JP2000151905A - Image sensor based on cell structure - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image sensor system that is compact, offers case of manufacture and is used for a scanner or the like. SOLUTION: The image sensor system is provide with a 1st photosensitive element 32A that provides a 1st signal in response to light, a 2nd photosensitive element 32B that provides a 2nd signal in response to light, a 1st lens 30A that focuses a light from a 1st part of an image onto the 1st photo sensing element, a 2nd lens 30B that focuses a light from a 2nd part of an image that is overlapped to part of the 1st part of the image onto the 2nd photosensitive element, and a coupling means that couples the 1st and 2nd signals from the 1st and 2nd photosensitive elements to provide an output of a signal denoting the image.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

FIELD OF THE INVENTION The present invention relates generally to image sensors, and more particularly to scanners for multifunction peripherals, including document scanners, copiers, facsimiles, and scanning devices.

[0002]

2. Description of the Related Art Conventionally, various types of image sensors have been used in scanners, copiers, and facsimile (fax) machines to collect one line of image at a time.

For example, a conventional flatbed scanner uses a single lens optical system. The image on the document is
The image is imaged by a single lens that images the image onto one photosensitive semiconductor die, with the full width of the document in view. This lens is usually composed of several optical surfaces. The drawback of this approach is that the distance between the document and the lens, as well as the lens and the semiconductor die, must be quite long, which means that the height or thickness of the scanner tends to be large. I mean. For this reason, the device was extremely bulky.

The "f" number of a single lens optic in a flatbed scanner is very large. “F” refers to the angle at which light is collected. The larger the "f" number,
Light collection is reduced. The large "f" number is the main reason for requiring very bright light in a single lens scanner.

[0005] Substantial improvements have been realized by making flatbed scanners that lower the height of the device by bending the optical path with mirrors. Unfortunately, this adds to the cost of the equipment. First, many sophisticated mirrors are required. Second, these mirrors contribute to additional light loss. Third, it requires expensive semiconductor dies. The die had to be long enough to have a sufficient number of pixels to cover the entire width of the image. The width of the die must typically be greater than a minimum to obtain an acceptable aspect ratio (length / width). If the aspect ratio of the die is too large, the stability is lost, and there is a possibility that the die may be damaged during a semiconductor process such as sawing and die attachment. The width of the die is wider than the minimum required to give the circuit an acceptable aspect ratio, which is more expensive than if it could be made with the smallest circuit width. And fourth, the equipment still tends to be relatively thick.

In order to develop a cheaper and smaller device, a SELFOC lens is used in a different type of scanner sensor called a contact image sensor. In the SELFOC lens, a plurality of rods are arranged with their end faces aligned. The optical path is along the rod axis. S
In the material forming the ELFOC lens, the refractive index changes according to the radial distance from the rod axis. The radially varying refractive index causes the rod to act as a lens. Each rod functions as a lens with a narrow field of view. By lining the rods side by side, a linear image can be transferred.
When the SELFOC lenses are arranged in a line, the fields of view of the rod lenses overlap. To form a linear image, the rod must be cut to length and positioned so that the image of each lens is a positive (non-inverted) image at 1: 1 magnification.

[0007] A contact image sensor is much smaller than a single lens system. These are generally cheaper, but have a much smaller depth of field and lower image quality.

[0008]

SUMMARY OF THE INVENTION There are two methods for manufacturing a lens.
There is a street. One is to use a glass rod. These glass rods are made by immersing them in a dopant and diffusing the dopant into the rod. It is this dopant that provides the gradient for the deflection. Then, the rod is cut to a certain length, joined by epoxy, and both ends are ground. This tends to be a relatively tedious and expensive method.

A less expensive approach is to use plastics extruded by multiple extrusion steps, each with a different refractive index. This multiple extruded plastic is later melted and the refractive index is more continuously graded. This plastic SELFO
C generally has a poor image quality because the refractive index generally still changes stepwise and does not change smoothly. A major drawback of conventional contact image sensors is that the photosensitive semiconductor die cannot be made as a single component. Therefore, to produce the long columns needed to match the width of the document,
A large number of dies must be cut to very close tolerances and then assembled very carefully in detail. The production of such a long line-shaped die is the most difficult assembly process when producing a contact image sensor. The sensor die is also the most expensive part of these sensors.

Semiconductor dies must be assembled to maintain continuous pixel spacing, even between pixels between adjacent dies. It is unacceptable for tolerance deviations to accumulate over all about twenty semiconductor dies employed to fabricate a contact image sensor. The pixels of conventional contact image sensors have a poor "fill factor". As much as necessary for the die sawing process, only the gap between the die and the minimum spacing of the active area of the die with respect to the saw edge,
The size of the pixels at the edge of the die needs to be reduced.
In order for all pixels in the array to respond equally to light, the size of all pixels must be reduced to be the same as the edge pixels. The fill factor is the active area of a pixel divided by the total area used by that pixel (the reciprocal of the number of pixels per inch). A contact image sensor with 600 pixels per inch typically has a 30% fill factor. A small fill factor means that the amount of light collected per pixel is smaller and it is necessary to either brighten the light or slow down the scanner. If it becomes technically necessary to obtain 1200 pixels per inch, it becomes impossible to manufacture a contact type image sensor.

[0011] In the conventional contact type image sensor, the steps of die sawing and die attaching require the die to be sawed to a very precise tolerance (typically 10 microns). Yes, and it is very difficult to achieve assembly tolerances that maintain such spacing. Doing these with standard automatic die attach equipment is too difficult.

In many cases, the sawing and assembling steps cause the semiconductor die to chip and crack, resulting in low yield and high cost. The die width is often about 0.027 inches (0.069 cm) wide by 0.5-0.8 inches (1.2-2 cm) long. With non-square aspect ratios, die sawing is very difficult. The smaller the number of die required, the more
Of course, the number of dies that need to match each other is small. Acceptable aspect ratios are 20 to 30:
It is often assumed by manufacturers that the die is wider so that the number of dies used to obtain a certain width can be reduced, but this results in additional cost due to wasted die area.

[0013] The alignment is very complex and is usually done manually. However, this often leads to significant losses with a yield loss of the order of 30% even with careful work. Alignment errors often result in scrapping the entire array of sensor dies. Since the die is the largest cost in the contact image sensor, this yield loss greatly affects the cost.

A scanner has conventionally been provided with a white light illumination device, and performs color separation by a prism or a filter. The white light was typically a tubular lamp such as a fluorescent lamp. These lamps were not color stable and required adjustments to control intensity and color. Also, the life was limited.

With an LED, color separation can be performed during illumination. Using three or more colors, the color of illumination that changes with each exposure is read three times with a detector. Using four or more LEDs gives a more realistic tint (ie,
Color rendition).

The next difficulty relates to placing the die on the PC board. Because there is a pattern on the PC board,
The edges of the board tend to float, which means that from an optical point of view, it is difficult to accurately position the semiconductor die to align with the optics of the sensor.

A person of ordinary skill in the art has long sought an inexpensive and small system having a quality equal to or better than that of a single lens system, but has long been an unsolved problem.

[0018]

According to the present invention, a series of plastic molded lenses are used to image a series of photosensitive dies. Each lens combines with one die to create one cell. The slight overlap of the fields of view of the plurality of cells on the image allows manufacturing tolerances for photosensitive dies and the like. The overlap may be greater than the minimum so that the image is reproduced properly. The lens can be enlarged (reduced) so that the die can be smaller than the image. By utilizing the overlapping of the images, a plurality of small images obtained from each cell can be combined to obtain an output signal for reproducing an original image.

The lens in each cell may be a single optical element, but usually has several optical surfaces.
By using more optical surfaces, it is possible to correct optical aberrations such as chromatic aberration. Further, the lens may be a refractive lens, or may function as both a refractive lens and a reflective lens. The shape of the optical surface may be spherical, non-spherical, diffractive, or a combination thereof.

The advantages of the present invention will become apparent to those skilled in the art from a reading of the following detailed description when taken in conjunction with the accompanying drawings.

BEST MODE FOR CARRYING OUT THE INVENTION

Referring now to FIG. 1, an image sensing device 10
Which is a scanner, a facsimile (fax) device, a copy machine, or the like. The image detection device 10 is set to detect the image 14 on the top of the document 12,
The scanning line is moved in the vertical direction in the direction indicated by the arrow 16 so as to detect "one" scan line across the width of the document 12 at one time.

Illumination of document 12 is provided by a series of light-emitting diodes (LEDs) of different colors. LED1
8, 20, 22 and 24 can be red, blue, green and yellow so that the various color images 14 can be detected, but need not be in this order.
The LED of FIG. 1 is shown in the plane of the paper. Also,
These LEDs may be arranged in a row perpendicular to the plane of the paper of FIG. The lighting device is an end feeder.
illuminator).

During one scan, one of the light emitting diodes (1
From 8), the light travels along the optical path 26. Light travels along a line 28 through a molded plastic lens array 30 to a photosensitive semiconductor die 32 embedded in a molded lead frame 34.

In a full color image, three scans, usually referred to as red, green and blue scans, are used. The spectral distribution of each scan can affect the scanner's ability to reproduce the correct color. To improve the spectral distribution, LEDs of two or more colors may be turned on each time scanning is performed.

The photosensitive semiconductor die 32 is connected to circuitry 36 to extract signals therefrom and convert them into signals usable by a particular scanner, fax machine, copier, or the like. The cable 38 supplies power to the image detection device 10.

Next, referring to FIG.
Image sensor 4 (shown vertically in FIG. 1)
0 shows a state viewed from the lateral direction. The image sensor 40 is configured by a molded lens array 30, and the lens array includes, for example, each lens 30A, 30B, 30C,
It is composed of 30D and 30E. These lenses are
Although shown schematically as a single lens element, it is most commonly comprised of multiple elements. As shown in FIG. 2, each of these lenses is a lens 30A, 30A.
B, 30C, 30D, and 30E cover cell fields of view designated as cell fields of view 42A, 42B, 42C, 42D, and 42E, respectively. These fields of view overlap and overlap 44A, 44B, 44C.
And 44D. Lens array 3
0 sets the cell visual fields 42A, 42B, 42C, 42D and 42E to the photosensitive semiconductor dies 32A, 32B, 32C, 3
Focus on 2D and 32E, respectively. Each lens and associated die form a cell of the image sensor 40.

The molded lead frame 34 has an alignment function (not shown) so that the molded lead frame 3
4 can be accurately positioned. The alignment function allows the molded lead frame to be accurately positioned on the assembling apparatus, and the alignment function allows the final scanner assembly to be accurately positioned by the final manufacturing process of the entire scanner assembly. Accurate alignment allows the semiconductor die 32 to be accurately positioned with respect to the molded lead frame and thus with respect to the optical system 30.

Referring now to FIG. 3 (prior art), prior art photosensitive semiconductor dies 50A and 50B are shown. On the surface of the semiconductor die 50A are a series of pixels 52, 54,
56, 58 and 60 are produced. Similarly, semiconductor die 50B also includes pixels 62, 64, 66, 68 and 70
Having. The pitches 53 and 57 between the pixels must be kept the same and, for example, the pitch 61 must be the same between the pixels at both ends of the semiconductor die. It is not permissible for tolerance shifts to accumulate over all ten or more semiconductor dies 50. This is because the semiconductor dies 50A and 50B have very narrow tolerances.
Means that the gap, designated as gap 72, must be sawed to within a tolerance of typically 0.0002 inches (5 μm).

Prior art pixels have a poor "fill factor". This means that the spacing between all pixels, including pixels 60 and 62 at both ends of the die, must be kept the same, reducing from the maximum possible size. Die assembly is difficult to maintain a constant pixel pitch across the gap between the dies. Assembling such dies with standard automatic die attach equipment is very difficult. It is possible to make accurate die attach with very expensive custom equipment, but it is usually done manually.

The sawing process for semiconductor dies usually produces chipping and cracking at the edges. The dies of conventional contact image sensors cannot be tolerated with normal amounts of chipping and cracking, so the sawing process must be maintained at a fine level.

Also, in the assembly process, the die may chip and crack. Butts against each other can cause the die to chip and crack. For this reason, the manual or automatic die attaching operation is more difficult than in a normal operation, which causes a reduction in yield.

Referring now to FIG. 4, semiconductor dies 32A and 32B are shown. It should be noted that the new semiconductor die is approximately one-third or less in size compared to prior art semiconductor dies 50A and 50B. 1 pixel
Since they have a fill factor of 00%, they are adjacent to each other and there is no problem with the arrangement of the pixel gaps.
For example, the pixels of the semiconductor die 32A are
2, 84, 86, 88 and 90, which can be arranged by conventional semiconductor fabrication steps. The exact location where the image 14 connects the image on the semiconductor die 32A is not important and is corrected by image processing. Therefore,
Due to the overlap 44 and the image processing, the distance 104 between the dies is not important. Also, the die's sawed edge and distance 102 from the pixel are not critical.

In operation, a document 12 having an image 14 passes through the image sensing device 10. Light emitting diodes 18, 20, 22, and 24 illuminate a continuous band of light in a narrow band-like area on image 14, and light travels toward lens array 30 according to the optical path indicated by arrow 28. Assuming that the lens array is colorless, the lens array includes LEDs 18, 20, 22
And 24, respectively, can be focused on the plane of the semiconductor die 32.

As shown in FIG. 2, the image 14 is
0A, 30B, 30C, 30D, and 30E, each portion 42A, 42B, 42C, 42D, and 42E of the image is imaged through lens array 30 with a typically 3: 1 to 6: 1 reduction. The overlap is indicated by overlap portions 44A, 44B, 44C and 44D.

Each portion of image 14 is imaged on a photosensitive pixel of semiconductor die 32. In the overlap region 44, the target image 14 may be imaged on both semiconductor dies 32. It should be noted that accurate alignment is not important because such overlaps are matched and removed by image processing and the original image 14 is re-formed.

It can be seen that the entire image sensing device 10 can be manufactured at a much lower cost while providing higher quality than the prior art.

For example, the price of the molded lens array 30 is six times cheaper than a glass SELFOC array and twice as cheap as a plastic SELFOC array.

Also, molded lead frames are more than twice as cheap as printed circuit boards required in the prior art. The molded leadframe also simplifies assembly into other components. Further, from 3: 1 to 6: 1
As a result, the size of the semiconductor die can be reduced to 1/3 to 1/6 of that of the conventional contact type image sensor, so that the cost is reduced in proportion thereto. Further, with this configuration, 30% of the conventional contact image sensor can be obtained.
The cost of assembly yield loss is virtually eliminated.

Furthermore, the alignment of the semiconductor dies is no longer important, which speeds up the assembly. At the same time, the lens array is
Since it can operate at higher resolution than the LFOC array,
The image quality will be better. Good resolving power means that the image sensor 40 can operate at much more than 1200 dots per inch compared to prior art SELFOC systems, which are limited to about 600 dots per inch.

The SELFOC-based image sensor is
When used in a scanner with 600 pixels per inch, the depth of field is limited to around 1 mm. The new device 40 can be designed to use much greater depth of field using conventional optical design techniques. This means that the fluctuation range can be larger without being significantly affected by the distance between the document 12 and the lens array 30.

When the f-number is large, the same SELFOC
Although more light is required as compared with the above, light can be collected more efficiently by improving the fill factor.
When the resolution exceeds 600 pixels per inch, the fill factor is immediately impaired in the contact type image sensor. Therefore, a large fill factor allows a new sensor with excellent light collection even at a resolution with a large number of pixels per inch. Can be brought. This has a dramatic effect as the resolution increases.

The semiconductor die 32 may utilize CCD or photodiode technology. It is possible to make the die in CMOS, bipolar or amorphous silicon.

While the invention has been described in conjunction with specific best modes, it should be understood that many alternatives, variations and modifications will be apparent to those skilled in the art in light of the above description. It is therefore intended to cover the above alternatives, modifications and variations which fall within the spirit and scope of the appended claims. All matters described herein or shown in the accompanying drawings are to be interpreted merely in an illustrative and non-limiting sense.

[Brief description of the drawings]

FIG. 1 is a schematic side view of a system incorporating a cell-based image sensor of the present invention.

FIG. 2 is a schematic diagram of a light detection unit of the present invention.

FIG. 3 is part of an assembly of a contact image sensor semiconductor die array (prior art).

FIG. 4 is a portion of a semiconductor die configuration of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Image detection apparatus 12 Document 14 Image 16 Vertical direction 18,20,22,24 Light emitting diode 26,28 Optical path 30 Lens array (optical system) 32 Semiconductor die 34 Lead frame 36 Circuit 38 Cable 42 Cell view 44 Overlap

 ──────────────────────────────────────────────────続 き Continued on the front page (71) Applicant 399117121 395 Page Mill Road Palo Alto, California U.S.A. S. A.

Claims (20)

[Claims] A first photosensitive element for providing a first signal in response to light; a second photosensitive element for providing a second signal in response to light; and transmitting light from a first portion of the image to the first photosensitive element. A first lens converging on a photosensitive element, and a second lens of the image overlapping a portion of the first portion of the image.
A second lens for focusing light from a portion on the second photosensitive element; and the first and second light from the first and second photosensitive elements.
Means for combining signals and outputting a signal representative of the image. 2. The sensor system according to claim 1, wherein said first and second photosensitive elements are spaced apart. 3. The first and second photosensitive elements comprise a plurality of pixel elements, and light from the first and second portions of the image is focused on the elements below the plurality of pixel elements. The sensor system according to claim 1, wherein 4. The sensor system according to claim 1, wherein said first and second lens elements are a single unit. 5. The sensor system according to claim 1, wherein said first and second photosensitive elements are on separate carriers. 6. The sensor system according to claim 1, wherein said first and second photosensitive elements are on a single carrier. 7. The sensor system according to claim 1, wherein said first and second lens elements invert said first and second portions of said image, and said combining means includes means for inverting said image again. 8. The method according to claim 1, wherein said first and second lens elements are in proximity to each other.
The sensor system according to claim 1, comprising a light source that illuminates the image. 9. A first photosensitive semiconductor die for providing a first signal in response to light, a second photosensitive semiconductor die for providing a second signal in response to light, and light from a first portion of the image. A first lens that focuses on the first photosensitive semiconductor die; and a second lens of the image that overlaps a portion of the first portion of the image.
A second lens for focusing light from the portion onto the second photosensitive semiconductor die. 10. The sensor system according to claim 9, wherein said first and second photosensitive semiconductor dies are spaced apart. 11. The first and second photosensitive semiconductor dies include a plurality of pixel elements, and light from the first and second portions of the image is focused on the elements below the plurality of pixel elements. The sensor system according to claim 9. 12. The sensor system according to claim 9, wherein said first and second lenses are a single unit molded of plastic. 13. The sensor system according to claim 9, wherein said first and second photosensitive semiconductor dies are on separate die carriers. 14. A molded single plastic carrier, wherein said first and second photosensitive semiconductor dies are capable of accurately holding said first and second photosensitive semiconductor dies with respect to said first and second lenses. The sensor system according to claim 9, wherein: 15. The sensor system according to claim 9, wherein said first and second lenses invert said first and second portions of said image. 16. The sensor system according to claim 9, further comprising a plurality of light emitting diodes proximate to the first and second lenses and illuminating the image. 17. A method for focusing light from a first portion of the image onto a first photosensitive element, and a second portion of the image overlapping a portion of the first portion of the image.
Focusing light from a portion onto a second photosensitive element; and the first and second light sources from the first and second photosensitive elements.
Combining the signals. 18. The method according to claim 18, wherein light from a first portion of the image is
The step of focusing on a photosensitive element focuses light on the element below the complete first photosensitive element, and the step of focusing light from a second portion of the image on the second photosensitive element comprises: 18. The method of claim 17, wherein light is focused on the element below a complete second photosensitive element. 19. The step of focusing light through the first and second lens elements inverts the first and second portions of the image, and the step of combining the first and second signals comprises the step of combining the first and second signals. 2. The method of claim 1, further comprising inverting first and second signals to reinvert the image and recombining overlapping regions.
7. The image detection method according to 7. 20. The method according to claim 17, further comprising the step of illuminating the image.