2017년 4월 24일 월요일

CCD image sensor

CCD image sensor

CCD image sensor (sea sea D image sensor British: CCD image sensor) is a semiconductor element used in a video camera, a digital camera, a photosensor in one of the solid-state image sensing devices widely. Is often called merely CCD [1]; [2] [3] [4] [5].

Table of contents

Summary

History

CCD

Generally it is connected electrically, and, with the normal semiconductor integrated circuit, signals are exchanged between each element elaborated on the top by the wiring by the pattern of the metal layer. In contrast, using the electric charge-like combination between the element next to each other, the element that signals are exchanged by the state of the electric charge being sent off in sequence is Charge-Coupled Device (CCD, charge coupled device).

I assume it image pick-up device in combination with a light receiving element, and there is the various application including the storage device of the sequential access equivalent to the delay line in the analog filter with a necessary comb filter in NTSC. There are the weak points such as smear phenomena while there is the advantage that there is at a light receiving area widely not to have area stolen by wiring for image pick-up device.

CCD image sensor and CMOS image sensor

 
Example of the image sensor device

Major solid image pick-up device includes CMOS image sensor other than CCD image sensor now. A lot of CCD image sensor had been used for a long time, but the spread of CMOS sensors advanced with the low-priced article and cell-phone, and a CMOS sensor exceeded it by the total shipment number in late 2004.

The production of the CCD is more difficult than a CMOS sensor, and the company to have of the production is limited for more complicated structure. In addition, it becomes the semiconductor which is considerably more expensive than a CMOS sensor (the details mention it later).

Purpose of the image sensor

I let the light receiving plane of the imaging element image the rays from a subject by the optical systems such as lenses, and the function of imaging element electrically converts the light and shade by the light of the image to the quantity of the electric charge and is to begin to read it sequentially, and to convert it into an electrical signal.

In the solid-state image sensing device, I electrically convert it with the row of the large number of light receiving element formed on one piece of silicon substrate. I produce an electric charge by light energy when I irradiate a light receiving element with light. It is main movement to transfer this electric charge by a CCD element outside. In addition, I may operate CCD itself for the transfer as a light receiving element when I use independent photodiode as a light receiving element (later description).

A principle and structure

 
I control the voltage and transfer an electric charge

As for the CCD, the structure that approaches it, and displayed MOS capacitors is basic. The CCD established a large number of electrodes on the oxidation film of the silicon substrate surface with kind of the MOS structure semiconductor element and created potential well (English version) (well of the electric potential) by giving each electrode of the MOS structure the neighboring, different voltage and allowed maintenance to transfer an electric charge in semiconductors using this.

The electric charge of each element is transferred to the next element all at once by controlling the voltage in addition to each electrode appropriately. A bucket brigade type can in this way pick up the electric charge that each element maintains outside sequentially. It is the operation of the digital circuit same as parallel = serial number shift register and is just said to be the shift register who can treat analog quantity. If I give you a delay equivalent to the transfer number of times for the number of elements and take the electric charge that I input using this property from one line of edge out of the contralateral edge, I can operate it as delay line (derailleur in).

One-dimensional image sensor

I say the linear image sensor and arrange photodiode to one line and locate CCD in this in parallel. I transfer the electric charge that photodiode electrically converted by one exposure to a CCD element corresponding to each pixel all at once and I give CCD a transfer pulse successively and begin to read an electric charge sequentially. The next exposure is enabled when I finish outputting the electric charge of all pixels.

Because the one-dimensional image sensor can electrically convert only a linear image, I relatively move a subject and image sensor or cover the whole subject by relatively moving the same class by an optical system. It is a method used with a facsimile and a copier, image scanner. Movement time is second order and slow one, and the subject is almost limited to a still image.

Two-dimensional image sensor

I said the area image sensor and spread it all into a plane form by grillwork or other placement. There is some structure by the sequence of CCD and the light receiving element.

Inter-line type

 
Structure of the inter-line type Charge Coupled Device imager

Inter-line type CCD image sensor is used many with a general video camera and digital camera in two-dimensional image sensor. I locate the photodiode of the light receiving department and the perpendicular transfer CCD of the electric charge transfer department every one line in turn and am the structure that connects the edge of the perpendicular CCD line to each element of the stability transfer CCD, and, as a whole, located it in the pectinated form. The CCD for each transfer is covered with a shading film not to electrically convert it. In addition, a transfer gate acting on each element of the perpendicular CCD corresponding to photodiode and the pixel of each pixel as an analog switch between is put. I do the reading for one as follows and perform it.

  1. I close a transfer gate.
  2. I am exposed to light in photodiode and accumulate an electric charge.
  3. I open the transfer gate and transfer an electric charge to each perpendicular transfer CCD from photodiode all at once.
  4. I close a transfer gate.
  5. I transfer an electric charge of each perpendicular transfer CCD for once and transfer the electric charge of the pixel equal to the edge of each line to horizontal transfer CCD.
  6. I give horizontal transfer CCD a transfer pulse sequentially and output all horizontal pixels.
  7. It is repeated until I return to 5 and retrieve all pixels of the perpendicular transfer CCD.

The pixel of all all areas would be in this way scanned sequentially.

Because it is necessary to perform 30-60 times per second of exposure, transfer, reading to use it for a video camera, as for the electric charge transfer from photodiode to perpendicular CCD, it is performed again during each stability return wire period during a verticality return wire period beginning to read it from stability transfer CCD.

Other model

 
Structure of the full-frame transfer type Charge Coupled Device imager
 
Structure of the frame transfer type Charge Coupled Device imager

I have full-frame transfer type structure, CCD for the light receiving and CCD for the transfer electrically converting with each element of the perpendicular transfer CCD directly elsewhere, and there is the frame transfer type CCD transferring to CCD for the transfer from CCD for the light receiving during a verticality return wire period. The ratio (opening rate) of the element area that the light is because CCD for the transfer is unnecessary as for the full-frame type is big and is more sensitive than an inter-line type. On the other hand, the animation photography is difficult because a mechanical shutter becomes required not to receive light during electric charge transfer. The frame transfer type is similar to a full-frame type at the high point of the opening rate, but an element area grows big as much as I have CCD for the transfer particularly.

Because there is a reading circuit in the other side of the light receiving department, the backside-illumination type CCD can do a light receiving area widely and is more sensitive than the tip of the area. However, thermal capacity largely decreases by sharpening a silicon substrate layer for backside-illumination, and dropping it, and the fever by the dark current increases. Some kind of cooling means is necessary to suppress the influence of the heat noise by this; of the small size price reduction of the photography apparatus is a big hindrance.

History

I came up with the idea, and the CCD was invented by Willard boiling and George E Smith who were in American telegraphic communication telephone[6]Bell Laboratories in 1969 by a theory of the magnetic bubble memory by applying it to a semiconductor [7]. Fairchild semiconductor, RCA, various companies including Texas Instruments adopted the invention and began program development. Fairchild became the first company which I used to a commercial device. By 1974, it released the two-dimensional device of the 100x100 cell with the one-dimensional element of 500 cells. Cause, Sony of the leadership by Kazuo Iwama began a big development effort to CCD including the investment of the notable sum. Sony was able to mass-produce CCD for company's camcorders after all [8].

After invention in Bell Laboratories, an applied study as as delay line or the image sensor advanced, and the color television camera which was practical in the 1980s was manufactured, and a trial manufacture camera replaced it in an image pickup tube by progress of the semiconductor processing technique in the late 1970s. A digital still camera is born in the 1990s and is substituting with a silversalt camera in the early 21st century.

Work about the CCD was accepted, and, in January, 2006, boiling and Smith won the Charles Stark ドレイパー prize of the U.S. technical academy. [9] In 2009, two people won Nobel Prize in Physics. [10]

Production

The CCD image sensor is made from a silicon wafer like other semiconductor integrated circuits. I do not change with the production of other semiconductor integrated circuits basically, but the consideration to a light receiving side is demanded, and the normal moulding is maintained by the implementation to a substrate without being performed.

Applied

  • Digital camera
  • Video camera
  • Cell-phone with camera
  • Copier, facsimile
  • Gastrocamera (upper part gastrointestinal tract endoscope)

Color imaging with the CCD image sensor

 
Pixel placement of the LCD indicating element
 
Buyer sequence of the single plate-type color camera
 
Space placement of the CCD element of 3 boards type color camera
 
Version (green resolution priority) of 3 boards type

Like many other color photography methods, there is a multi-disc (3 boards) method to photograph a single plate method to elaborate each color of RGB to one piece of sensor and the thing which separated each color of RGB optically with each sensor.

Single plate method

CCD image sensor itself does not have the discriminability of the color like other image pickup tubes and solid-state image sensing devices. The sensitivity properties every wavelength of the general CCD (spectrum properties) have peak properties of the slow chevron at about 300nm - 800nm and cover a visible light range. Therefore, it is necessary to perform a color separation in the three primary colors of the light by a color filter to perform color photography [11]. The three primary colors resolution includes a thing by a method and the subtractive color process by the additive color process, and the simple additive color process of postprocessing be theoretically superior to color reproduction characteristics, and it is said that the subtractive color process that there is much light passage quantity is superior in sensitivity, but the difference is not remarkable by the comparison at the product level because there are the performance of the element level and the difference by the later color processing technology.

Some methods are thought about in the mask pattern of the color filter. At first, like indicating element such as the LCD, a method to place a light receiving department of each RGB for each pixel spread all over the square is thought about. A model and the consistency with the value of the color are good every each pixel such as the image data, but each light receiving department becomes slim and must elaborate microstructure with again triple pace and is not all right in production.

Generally, it does aspect ratio with 1, or it is desirable for each light receiving region that it is near because it puts microlens with the solid-state image sensing device every light receiving department. Therefore, the technique that I spread a square all and use a filter of the buyer [12] sequence (Bayer arrangement) is common.

With the buyer sequence, as for the green resolution, N/2, the red and blue resolution create groups of RGB of the N unit for total pixels N of the CCD to become N/4 by operating interpolation using the output of neighboring pixels every each pixel. Because a picture is affected by a method of the interpolation operation, each camera maker devises a processing method originally. The spectral sensitivity of human eyes assumes green neighborhood a peak, and the reason why I make a green pixel 2 times here is that green resolution improves apparent resolution.

In addition, there are super CCD honeycomb of Fuji Film and other patterns (reference: English version), a method of Foveon distinguishing a color by depth not a filter again.

Multi-disc method

I may adopt 3 board methods to meet demand performance of high resolution, the high sensitivity with the color television cameras for the broadcast. This prepares for one piece of CCD image sensor in each RGB color each with the thing like the method of the image pickup tube era that the method with the filter of the microstructure such as the statement above was impossible and I let you divide it into each color by ダイクロイックプリズム and expose to light and take a color signal of RGB each. In 3 boards type, I may adopt technique to improve the apparent horizontal resolution because 1/2 pixel moves an element for two pieces of green horizontally as well as RGB equality method. In this case two colors of CCD for blue, the red goes down for common use, and the resolution decreases, but aims at what I can compensate it for by improvement of the green resolution. Furthermore, the color imaging of two pieces of CCD methods is possible when I make green CCD only one piece (2 board methods).

Other

It is not a practical thing, but there is technique to let you perceive a color using an optical illusion phenomenon. I refer to a top of the Ben ham.

Cooling CCD

I reduced a noise due to heat by cooling off and enabled long-time exposure. In addition, there is the direction that confuses shot noise with a thermal noise, but is a noise of the another kind really.

Size name of the CCD image sensor

There are two ways of methods about the size name of the image sensor. This is similar about the CMOS image sensor.

By the inch unit call; size

It is size to be called by the inch unit including 3/2 inch, 1/1.8 inch, the 1/2 inch. このサイズはイメージセンサの撮像面の実寸を示すものではなく、呼び名に相当する管径の撮像管の撮像面サイズと等しいことを表す。 2/3インチセンサの場合は(16.9mmではなく)2/3インチ撮像管に相当する対角11mm(8.8mm x 6.6mm)、1/1.8インチセンサでは対角8.93mm(7.18mm x 5.32mm)、1/2インチセンサでは対角8mm(6.4mm x 4.8mm)が実寸となる。 これはCCDイメージセンサの初期の用途がテレビカメラ用の撮像管を置き換えるものであったため、レンズなどの光学系を設計したり選択したりする際の便宜を考慮してこのような習慣が生まれたものである。 同様な理由により、特に断らない限り画面の縦横比は標準テレビ画面と等しい4:3が主流である。

なお日本では計量法により取引における「インチ」の使用が認められないためブラウン管などと同様「1/2.5型」などと表記する場合があるが同じことである。

規格名称

特定の規格により実サイズが規定されているもので、35mmフルサイズ135フィルムを用いる35mm判の画面サイズに等しい36mm×24mm)、APS-C(16.7mm×23.4mm)などがある。なお、フォーサーズ・システムについては約18.0mm×13.5mm(実際は17.3mm×13.0mm)の実寸が規定されているためこの範疇であるが、そのサイズ自身は「管径4/3インチの撮像管」に由来している。

CCDイメージセンサの製造

CCDイメージセンサの製造技術は、半導体部分の製造プロセス、カラーフィルタやマイクロレンズの生成などにおいて一般的なCMOSロジック製造技術とは異なる部分が多い。このため主要な製造メーカーは半導体メモリCPUASICなどとはかなり異なる。2004年における大手製造メーカーはソニーシャープ、松下電器産業(現パナソニック)、三洋電機で世界市場の約8割を占め、これはCMOSイメージセンサにおいて首位のソニー、2位の東芝とも6、7%程度のシェアしかない状況とは対照的である。このほか、KODAKもCCDイメージセンサを製造している。

このため、デジタルカメラやビデオカメラなどの製品では競合しているメーカーにCCDイメージセンサを供給する例は珍しくない。例として、ソニーが広範な供給を行っていた事実がはからずも明らかになった事件がある。

ソニー製CCD不具合問題

ソニーはCCD製造において大手であり、ビデオカメラやデジタルカメラを製造する会社へ供給していた。2002年10月にCCD素子の窓となるガラス板を接着するために用いていた接着剤を変更した。

2004年春、業務用ビデオカメラにおいてCCD素子とICパッケージのピンを結ぶボンディングワイヤが破断する不具合が見つかった。ボンディングワイヤリング装置に不適切な設定を行った状態で製造していたことが判明したため、ソニー製CCDを使用していた業務用ビデオカメラメーカー各社はリコールを行い事態は収拾したと思われた。

しかし、2004年夏から徐々に、再びボンディングワイヤが破断する問題が発生した。これは変更された接着剤に含まれるヨウ素がCCDチップとボンディングワイヤが繋がっている合金部に侵入し、欠陥が生じて合金部で破断が発生するものであった。 ボンディングワイヤリング装置の設定ミスが原因と思い込んでいたソニーは、問題のある接着剤でCCDを製造し続けたため、本問題を見過ごしてしまった。

2005年10月以降、ソニーおよびカメラ製造メーカー各社からリコール情報が発表された。 この問題によって不具合を生じるCCDは1000万個以上、問題を生じうるデジタルカメラ・ビデオカメラは100機種以上に及んだ。なお、ソニー以外のカメラ製造メーカーは「ソニー製CCD」に起因する不良であることを明言しているわけではないが、不具合の発生状況と公表された原因から同一原因であることが認めうるとされる。

また、富士写真フイルムの子会社富士フイルムマイクロデバイス開発のスーパーCCDハニカムの一部にも同様な不良が発生しており、 不具合の原因であるヨウ素化合物入り接着剤をCCD製造に用いるのは、当時の一般的な方法であったとも考えられる。

本節に関する参考文献

脚注

  1. ^ 神崎 洋治 (著), 西井 美鷹 (著) 「体系的に学ぶデジタルカメラのしくみ 第2版」日経BPソフトプレス; 第2版 (2009/1/29) [1]
  2. ^ 安藤 幸司 (著)「らくらく図解 CCD/CMOSカメラの原理と実践 」[2]
  3. ^ 加藤俊夫 半導体入門講座(Semiconductor JapanのWeb上講義)第16回 イメージセンサ [3]
  4. ^ http://www.roper.co.jp/Html/roper/tech_note/html/rp00.htm
  5. ^ http://www7.ocn.ne.jp/~terl/JTTAS/JTTAS-CMOS.htm
  6. ^ 現在はAT&Tの子会社。
  7. ^ 例えば、次の特許を参照のこと。 アメリカ合衆国特許第3,792,322号[4]アメリカ合衆国特許第3,796,927号 [5]また、特許の解説(日本語)として、以下の記事が詳しい[6]
  8. ^ たとえば、アメリカ合衆国特許第419,213号 [7]を参照のこと。
  9. ^ Charles Stark Draper Award, http://www.nae.edu/NAE/awardscom.nsf/weblinks/CGOZ-6K9L6P?OpenDocument 
  10. ^ Nobel Prize website, http://nobelprize.org/nobel_prizes/physics/laureates/2009/ 
  11. ^ 例えば、アメリカ合衆国特許第5,877,807号[8]
  12. ^ 「バイヤー」というカナ音記については http://d.hatena.ne.jp/ilovephoto/20130609/1370745789 を参照

文献

This article is taken from the Japanese Wikipedia CCD image sensor

This article is distributed by cc-by-sa or GFDL license in accordance with the provisions of Wikipedia.

Wikipedia and Tranpedia does not guarantee the accuracy of this document. See our disclaimer for more information.

In addition, Tranpedia is simply not responsible for any show is only by translating the writings of foreign licenses that are compatible with CC-BY-SA license information.

0 개의 댓글:

댓글 쓰기