measured IB of a SF device vs. VD with VG stepping from 1.3 V to 2.8 V; (b) The same data as in (a) but plotted against VDS−VDsat≈VD−VG+Vt with Vt as a fitting parameter; (c) The same data as in (b) plotted against 1/(VDS−VDsat) with P=(P1,P2) as two fitting parameters according to Equation (1).

Figure 3. The bias configuration of the SF under test. The red and blue solid circles symbolize electrons and holes, respectively.

Figure 4. The histograms of the measured VGS of the SF for stress time (t) from 0 to 100 min.

Figure 5. (a) The histograms of the threshold voltage shift (ΔVt) after 10-, 20-, 50-, and 100-min stress; (b) The inverse cumulative distribution function (ICDF) curves of ΔVt; (c) the constant ICDF contours against stress time (t).

Figure 6. (a) The histograms of the random noise changes (ΔRN) after 10, 20, 50, 100 min stress; (b) The inverse cumulative distribution function (ICDF) curves; (c) the constant ICDF contours as functions of stress time (t).

Figure 7. The correlation of the SF threshold voltage shift (ΔVt) after 10 min of HCI stress vs. after (a) 20 min, (b) 50 min, and (c) 100 min of stress, respectively. The linear least-square fit of the x/y ratio (red dash line) shows the continuous increase of the ΔVt as the stress time increases. The ΔVt increases are relatively uniform among all 1M devices, which is quite different from the random noise increases in Figure 8 below. Random colors are assigned to the data points to separate the dots from each other.

Figure 8. The correlation of the random noises (RN) before HCI stress (t = 0) vs. after (a) 10 min, (b) 20 min, and (c) 100 min stress, respectively The RN increases are noticeably nonuniform. The RN along the x = y red dash line remains relatively unchanged. The devices on the lower-right branches show a significant increase in RN. The population of the lower branch increases as stress time increases. Random colors are assigned to the data points to separate the dots from each other.

Figure 9. The 2D histograms of the correlation of the Vt shift and RN degradation shows dramatically different statistical behaviors. (a) The Vt change after 100-min stress versus that after 10-min stress. (b) The RN after 100 min stress versus that before the stress.

Figure 10. Generation of RTN traps during HCI stress. The 5000-frame waveforms before (t = 0) and after the HCI stress (t = 20, 100 min) with the corresponding histograms are shown for three selected examples. (a) Device (296, 137) shows one trap before stress and remains unchanged after stress. (b) Device (202, 1338) shows no trap before stress and one trap generated after 20 min of stress. (c) Device (400, 816) shows no trap before stress; however, one trap is generated after 100 min of stress. The RN unit is mV-rms.

Figure 11. Degeneration of the RTN discrete levels. During HCI stress, the non-RTN noises may be increased significantly such that the discrete RTN levels become indistinguishable. (a) Device (141, 1393) show such degeneration after 100 min of stress. (b) Device (481, 405) show degeneration after 20 min of stress. (c) Device (519, 1638) shows unsymmetric side peaks and unsymmetric degeneration after 20 min and 100 min of stresses. The RN unit is mV-rms.

Figure 12. For devices showing a single histogram peak, if the histogram is significantly different from the Gaussian distribution, they are counted as RTN-like devices. The ratio R expressed in Equation (2) is defined as the red area versus the total area under the black histogram. The R values in examples (a) and (b) are 36% and 28%, respectively. The RN unit is mV-rms.

Figure 13. Devices with amplitude distributions close to Gaussian are considered as non-RTN devices. The deviation ratio R is 7% for device (587, 492) in (a) and 9% for device (124, 1349) in (b). The RN unit is mV-rms.

Figure 14. The RN distribution of the RTN and non-RTN devices, sorted by the improved algorithm: (a) before HCI stress, (b) after 20 min stress, and (c) after 100 min stress. The RTN devices clearly contribute to and dominate the long tails of the RN histograms. The number of RTN devices (Nx) (with the R-threshold set to 15%) increases systematically as the stress time increases.

Figure 15. The count of RTN devices increases consistently as stress time increases. N2 is the number of devices showing two or more peaks in amplitude histograms. Nx is N2 plus the number of RTN-like devices determined by setting the R-threshold to 10%, 15%, and 20%, respectively.

Figure 16. (a) The Vt shift and (b) the RN degradation trends against the effective stress defined in Equation (3), where the effectiveness factors are treated as empirical fitting parameters such that all the constant-ICDF points for different voltages fall onto a family of continuous and smooth curves. The fitting results are listed in Table 1.

Go to the original article...

Image Sensors World        Go to the original article...

The first article in the Sensors special issue for IISW2023 is now available:

https://www.mdpi.com/1424-8220/23/18/7959

Chao et al. from TSMC in a paper titled "Random Telegraph Noise Degradation Caused by Hot Carrier Injection in a 0.8 μm-Pitch 8.3Mpixel Stacked CMOS Image Sensor" write:

In this work, the degradation of the random telegraph noise (RTN) and the threshold voltage (Vt) shift of an 8.3Mpixel stacked CMOS image sensor (CIS) under hot carrier injection (HCI) stress are investigated. We report for the first time the significant statistical differences between these two device aging phenomena. The Vt shift is relatively uniform among all the devices and gradually evolves over time. By contrast, the RTN degradation is evidently abrupt and random in nature and only happens to a small percentage of devices. The generation of new RTN traps by HCI during times of stress is demonstrated both statistically and on the individual device level. An improved method is developed to identify RTN devices with degenerate amplitude histograms.

 

Figure 1. Simplified test chip architecture. The device under stress is the source follower (SF) NMOS in the 4 × 2-shared pixels on the top layer. The PD0–7 are the photodiodes, and the TG0–7 are the transfer gates in each 4 × 2-shared pixel. The total number of SF is 628 × 1648 = 1.03 M.

Figure 2. (a) The measured IB of a SF device vs. VD with VG stepping from 1.3 V to 2.8 V; (b) The same data as in (a) but plotted against VDS−VDsat≈VD−VG+Vt with Vt as a fitting parameter; (c) The same data as in (b) plotted against 1/(VDS−VDsat) with P=(P1,P2) as two fitting parameters according to Equation (1).

Figure 3. The bias configuration of the SF under test. The red and blue solid circles symbolize electrons and holes, respectively.

Figure 4. The histograms of the measured VGS of the SF for stress time (t) from 0 to 100 min.

Figure 5. (a) The histograms of the threshold voltage shift (ΔVt) after 10-, 20-, 50-, and 100-min stress; (b) The inverse cumulative distribution function (ICDF) curves of ΔVt; (c) the constant ICDF contours against stress time (t).

Figure 6. (a) The histograms of the random noise changes (ΔRN) after 10, 20, 50, 100 min stress; (b) The inverse cumulative distribution function (ICDF) curves; (c) the constant ICDF contours as functions of stress time (t).

Figure 7. The correlation of the SF threshold voltage shift (ΔVt) after 10 min of HCI stress vs. after (a) 20 min, (b) 50 min, and (c) 100 min of stress, respectively. The linear least-square fit of the x/y ratio (red dash line) shows the continuous increase of the ΔVt as the stress time increases. The ΔVt increases are relatively uniform among all 1M devices, which is quite different from the random noise increases in Figure 8 below. Random colors are assigned to the data points to separate the dots from each other.

Figure 8. The correlation of the random noises (RN) before HCI stress (t = 0) vs. after (a) 10 min, (b) 20 min, and (c) 100 min stress, respectively The RN increases are noticeably nonuniform. The RN along the x = y red dash line remains relatively unchanged. The devices on the lower-right branches show a significant increase in RN. The population of the lower branch increases as stress time increases. Random colors are assigned to the data points to separate the dots from each other.

Figure 9. The 2D histograms of the correlation of the Vt shift and RN degradation shows dramatically different statistical behaviors. (a) The Vt change after 100-min stress versus that after 10-min stress. (b) The RN after 100 min stress versus that before the stress.

Figure 10. Generation of RTN traps during HCI stress. The 5000-frame waveforms before (t = 0) and after the HCI stress (t = 20, 100 min) with the corresponding histograms are shown for three selected examples. (a) Device (296, 137) shows one trap before stress and remains unchanged after stress. (b) Device (202, 1338) shows no trap before stress and one trap generated after 20 min of stress. (c) Device (400, 816) shows no trap before stress; however, one trap is generated after 100 min of stress. The RN unit is mV-rms.

Figure 11. Degeneration of the RTN discrete levels. During HCI stress, the non-RTN noises may be increased significantly such that the discrete RTN levels become indistinguishable. (a) Device (141, 1393) show such degeneration after 100 min of stress. (b) Device (481, 405) show degeneration after 20 min of stress. (c) Device (519, 1638) shows unsymmetric side peaks and unsymmetric degeneration after 20 min and 100 min of stresses. The RN unit is mV-rms.

Figure 12. For devices showing a single histogram peak, if the histogram is significantly different from the Gaussian distribution, they are counted as RTN-like devices. The ratio R expressed in Equation (2) is defined as the red area versus the total area under the black histogram. The R values in examples (a) and (b) are 36% and 28%, respectively. The RN unit is mV-rms.

Figure 13. Devices with amplitude distributions close to Gaussian are considered as non-RTN devices. The deviation ratio R is 7% for device (587, 492) in (a) and 9% for device (124, 1349) in (b). The RN unit is mV-rms.

Figure 14. The RN distribution of the RTN and non-RTN devices, sorted by the improved algorithm: (a) before HCI stress, (b) after 20 min stress, and (c) after 100 min stress. The RTN devices clearly contribute to and dominate the long tails of the RN histograms. The number of RTN devices (Nx) (with the R-threshold set to 15%) increases systematically as the stress time increases.

Figure 15. The count of RTN devices increases consistently as stress time increases. N2 is the number of devices showing two or more peaks in amplitude histograms. Nx is N2 plus the number of RTN-like devices determined by setting the R-threshold to 10%, 15%, and 20%, respectively.

Figure 16. (a) The Vt shift and (b) the RN degradation trends against the effective stress defined in Equation (3), where the effectiveness factors are treated as empirical fitting parameters such that all the constant-ICDF points for different voltages fall onto a family of continuous and smooth curves. The fitting results are listed in Table 1.

Go to the original article...

Image Sensors World        Go to the original article...

 ISW is building a new comprehensive Image Sensor Industry List. Click for more details.

Go to the original article...

Image Sensors World        Go to the original article...

I apologize to those of you who tried to read the Pixim data sheets.  I forgot to set the share flag on the folder.  The files are available for viewing by anyone now.

Go to the original article...

Image Sensors World        Go to the original article...

Teledyne

CMOS Sensor Product Support	Waterloo, Ontario, Canada	Link
Senior Manager, Digital & Analog Mixed Signal IC Design	Camarillo, California, USA	Link
MBE Growth Production Engineer (US Citizen)	Camarillo, California, USA	Link
ASIC Design Engineer (US Citizen or equivalent)	Goleta, California, USA	Link
Director - Market Development	Grenoble, France	Link
Senior Scientist (US Citizen)	Acton, Massachusetts, USA	Link

Lumotive

Optoelectronics Systems Engineer	Redmond, Washington, USA	Link
Optoelectronics Systems Engineer	Vancouver, British Columbia, Canada	Link

University of Arizona

Optical Sciences Postdoctoral Research Associate I	Tucson, Arizona, USA	Link
Physics Postdoctoral Research Associate I	Tucson, Arizona, USA	Link
Optical Sciences Postdoctoral Research Associate	Tucson, Arizona, USA	Link

Sandia National Laboratories

Nanophotonics - Postdoctoral Appointee

Albuquerque, New Mexico, USA

Link

Santa Clara University

Assistant Professor, Electrical and Computer Engineering (Tenure-track)

Santa Clara, California, USA

Link

Go to the original article...

Image Sensors World        Go to the original article...

IEEE Nuclear Science Symposium and Medical Imaging Conference - 4-11 Nov 2023 - Vancouver, British Columbia, Canada  - Website

Semi MEMS and Sensors Executive Conference - 6-8 Nov 2023 - Phoenix, Arizona, USA - Website

Coordinating Panel for Advanced Detectors Workshop - 7-10 Nov 2023 - Menlo Park, California, USA - Website

Compamed - 13-16 Nov 2023 - Dusseldorf, Germany - Website

Fraunhofer IMS 10th CMOS Imaging Workshop -  21-22 Nov 2023 - Duisburg, Germany - Website

RSNA 109th Scientific Assembly and Annual Meeting  - 26-30 Nov 2023 - Chicago, Illinois, USA - Website

Return to Conference List index 

Go to the original article...

Image Sensors World        Go to the original article...

Since someone just pointed out that the Pixim weblink in the ISW vendor list was dead, I decided to put up Pixim datasheets next.  These are different from most because Pixim sold a chipset that included the sensor and a matching processor to turn the chip outputs (basically timing data) into digital image data.  Pixim also made several demonstration cameras. Sony bought Pixim in 2012 and stopped making Pixim-type chips a couple of years later because they felt they had a better way to do HDR imaging. The debate continues.

Link to the Pixim folder

Return to the Documentation List

Go to the original article...

Image Sensors World        Go to the original article...

OMNIVISION Announces New Low-power, Enhanced-performance 2MP Image Sensor for Security Surveillance Cameras
 
The OS02N features a 2.5-micron enhanced-performance FSI pixel with on-sensor DPC for higher sensitivity, performance and reliability while remaining cost-effective
 
SANTA CLARA, Calif. – September 27, 2023 – OMNIVISION, a leading global developer of semiconductor solutions, including advanced digital imaging, analog, and touch & display technology, today announced the new OS02N, a 2-megapixel (MP) frontside illumination (FSI) image sensor with optimized defective pixel correction (DPC) algorithm for higher sensitivity, improved performance and increased reliability for IP and HD analog security cameras, including professional surveillance and outdoor home security cameras. The OS02N supports always-on with its low-power capability.
 
“Customers need high-performing security cameras that produce sharp, high-resolution images with low power consumption for extended battery life. The OS02N meets these requirements and is also a cost-effective solution,” said Cheney Zhang, senior marketing manager, OMNIVISION. “The OS02N uses FSI technology, which has a large pixel size for better quantum efficiency and excellent signal-to-noise ratio, resulting in high sensitivity in low-light conditions and dramatically improved image quality and performance. It has a 1/3.27-inch optical format and is designed to be pin-to-pin compatible with our OS04L and OS04D image sensors.”
 
The OS02N features a 2.5-micron pixel based on OMNIVISION’s OmniPixel®3-HS technology. This enhanced-performance, cost-effective solution uses FSI technology for true-to-life color reproduction in both bright and dark conditions. Optimized DPC algorithm improves sensor quality and reliability above and beyond standard devices by providing real-time correction of defective pixels that can result throughout the sensor’s life cycle, especially in harsh operating conditions. The OS02N features 1920x1080 resolution at 30 frames per second (FPS).
 
The OS02N supports MIPI and DVP interfaces. It is sampling now and will be in mass production in Q1 2024. For more information, contact your OMNIVISION sales representative: www.ovt.com/contact-sales.

 

Go to the original article...

Image Sensors World        Go to the original article...

Sheba Microsystems Launches Revolutionary MEMS Autofocus Actuator for Active Athermalization in Embedded Vision Cameras

Breakthrough µPistons™ technology uniquely solves decades-long embedded vision camera industry’s problem of lens thermal expansion. Novel product unlocks unparalleled resolution and consistent high-quality imaging performance for automotive, action, drone, mobile robotics, security and surveillance, and machine vision cameras.

TORONTO--(BUSINESS WIRE)--Sheba Microsystems Inc., a global leader in MEMS technologies, today announced the launch of its revolutionary new product, the MEMS Autofocus Actuator for Active Athermalization in Embedded Vision Cameras used in automotive, action, drones, machine vision, security and surveillance, and mobile robotics.

The first-of-its-kind solution tackles the long-standing industry problem of embedded vision cameras’ inability to maintain image quality and focus stability during temperature fluctuations as optics undergo thermal expansion.

While smartphones use autofocus actuators and electromagnetic actuators including voice coil motors (VCMs), these actuators are unreliable for achieving active athermalization in embedded vision cameras due to extreme environmental conditions. Embedded vision camera optics are also 30 times larger than smartphone optics. Other autofocus systems in-market such as tunable lenses lack thermal stability and compromise optical quality.

“MEMS actuators are fast, precise, and small in size, and are actually uniquely suited to solve thermal expansion issues, because they are thermally stable and maintain consistent performance regardless of temperature changes,” said CEO and co-founder Dr. Faez Ba-Tis, PhD. “Because of these known advantages, there have been previous industry attempts at incorporating MEMS actuators into cameras, but because they failed drop tests they were quickly abandoned. Sheba’s new design solves for all of these previous blockers, which opens up limitless possibilities for embedded vision camera innovation.”

Sheba’s proprietary technology compensates for thermal expansion by uniquely moving the lightweight sensor, instead of moving the lenses. The silicon-based MEMS actuator platform actuates the image sensor along the optical axis to compensate for thermal expansion in the optics. The weight of the image sensor represents only 2-3 % of the optical lens weight, which makes it easier to handle, enabling ultra-fast and precise autofocus performance even when temperatures fluctuate.

Sheba’s novel piston-tube electrode configuration takes advantage of a larger capacitive area, allowing for substantial stroke and increased force. In contrast to traditional MEMS comb-drive electrode configuration, Sheba’s µPistons™ design makes the MEMS actuators uniquely resilient against severe shocks, since the electrodes are well-supported and interconnected with each other.

Sheba’s new MEMS actuator has successfully passed drop tests as well as other reliability tests, including thermal shock, thermal cycling, vibration, mechanical shock, drop, tumble, and microdrop tests. It is also highly rugged, which helps maintain image focus during high shocks in action cameras or machine vision environments.

“Digital camera technologies are increasingly used in almost every aspect of our lives,” said Ba-Tis. “From sharing photos of our travels in social media, to experiencing new artificial intelligence innovations powered by machine vision, and accelerating the deployment of autonomous vehicles in our communities, high quality images are imperative to not only capture our most memorable events, but to also keep us safe. In situations where split-second decisions are critical, image quality becomes paramount.”

Sheba’s MEMS actuator offers lens design flexibility and is suitable for near and far-field imaging. It is easily integrated into existing systems and scaled up on mass production tools for automotive, action, drone, mobile robotics, security and surveillance, and machine vision cameras.
Sheba is offering evaluation kits to interested customers, so they can test and evaluate the new product in their own labs to ensure the reliability of the technology. The kit includes camera samples, a daughter board with the MEMS driver, interposer, and camera test jig to perform mechanical reliability tests, software, and user manual.

To learn more about Sheba Microsystems or to order an evaluation kit for your organization, visit www.shebamicrosystems.ca.

 

Go to the original article...