JP2001174415A - Defect detection optical system and surface defect inspection device - Google Patents

Abstract

(57) Abstract: An object of the present invention is to provide a defect detection optical system and an inspection device that can easily adjust the detection sensitivity and can detect minute unevenness defects. A light projection system that irradiates a light beam having a width in a direction perpendicular to a main scanning direction to relatively scan a face plate;
A light receiver having n (where n is an integer of 2 or more) light receiving elements arranged in a perpendicular direction and receiving reflected light from the face plate is provided, and an image of the scanning position of the face plate is formed on the n light receiving elements. A light receiving system, a fringe pattern filter inserted in the light receiving system and having a light-receiving surface adjacent to the light-receiving surface of the light-receiving element substantially opposite to the right side and the left side of the light-receiving surface, and having a light-transmitting or light-shielding relationship reversed; A detection circuit for generating a detection signal corresponding to the difference in the amount as a signal for detecting a defect, wherein the reflected light from the surface in the case where there is no defect is substantially applied to the central portion.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a defect detection optical system and a surface defect inspection apparatus, and more particularly, to the adjustment of detection sensitivity in a surface defect inspection apparatus for a magnetic disk or a surface plate such as a glass substrate (glass substrate). The present invention relates to a defect detection optical system and a surface defect inspection apparatus that can easily detect minute irregularities and defects, and can accurately detect the size and depth or the magnitude and height of the irregularities on the face plate surface. .

[0002]

2. Description of the Related Art Hard magnetic disks used as recording media for computer systems include a substrate disk (substrate) as a material or a magnetic disk coated with a magnetic film (for convenience, these are collectively referred to as a magnetic disk or simply a disk). ), The defects existing on the surface and their sizes are inspected. In recent years, discs having a size of 3.3 inches or less have become mainstream, and their recording density has been dramatically increased by the use of GMR heads. In this type of disk, a glass disk having a smaller coefficient of thermal expansion is used from an aluminum substrate, and its thickness is as thin as about 0.6 mm to 0.8 mm.

FIG. 7 shows a configuration of a main part of a conventional magnetic disk surface defect inspection apparatus 10. The surface defect inspection apparatus 10 shown in FIG. 7A includes a rotation mechanism 2 and a detection optical system 3.
And a surface defect detection processing unit 4. The disk 1 to be inspected is a spindle 21 of the rotating mechanism 2.
And rotated by the drive of the motor (M) 22.
Detection optical system 3 on the other hand, the laser beam L _T focusing lens from the laser light source 311 of the light projection system 31 312
And it is focused by, and forms a spot S _P on the surface of the disk 1 is irradiated with the surface of the disk 1. For example, the spot S _P by the movement of the X-axis direction of the disk 1 is moved in the direction of the radius R of the disk 1, thereby the surface of the disk 1 is scanned spirally. In this case, in order to minimize the scanning time, the spot S _P is an elliptic shape having a minor axis phi ₁ and diameter phi ₂ as shown as FIG. (B) on the lower side, the longer diameter phi ₂ in the scanning direction A method is adopted in which the scanning width is widened by setting the angle at a right angle.

[0004] The defect F which is present on the surface, to scatter the light of the spot S _P. The reflected light S _R is condensed by the condensing lens 321 of the light receiving system 32 and is converted into a photoelectric conversion element, for example,
The light is received by a light receiver 322 formed of an avalanche photodiode (APD) or a photomultiplier tube (PMT).
The output signal of the light receiver 322 is input to the signal processing circuit 41 of the surface defect detection processing unit 4. Here, the defect F is detected, and the size of the defect is classified or calculated based on the amplitude of the output signal. The signal processing circuit 41 detects a defect and classifies or calculates the size.
A circuit for detecting a defect by so-called sampling, which amplifies an output signal from the light receiver 322;
A sampling circuit for detecting a peak value of a defect by sampling a peak value of an output signal of a defect exceeding noise among output signals amplified by receiving a pulse from the rotary encoder 23, and further digitally converting the sampled peak value. An A / D converter for converting the value into a value, a position data generating circuit for generating position data on the disk in response to a pulse from the rotary encoder 23, and the like.

The data of the size of each defect and the data of the position on the disk are A / D-converted in a signal processing circuit 41 and input to a data processing device 44 comprising an MPU 42 and a memory 43 and the like. Then, here, the number of each size is counted, and the result is printed out by the printer (PR) 45 together with the position of the defect on the disk 1. Alternatively, the size of the display (C
(RT) 46 and the like, along with the position on the disk, and the count value is also displayed separately. The rotary encoder 23 is provided adjacent to or engaged with the rotation shaft of the motor 22, detects the reference rotation position and the rotation amount of the disk, and sends each pulse signal to the signal processing circuit 41. I do.

There are various shapes of the defect F in the disk 1. One example is shown in FIG. 8, the defect F _h, be those dished, have a relatively large diameter D _h, shallow depth d _h in comparison with this. Defect F _P has been made in view of the well-shaped, the diameter D _P is relatively small, the depth d _P is large, those called pits. Note that the two defects F _h and F _P are often present in isolation. On the other hand, F _S is a linear scratch defect, and its cross-sectional width w _S and depth d _S vary. There are, of course, defects F in forms other than these shapes. In addition to such concave defects, there are various sizes and heights of protrusion defects such as a defect called a convex foreign material generated by attachment of fine particles or the like.

[0007] Therefore, a defect of various shapes and sizes, in order to better detect, respectively, the surface defect inspection apparatus 10 described above, the projection angle of the laser beam L _T of the light projecting system 31 theta
_T , the light receiving angle θ _R of the light receiving system 32, and the light receiving device 322 (AP
Via the control panel 47, the voltage V to be applied to D), the gain related to the amplifier incorporated in the signal processing circuit 41, the threshold voltage E for removing noise, the laser output of the laser light source 311 and the like, which are related to the detection sensitivity. Set each to optimal. The sensitivity adjustment is performed by using an actual disk having a sample defect such as a dish-like defect, pit defect, or scratch defect of a known size or an actual disk having a protrusion of a specific height as a sample. Is being done. As an invention of this type, there is an application by the applicant of Japanese Patent Application Laid-Open No. 10-325713, entitled "Surface Defect Inspection Method and Inspection Apparatus". This is because a simulated defect row whose size is gradually increased or decreased by performing a defect inspection using a radial sensitivity calibration disk in which the size of the convex portion or the concave portion of each simulated defect row is gradually different. Are displayed as test results in a radial pattern, and the detection sensitivity is adjusted in accordance with the test results.

[0008]

However, as the unevenness defects to be detected become minute pit defects or needle-like defects, it becomes more difficult to adjust the detection sensitivity. Moreover, in the above-described conventional defect detection system, the light receiving device detects a concave defect or a convex defect (including a foreign substance) by comparing the level of reflected light or scattered light with a reference level. Therefore, as the number of irregularities to be detected becomes smaller, the level change becomes smaller in the mode in which scattered light is received, and the difference from the background noise disappears, making defect detection more difficult. As a result, it becomes difficult to distinguish convex defects, particularly foreign matters and concave defects. Even if a defect of the same size as the conventional one is to be detected, recently, it has been required to improve the accuracy of measuring and classifying the shape of the unevenness.
In addition, since the concave defect is located at a lower position with respect to the disk surface with respect to the objective lens, and the convex defect is located at a higher position with respect to the disk surface, it is located at a position closer to the disk surface. Therefore, it is difficult to accurately separate the sizes of the two in terms of the amount of received light. SUMMARY OF THE INVENTION An object of the present invention is to solve such a problem of the prior art, and to provide a defect detection optical system and an inspection apparatus which can easily adjust the detection sensitivity and can detect minute unevenness defects. It is in. Another object of the present invention is to provide a defect detection optical system and an inspection device capable of distinguishing uneven defects on the surface of a face plate. Still another object of the present invention is to provide a defect detection optical system and an inspection apparatus which can accurately detect the size and depth or the size and height of an uneven defect on a surface of a face plate and can easily classify the defect. It is in.

[0009]

A feature of the defect detection optical system and the surface defect inspection apparatus of the first invention for achieving the above object is that the surface of the face plate is scanned with a light beam and the light beam is scanned. In a defect detection optical system that receives light reflected from the surface by a light receiver and generates a signal for defect detection, a light beam having a width in a direction perpendicular to the main scanning direction is irradiated. A light projection system that relatively scans the face plate, and a light receiver that has n (where n is an integer of 2 or more) light receiving elements that are arranged in a perpendicular direction and receive reflected light from the face plate; The light-receiving system that forms the image of the image on n light-receiving elements and the light-receiving surface of the light-receiving element inserted in this light-receiving element are substantially opposite in transmission or light-shielding between the light-receiving elements adjacent to the right and left sides from the center. Stripe pattern And a detection circuit for generating a detection signal corresponding to the difference in the amount of light received by the adjacent light receiving element as a signal for defect detection, and the reflected light from the surface when there is no defect is substantially at the center. Is irradiated. Further, the feature of the defect detection optical system and the surface defect inspection apparatus of the second invention is that the stripe pattern filter has alternately transmission and light-shielding stripe patterns corresponding to the width in the arrangement direction of the light receiving elements, and The left and right patterns are substantially displaced in the arrangement direction by the width of the light receiving elements in the arrangement direction at the center, and the detection circuit receives an electric signal from an adjacent light receiving element and outputs a signal corresponding to the difference. , A plurality of differential amplifier circuits, receives a signal corresponding to the difference from these differential amplifier circuits, and generates a detection signal obtained by adding the signals according to the difference.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS As described above, the relationship between transmission and shading is substantially reversed between adjacent light receiving elements on the right and left sides substantially from the center of the light receiving element. When the reflected light from the surface is substantially applied to the central portion, there is no difference in the amount of received light between adjacent elements, but when there is a defect on the surface, the reflected light is directed to the left or right from the center of the light receiving element. Because of the fluctuation, a difference occurs in the amount of received light between adjacent elements, which becomes a defect detection signal. In other words, when a defect is scanned, the specularly reflected light from the defect is deflected to one of the right and left sides of the pattern, so that the light receiving position of each light receiving element is shifted by that amount according to the size of the defect. In addition, the order of the left and right shifts of the concave defect and the convex defect due to the relationship between the dent and the protrusion appears reversely. Since the shift amount changes in accordance with the size of the defect, a detection signal corresponding to the shift amount of the light receiving position can be obtained by the detection circuit, whereby the size of the defect and the irregularity defect can be detected. In this case, detection is performed based on a difference in light reception level between adjacent elements, and is not a comparison between the level of scattered light or reflected light and a reference level. Defect detection can be performed with high accuracy. Further, in the present invention, since the light receiving elements are arranged and detected,
The number of arrays of light receiving elements is increased to detect minute defects by each light receiving element, and the size of the defect can be easily determined by data processing of the detection relationship with adjacent elements, enabling highly accurate classification. become.

Further, in the second invention, since a large number of signals corresponding to the difference between adjacent light receiving elements are added to obtain a detection signal, the detection signal is affected by a blank portion (dead zone) between the light receiving elements. This makes it possible to detect gentle irregularities such as dish-shaped defects, which are difficult to detect only by the difference between adjacent light receiving elements. In this case, in particular, the detection value of the difference between the adjacent light receiving elements is divided into odd and even numbers, the difference between the next adjacent light receiving element is taken, the difference is added, and then the integrated value is obtained. can do. As a result, it is easier to adjust the detection sensitivity than when directly detecting the intensity of reflected light or detecting the intensity of scattered light, and it is possible to detect minute unevenness defects. Classification accuracy can be improved.

[0012]

FIG. 1 is a block diagram showing an embodiment of a surface defect inspection apparatus to which the present invention is applied. FIG. 2 is a drawing showing a detection optical system in a direction parallel to the drawing (θ direction) and perpendicular to the drawing. FIG. 3 is an explanatory view of development of a light projecting / receiving system in one direction (R direction), FIG. 3 is an explanatory view of a relationship between a stripe pattern filter and an APD array sensor, and FIG. FIG. 5 is an explanatory diagram of a waveform detected by a regular reflection light of a convex defect, and FIG. 6 is a block diagram of another embodiment of a surface defect inspection apparatus to which the present invention is applied. In FIG. 1, reference numeral 100 denotes a surface defect inspection device, and reference numeral 50 denotes a detection optical system. 51 is a light projecting system for detecting optical system 50, the light projecting system 51, the laser beam L _T than the laser light source 511 receives a beam expander 512, in one direction (R direction) perpendicular to the paper surface in the drawing By expanding the laser beam, the focal position is shifted to the front. The beam expanded here passes through a cylindrical lens 513 and a focusing single lens 514, and is focused on the surface of the disk 1 by shifting the focal position toward the beam waist in the radius R direction by an offset Δd (see FIG. 2A). The beam spot _Sp is expanded to an elliptical shape (see FIG. 2B). In FIG. 1, in a direction parallel to the paper surface of the drawing (the disk rotation direction, that is, the θ direction), the light beam that has passed through the cylindrical lens 513 is As shown in the drawing, the light beam is focused on the surface of the disk 1 through the focusing single lens 514 at the inspection point S in a point-like manner (FIG. 2).
(Refer to point ◆ in (b)). As shown in FIG. 2 (b), the projection angle is about 35 ° with respect to the normal on the disk 1 in the θ direction.
゜. And in the direction (R direction) perpendicular to the paper surface,
The light has a constant width W (for example, about 150 μm to about 200 μm), and the width W is a length covering the arrangement range of the light receiving elements on the imaging plane. In this respect, it is different from the beam spot S _p of FIG.

In FIG. 1, a light receiving system 52 has an objective lens 521, which receives specularly reflected light from an inspection point S of the disk 1 and converts it into parallel light to form a stripe pattern filter 522. Through the imaging lens 523. The imaging lens 523 includes the APD array sensor 52
An image of the inspection point S is formed on the light receiving surface of No. 4. As shown in FIG. 3, the APD array sensor 524 includes an element that receives light corresponding to each stripe width of the stripe pattern filter 522 along a radius R direction of the disk 1 (a direction perpendicular to the plane of FIG. 1). (For example, 22), and the relationship with the field of view A of the laser beam corresponds as shown in FIG.
The width W (about 150 μm to about 200 μm) covers the range of the visual field A (about 120 μm). In addition,
In the figure, for convenience of explanation, the number of light receiving elements is not all 22 but is reduced.

FIG. 2A is a development explanatory view of a light emitting / receiving system for explaining a direction (θ direction) parallel to the paper surface and one direction (R direction) perpendicular to the paper surface. The upper side is parallel to the paper surface. In the θ direction, the lower side is the R direction perpendicular to the paper surface. As shown in FIG. 2 (a), in the θ direction, test laser beam L _T than the laser light source 511, beam expander 512, the focus single lens 514 via the focus single lens 514 to the surface of the disk 1 as a focus Point S
Is focused as a spot (see the point ◆ in FIG. 2B),
The specularly reflected light from the inspection point S passes through the objective lens 521, the stripe pattern filter 522, and the imaging lens 523, and is received by each light receiving element of the APD array sensor 524. On the other hand, as shown in the lower side of FIG. 2 (a), in the radial direction R, the laser beam L _T than the laser light source 511, beam expander 512, a focus Thing lens 51
After that, the focussing lens 514 focuses on the position just before the surface of the disk 1 by the offset Δd and focuses on the inspection point S in an elliptical shape with a width W (FIG. 2).
(B)). The specularly reflected light from the inspection point S is transmitted to the objective lens 521, the stripe pattern filter 522, and the imaging lens 523.
APD array sensor 52 through imaging lens 523 through
4, light receiving elements 524a, 524b, 524c,.
An image of the inspection point S is received by these components along the 24n arrangement direction. At this time, the regular reflection light to the stripe pattern filter 522 based on the width W of the inspection point S and the light receiving element 524
As shown by dotted lines in FIG. 3, the relationship between a and 524n corresponds to the arrangement pattern of the light receiving elements and the arrangement pattern width of the stripes and the monochrome arrangement corresponding to the width of each light receiving element. The width is, for example, 0.5 mm on the light receiving element and about 5 μm on the surface of the disk 1. The radial width W of the beam spot S _p on the disk 1 covers the entire light-receiving element.

As shown in FIG. 1, the APD array sensor 5
N light receiving elements 524a to 524n arranged in a number of 24
Are input to the defect detection unit 400 corresponding to the signal processing circuit 41 of the surface defect detection processing unit 4 in FIG. Here, data of the defect F and its depth or height are generated. Those defect data are MP
The data is input to a data processing device 44 comprising a U42 and a memory 43 and the like. The data processing device 44 classifies the unevenness, detects the depth or height, and outputs the result together with the position of the defect on the disk 1 to a printer (PR). ) 45
Is printed out. These are displayed on the display (CRT) 46 and the like together with the position on the disk, and the count value is also displayed separately. In the present embodiment, the classification performed by the conventional signal processing circuit 41 is performed by comparing the data processing device 44 with reference data by program processing. The conventional signal processing circuit 41
An embodiment in which the defect detection unit generates the classification data of the unevenness defect in the same manner as that described above will be described in the next embodiment, and is omitted here.

The defect detecting section 400 includes light receiving elements 524a to 524a.
The relationship between the stripe pattern of the stripe pattern filter 522 in FIG. 3 and the APD array sensor 524 and the detected waveform are described below.
Reference numeral 2 denotes a light-shielding fringe pattern shifted right and left in the radius R direction in the visual field A at the center of each light receiving element. The black part is a light-shielding pattern, and the white part is a transparent transmission pattern. The width of each of the black and white patterns corresponds to the width of one light receiving element in the arrangement direction of the light receiving elements 524a to 524n. That is, the pattern is shifted up and down by one pattern width with respect to the central portion O of each light receiving element. In the pattern PR of the stripe pattern filter 522 on the right side from the center of the drawing, the odd-numbered number from the top is transparent. It is a transmission pattern, and the even-numbered light-shielding patterns are black. On the other hand, in the pattern PL on the left side from the center of the drawing, the even-numbered number from the top is a transparent transmission pattern, and the odd-numbered number is a black light-shielding pattern.

Such a fringe pattern captures the difference in the amount of light received between adjacent light receiving elements in relation to the position of the left and right shifts of the received light. For example, the light receiving elements 524a and 524b Right side of drawing (right side PR)
As the light receiving position of the reflected light from the disk surface shifts to the right, the light receiving amount of the light receiving element 524a increases, the light receiving amount of the light receiving element 524b does not change, and the left side of the drawing (the left side) Similarly, in the pattern PL), the light receiving amount of the light receiving element 524b increases as the light receiving position of the reflected light from the disk surface shifts to the left, and the light receiving amount of the light receiving element 524a does not change. Therefore,
By obtaining the difference between these two received light amounts as a detection signal, it is possible to determine to which side the specularly reflected light is moving. In view of the above, the stripe pattern is not limited to the stripe pattern, and may be, for example, an inclined pattern in which the light-shielding pattern increases from the vicinity of the center toward the outside. In other words, the stripe pattern filter 522 in this case only needs to have the transmissive or light-shielded relationship reversed in the light receiving elements adjacent to the right and left sides substantially from the center of the light receiving element. In such a pattern, when the reflected light from the disk surface is received near the center position of the light receiving element, the difference between the detection signals between the adjacent light receiving elements becomes almost zero.

Returning to FIG. 1, the average value calculation circuit 401 will be described.
a is a light receiving element 524a sandwiching the light receiving element 524b.
The average value of the level of the detection signal between the light receiving element 524c and the light receiving element 524c is calculated, and the average value calculating circuit 401b calculates the average value of the level of the detection signal between the light receiving element 524b and the light receiving element 524d sandwiching the light receiving element 524c. . Similarly, the average value calculation circuit 401n-2 calculates the average value of the levels of the detection signals of the light receiving elements 524n-2 and 524n sandwiching the light receiving element 524n-1. This makes it possible to obtain an average value of two light receiving elements in the light receiving state of the same stripe pattern adjacent to each other with one light receiving element interposed therebetween. That is, a difference between adjacent light receiving elements is not simply taken here. The reason will be described later.

The differential amplifier 402a includes a light receiving element 524a
The average value of the level of the detection signal between the light receiving element 524c and the light receiving element 524c is received from the average value calculating circuit 401a.
4b, and detects a level difference between the detection signal and the average value calculation circuit 401a. When the level of the detection signal is larger than the level of the detection signal of the light receiving element 524b, a signal corresponding to the level difference is generated as a positive signal. The differential amplifier 402b receives the average value of the level of the detection signal between the light receiving element 524b and the light receiving element 524d from the average value calculating circuit 401b, detects the level difference between the detection signal of the light receiving element 524c, and detects the average value. The level of the detection signal of the calculation circuit 401b is equal to the light receiving element 524.
When the level is higher than the level of the detection signal c, a signal corresponding to the level difference is generated with a positive signal. Similarly, the differential amplifier 402n-2 receives the average value of the level of the detection signal of the light receiving element 524n-2 and the light receiving element 524n from the average value calculating circuit 401n-2, and receives the average value from the average calculating circuit 401n-2.
1 is detected, and the average value calculation circuit 4
The level of the detection signal of 01n-2 is 524n-1
A signal corresponding to the level difference is generated with a positive signal when the level is higher than the level of the detection signal.

As described above, the level difference between the light receiving signals is obtained. In this case, originally, the level difference between the light receiving signals obtained from the adjacent light receiving elements may be simply taken, but here, for the following reason, the average value of the one adjacent light receiving element is taken. That is, it has been discovered that there is a phenomenon in which the change range of the reflected light is dispersed even for a minute defect. There is a dead zone due to the blank between the light receiving elements, and when the reflected light from the defect is received in a state shifted up or down so as to straddle the blank, one of the light receiving elements on either side sandwiching it This is for receiving light. Thereby, the defect detection rate can be improved.

The differential amplifiers 402a, 402b,.
The respective outputs of n-2 are integrated circuits 403a and 403, respectively.
b,... 403n−2 and are integrated. Each integrated value is obtained by a peak detection / sample hold circuit 404.
, 404b,... 404n-2, and a peak value is detected, and the peak value is detected by each of the peak detection / sample / hold circuits 404 (peak detection / sample / hold circuits 404a, 404b,.
(as representative of n-2). The held peak values are respectively input to A / D conversion circuits (A / D) 405a, 405b,... 405n-2 and converted into digital values, and the converted values are input to the memory 406. The integration circuit 40
The output of each integral value of 3a, 403b,.
The signal is input to the defect detection circuit 407. Defect detection circuit 407
Has a hysteresis comparator, compares the output of each integrated value of the integrating circuits 403a, 403b,... 403n-2 with a predetermined value Vth under an OR condition, and when the level of the integrated value becomes equal to or more than a predetermined value. The defect detection signal DET is stopped when the respective integrated values of the integration circuits 403a, 403b,... 403n-2 fall below a certain value (corresponding to the vicinity of the peak value). This DET
Is sent to the memory 406 as the write control signal WR. The write control signal WR is
Each integration circuit 40 is slightly delayed and used as a reset signal RS.
403n-2 and each peak detection / sample hold circuit 404a, 404b,.
Is input, and the previous value is reset.

As a result, the peak value in the period from the generation of the defect detection signal DET to the end of the signal is indicated by A
/ D405a, 405b,... 405n-2 to memory 4
06 and stored as defect data. The level at this time indicates the depth of the concave defect,
The height of the convex defect is represented. Such defect data is stored in the updated address of the memory 406 together with the position data of R and θ. This address is updated each time writing is performed, and the position coordinate data of R and θ are generated by the scanning position coordinate generation circuit 408. In addition,
The write control signal WR may be generated from the peak detection / sample and hold circuit 404 at the timing of peak detection instead of the falling of the defect detection signal DET. The scanning position coordinate generation circuit 408 receives an index signal indicating the rotation reference position of the disk 1 and an angle signal θ according to the rotation from the rotary encoder 23 (see FIG. 7), and further receives the current signal from the data processing device 44. Receiving the coordinate position R in the radius R direction, the position coordinate data of R and θ are generated and sent to the memory 406 as position data. As a result, a value corresponding to the depth or height of the defect and the position where the defect exists are stored in the memory 406 each time a defect is detected. 409 is
A / Ds 405a, 405b,... 405n-2, peak detection / sample and hold circuits 404a, 404b,.
4n-2, a clock generation circuit for sending a clock CLK to the data processing device 44 and the like.

Here, each of the differential amplifiers 402a to 402n
The detection signal of -1 will be described with reference to FIG. 3, and as described above, in the pattern PR of the stripe pattern filter 522 on the right side from the center of the drawing, an odd-numbered transmission pattern from the top is the transmission pattern. Therefore, when the regular reflection light from the inspection point S of the laser beam L _T is much incident to the right of the striped pattern PR is the light receiving element 524
Since the two odd-numbered light receiving elements before and after the even number corresponding to the incident pattern receive a large amount of light through the transmission pattern, the average value calculation circuit 401a that takes the average value of the two is used first. Of the average value calculation circuit 401 (as one of the odd-numbered average value calculation circuits 401a to 402n-2), it is possible to obtain a positive and large signal. At this time, the level of the detection signal of the even-numbered light receiving element sandwiched between the two is close to zero because light reception is blocked by the light blocking pattern. As a result, the average value calculation circuit 4
01 and the odd-numbered differential amplifier 402 (differential amplifier 40a) starting with the differential amplifier 402a receiving the signal of the average value.
2a-402n-2) generate a large positive output and likewise provide it to odd-numbered integration circuit 403 (as odd-numbered one of integration circuits 403a-403n-2). Is sent. In the following, this detection signal is referred to as an odd channel detection signal.

On the other hand, FIG.
In the pattern PL on the left side from the center of the surface,
The third is the transmission pattern. So, laser bee
Mu L _TSpecularly reflected light from the inspection point S is the left stripe pattern P
When a large amount of light is incident on L, it corresponds to the incident pattern.
The two even-numbered light receiving elements before and after the odd-numbered
Because they receive a lot of light through the excess pattern,
A positive and large signal is sent to the average value calculation circuit 401 that takes an average value.
Obtainable. At this time, the odd number sandwiched between the two
The detection signal level of the
Light reception is more blocked and is close to zero. As a result, average value calculation
The even-numbered differential amplifier receiving the average signal from the circuit 401
Amplifier 402 (even number of differential amplifiers 402a to 402n-2)
The first one) produces a large positive output
To the even-numbered integration circuit 403 (integration circuits 403a to 403
3n-2 as the even one). What
In the following, this detection signal is referred to as the detection signal of the even-numbered channel.
That.

In the light receiving element in which the positions for receiving the reflected light are arranged as described above, a detection signal is generated regardless of whether the light is deflected from the center to the right or to the left. In this case, since the odd-numbered array and the even-numbered array are displaced by one light receiving element, a detection signal of a small detection area corresponding to an area of at least one element can be obtained. it can. This makes it possible to detect minute pit defects and needle-like protrusions. Here, FIG. 4 the differences of the specular reflection light of the concave defects and convex portions defect (a), will be described with FIG. 5 (a), when the laser beam L _T encounters the recess defects when scanning the disk 1 if the angle of the inclined portion when the laser beam L _T enters the recesses defects and [delta], in the θ direction as shown in FIG. 4 (a), the specular reflection light is to be rotated clockwise 2.delta.. As a result, the specularly reflected light is shifted by 2δ in the θ direction from the inspection point S. Along with this, the regular reflection light also shifts by 2δ, and a large amount of regular reflection light concentrates on the right side of the stripe pattern PR side from the center position of the stripe pattern filter in FIG. 3, and then returns to the center position of the stripe pattern,
Conversely, when leaving the laser beam L _T from the recess defect,
The angle δ of the inclined portion becomes the opposite direction, conversely, a large amount of specularly reflected light concentrates on the left stripe pattern PL, and then returns to the original center.

Conversely, in the case of a convex defect as shown in FIG. 5 (a), the relationship of the inclination angle is opposite to that of the concave portion, and a large amount of specularly reflected light is first concentrated on the left stripe pattern PL. A large amount of specularly reflected light concentrates on the right side of the stripe pattern PR, and returns to the original state. At this time, right and left shakes occur according to the size or depth of the defect. That is, the inclination angle δ becomes large for a deep defect or a high defect. In the case of a large defect, the distance from the inclination angle δ to the next returning inclination angle δ becomes longer. in this way,
In the case of a concave defect, the reflected light from the defect swings to the right and then to the left and returns to the original center position. In the case of a convex defect, the reflected light from the defect swings to the left and then to the right to return to the original center position. Return to position. That is, in the case of the concave defect, the detection signal of the odd channel is generated first and the detection signal of the even channel is generated later. In the case of the convex defect, the detection signal of the even channel is generated first and the detection signal of the odd channel is generated. Will happen later. In any case, the reflected light swings right and left when there is a defect. The magnitude of the swing corresponds to the depth or height of the defect. Therefore, the integral amplification circuit 403
In this case, an integrated signal whose amplitude corresponds to the depth or height of the defect can be obtained.

A peak detection / sample hold circuit 404
(Peak detection / sample hold circuits 404a, 404
b,... 404n-2) respectively detect a peak value corresponding to the amplitude of the signal of the integrated value. That is, each A / D 405 (A / D 405a, 405
b,... 405n−2) and stored in the memory 406 together with the defective coordinate position in response to the defect detection of the defect detection circuit 407. As a result, the defect can be detected without comparing with the threshold value as in the related art. The defect data stored in the memory 406 is
After the entire surface of the disk 1 has been scanned, the data is read out by the data processing 44, and the depth and height of the defect are calculated according to the peak value. In this case, depending on which of the detection data of the odd-numbered channel and the detection data of the even-numbered channel has occurred first, it is possible to distinguish between the concave defect and the convex defect by data processing. Thus, the peak value can be divided into a depth value (concave portion defect) and a height value (convex portion defect). Further, since the n light receiving elements 524 are arranged in the R direction to detect defects individually, the light receiving elements 52
For example, when the number n of arrays of 4 is increased, for example, when m (where m <n) out of n defects are obtained simultaneously by each light receiving element, the size can be determined based on m. That is, the size of a defect can be easily determined by performing data processing on the detection relationship with an adjacent element, and high-precision classification becomes possible.

FIG. 6 shows an embodiment in which the size of a defect is detected and a gentle concave defect or a convex defect is detected without performing data processing. In this example,
A defect detector 400a is provided in place of the defect detector 400. Other configurations are the same as those in FIG. The defect detection unit 400a detects a difference between signals of adjacent elements among the light receiving elements 524a to 524n. The relationship between the stripe pattern of the stripe pattern filter 522 in FIG. 3 and the APD array sensor 524 is the same as that in FIG. This is the same as the embodiment.

The differential amplifier 410a includes a light receiving element 524a
And the level difference between the detection signals of the light receiving element 524b and
The level of the detection signal of the light receiving element 524a is
A positive signal is generated according to the level difference when the level is larger than the detection signal b, and a negative signal is generated when the level is lower than b. The differential amplifier 410b detects the level difference between the detection signals of the light receiving elements 524b and 524c, and outputs a positive signal when the level of the detection signal of the light receiving element 524b is larger than the level of the detection signal of the light receiving element 524c. Conversely, when it is small, a negative signal is generated according to the level difference. Similarly, the differential amplifier 410n-1 is connected to the light receiving element 524
When the level of the n-1 detection signal is higher than the level of the detection signal of the light receiving element 524n, a positive signal is generated according to the level difference, and when the level is lower, a negative signal is generated.

The differential amplifiers 410a to 410 of this embodiment
n-1 is characterized in that a signal corresponding to the difference between adjacent elements is generated, including positive and negative polarities, depending on which of the received light amounts is large. The differential amplifiers 410a, 410c, ...
Each of the detection signals of the odd-numbered differential amplifiers from the first to every other from the top is input through an addition / combination circuit 411 to an integration / amplification circuit 414 that integrates in both positive and negative directions. Further, each detection signal of the differential amplifiers 410b, 410d (not shown),...
The signal is input to the integrating amplifier circuit 414 via 3. First, the first to odd detection signals of the differential amplifiers 410a to 410n-1 will be described. As described above, in the pattern PR of the stripe pattern filter 522 on the right side from the center in the drawing, the odd-numbered detection signal from the top is used. since There has been a transmissive pattern, when the regular reflection light from the inspection point S of the laser beam L _T is much incident to the right of the striped pattern PR is positive the output of the odd-numbered detection signal of the addition synthesis circuit 411 And a large signal can be obtained. Conversely, a large negative signal is generated in the even-numbered signal. However, the signal is added and synthesized by the addition / synthesis circuit 412, added to the inverting amplifier 413, and converted in the positive direction. Similarly, a positive and large signal can be obtained. . This is input to the integration amplifier circuit 414, and the integration amplifier circuit 414 can obtain a positive signal having an amplitude proportional to the intensity of the specularly reflected light as an output thereof.

As described above, in the concave defect, a large amount of scattered light concentrates on the right side of the fringe pattern PR from the center position of the fringe pattern filter in FIG. Returning to the center position of the pattern, a large amount of scattered light concentrates on the left stripe pattern PL, and thereafter returns to the original center. As a result, a signal corresponding to the difference of the odd-numbered differential amplifier 410 has a detection waveform as shown in FIG. Conversely, in the case of a convex defect as shown in FIG. 5 (a), the reflected light has a tilt angle relationship opposite to that of the concave portion, and a large amount of scattered light is first concentrated on the left stripe pattern PL. Then, a large amount of scattered light concentrates on the right side of the stripe pattern PR, and returns to the original state. As a result, when a signal corresponding to the difference between the odd-numbered differential amplifiers 410 is viewed, a detection waveform opposite to that of the concave portion defect is obtained as shown in FIG. 5B. When the signal corresponding to the difference between the even-numbered differential amplifiers 410 for the concave portion defect and the convex portion defect is viewed, the signal is opposite to the above, and the concave portion defect has a detection waveform as shown in FIG. In the case of a partial defect, the detection waveform is reversed as shown in FIG. Therefore, by synthesizing a signal corresponding to the difference between the even-numbered differential amplifiers 410 and inverting them by the inverter 413, the detection waveforms of the odd-numbered and even-numbered differential amplifiers 410 are respectively applied to the odd-numbered differential amplifiers 410. Can be matched. As a result, at the input of the integrating amplifier 414,
In the case of a concave defect, a detection signal corresponding to the difference shown in FIG. 4B is obtained, and in the case of a convex defect, a detection signal corresponding to the difference shown in FIG. 5B is obtained. This is shown in FIGS. 4 (b) and 5 (b).
This is a detection signal corresponding to the difference shown in FIG. The amplitude of these detection waveforms increases according to the size or depth of the defect. In FIG. 6, an integrating amplifier 414 receives and integrates a detection signal corresponding to the difference. As a result, a waveform as shown in FIG. 4C in which a peak of the integral value appears in the positive direction for the concave defect is obtained as a detection signal, and a waveform in which a peak appears in the negative direction for the convex defect is shown in FIG. 5C. Such a waveform is obtained as a detection signal.

Therefore, in FIG.
Peak detection / sample hold circuit 41 provided after 4
In step 5, the peak value of the integration signal is detected and sampled and held. In this case, when a positive peak value is obtained, a concave defect occurs, and when a negative peak value is obtained, a convex defect occurs. Note that these peak values are converted by an A / D conversion circuit (A / D) 417 and input to the memory 406. The memory 406 is the memory 406 in FIG.
The defect detection circuit 416 is the same as the defect detection circuit 40 shown in FIG.
Here, there are hysteresis comparators on each of the positive side and the negative side, and when the level is higher than a predetermined level on the positive side and below the predetermined level on the negative side. In both cases, the detection signal DET is generated, and the detection signal DET falls when the integrated value becomes equal to or less than a predetermined value on the positive side and equal to or more than the predetermined value on the negative side, and the write signal WR is generated accordingly. Then, the value of the integrating amplifier circuit 414 and the value of the peak detection / sample hold circuit 415 are reset as the reset signal RS by delaying the write signal WR for a predetermined period. In this manner, the concave defect and the convex defect can be detected separately by their polarities, and the data of the depth and size can be obtained at the peak value. Here, the peak position of the integrated signal corresponds to the center position of the defect. Therefore, instead of generating a write control signal by the defect detection circuit 416, the write control signal WR is output from the peak detection / sample hold circuit 415 in accordance with the peak position.
May be generated.

The output of the integrating amplifier 414 is input to comparators 418a and 418b, and is compared with a predetermined threshold. The comparator 418a is for the positive integration signal shown in FIG. 4C, and the comparator 418b is for the negative integration signal shown in FIG. 5C. 4 (c) and FIG. 5 respectively.
As shown in (c), it is compared with the threshold value TH or -TH. As a result, as shown in FIG. 4D or FIG. 5D, a pulse having a positive or negative pulse width is output as the pulse width. This is received by the clock CLK from the clock generation circuit 420 and counted by the counter 419 to obtain the size of the defect as a count value. This count value is converted into an A / D conversion circuit (A / D
D) Data processing device 44 together with the A / D converted value of 417
To enter. The size of the defect may be determined by counting a signal indicating the coordinate position in the θ direction of the rotary encoder 23 instead of the clock CLK.

As described above, the configuration of the defect detection units 401 and 410 in the embodiment is merely an example, and the reflected light received by the light receiving element oscillates right and left in the stripe pattern corresponding to the defects of the concave and convex portions. If the detection signal is obtained from the adjacent light receiving element, and the detection signal of the difference according to the shift to the right or left is obtained, the processing of the subsequent detection signal can be freely designed. Of course, it is not limited to a circuit. The shape of the stripe pattern for obtaining the difference between the adjacent light receiving elements as a detection signal is described in the case of an increase in the amount of received light. If an inclined light shielding pattern is used as the pattern, either the increase in the amount of received light or the decrease in the amount of received light may be used. In the embodiment of FIG. 6, the detection values on the odd side and the even side are combined respectively in order to classify the irregularity defect. However, when simply detecting a gentle dish-shaped defect, either the odd side or the even side is used. Needless to say, only one of them may be combined to obtain a detection signal.

In the embodiment, a filter having a stripe pattern shifted at the center in the arrangement direction of the arranged light receiving elements is provided. This stripe pattern filter is divided into two at the center and corresponds to each. The light receiving element may be an individual element divided into two. In the embodiment, an example of a laser beam is used as the irradiation light for irradiating the inspection surface of the disk. However, when a laser beam is used, an S-polarized laser beam is particularly preferably used. However, the present invention is not limited to this laser beam, and it goes without saying that the irradiation light may be white light. Also,
Although the embodiments have been described mainly with respect to a device for inspecting a surface defect of a disk, the present invention is not limited to a disk, and it is needless to say that the present invention can be applied to inspection of an LCD substrate or the like. Further, in the embodiment, the scanning in the Rθ direction has been described, but it goes without saying that the scanning may be an XY two-dimensional scanning.

[0036]

As described above, according to the present invention, since the relationship between transmission and shading is reversed in the light receiving elements adjacent to the right side and the left side substantially from the center of the light receiving element, the defect is eliminated. When the reflected light from the surface in the case where there is no light is substantially radiated to the central part, there is no difference in the amount of received light between adjacent elements, but when there is a defect on the surface, the reflected light from the center of the light receiving element Since the light beam swings to the left or right, a difference occurs in the amount of received light between adjacent elements, which becomes a defect detection signal. In this case, detection is performed based on a difference in light reception level between adjacent elements, and is not a comparison between the level of scattered light or reflected light and a reference level. Defect detection can be performed with high accuracy. Further, in the second invention, since a detection signal is obtained by adding a large number of signals corresponding to the difference between the adjacent light receiving elements, the blank signal (dead zone) between the light receiving elements is not affected. It is possible to detect gentle irregularities such as dish-shaped defects, which are difficult to detect only by the difference between adjacent light receiving elements. In this case, in particular, the detection value of the difference between the adjacent light receiving elements is divided into odd and even numbers, the difference between the next adjacent light receiving element is taken, the difference is added, and then the integrated value is obtained. can do. As a result, it is easier to adjust the detection sensitivity than when directly detecting the intensity of reflected light or detecting the intensity of scattered light, and it is possible to detect minute unevenness defects. Classification accuracy can be improved.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of a surface defect inspection apparatus to which the present invention is applied.

FIG. 2 is a development explanatory view of a detection optical system in a direction parallel to the plane of the drawing (θ direction) in the drawing and a direction perpendicular to the plane of the drawing (R direction) in the drawing (R direction).

FIG. 3 is an explanatory diagram of a relationship between a stripe pattern filter and an APD array sensor.

FIG. 4 is an explanatory diagram of a waveform detected by regular reflection light of a concave defect.

FIG. 5 is an explanatory diagram of a waveform detected by regular reflection light of a convex defect.

FIG. 6 is a block diagram of another embodiment of the surface defect inspection apparatus to which the present invention is applied.

FIG. 7 is an explanatory diagram of a configuration of a main part of a conventional magnetic disk surface defect inspection apparatus.

FIG. 8 is an explanatory diagram of shapes of various defects on a disk.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Disc, 2 ... Rotating mechanism, 10 ... Surface defect inspection apparatus, 21 ... Spindle, 22 ... Motor, 3,50 ... Detection optical system, 31,51 ... Emission system, 311,511 ... Laser light source, 512 ... Beam Expander, 513: Cylindrical lens, 514: Focusing lens, 52
1: Objective lens, 522: Stripe pattern filter, 523
... imaging lens, 524 ... APD array sensor, 32,5
2: light receiving system, 400, 400a: defect detector, 410a
To 410n: differential amplifier, 411, 412: addition and synthesis circuit, 413: inverting amplifier, 414: integration amplifier circuit, 41
5 ... Peak detection circuit, 415 ... Sample hold circuit,
417 A / D conversion circuit (A / D), 4 Surface defect detection processing unit, 41 Signal processing circuit, 42 MPU, 43 Memory, 44 Data processing device, 45 Printer (P
R), 46 ... display, 100 ... surface defect inspection apparatus, L _T ... laser beam, S _P ... laser spot, phi ₁ ...
The minor axis of the spot, φ ₂ : the major axis of the spot, F: the defect.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2G051 AA71 AB02 AB07 BA08 BA10 BA11 BB09 CA03 CA07 CB01 CC07 DA07 DA08 EA03 EA05 EA08 EA09 EA14 EB01 EC01 EC03 5D112 AA24 BA03 GA19 JJ05 JJ09

Claims (9)

[Claims] 1. A defect detection optical system which scans the surface of a face plate with a light beam, receives light reflected from the surface by the light beam by a light receiver, and the light receiver generates a signal for defect detection. A light projection system that irradiates the light beam having a width in a direction perpendicular to the main scanning direction and relatively scans the face plate, and receives reflected light from the face plate arranged in the perpendicular direction a light receiving system having n light receiving elements having n light receiving elements (where n is an integer of 2 or more), and a light receiving system for forming an image of the scanning position of the face plate on the n light receiving elements; A light receiving surface of the light receiving element, a stripe pattern filter in which the relationship of transmission or light blocking is substantially reversed between right and left adjacent light receiving elements from the center, and a detection signal corresponding to a difference in the amount of light received by the adjacent light receiving element. For defect detection A defect detection optical system, wherein the reflected light from the surface when there is no defect is substantially irradiated to the central portion. 2. The method according to claim 1, wherein n is 4 or more,
An average value calculation circuit that calculates an average value of the levels of the respective electric signals obtained from the light receiving elements adjacent to the front and rear thereof in the arrangement direction of the light receiving elements, and the level of the electric signal from the light receiving elements of the self. 2. The defect detection optical system according to claim 1, wherein a difference from a level calculated by the average value calculation circuit is used as the detection signal. 3. The apparatus further comprises an integration circuit in both positive and negative directions and a peak detection circuit in both positive and negative directions, wherein the detection signal is integrated by the integration circuit and a peak is detected by the peak detection circuit. 3. The defect detection optical system according to claim 2, wherein the defect is detected and the concave defect and the convex defect are detected separately according to the peak value and the positive and negative polarities thereof. 4. The light-receiving device according to claim 1, wherein the face plate is a disk, and the stripe pattern filter has transmission and light-shielding stripe patterns alternately corresponding to the width of the light-receiving elements in the arrangement direction, and is substantially at the center of each of the light-receiving elements. Where the left and right patterns are shifted in the array direction by the width of the light receiving elements in the array direction, and the detection circuit receives an electric signal from the adjacent light receiving element and outputs a signal according to the difference, 2. The defect detection optical system according to claim 1, further comprising a plurality of differential amplifier circuits, receiving a signal corresponding to the difference from the differential amplifier circuits, and generating the detection signal obtained by adding the signals according to the difference. system. 5. The differential amplifier circuit of a number corresponding to each adjacent light receiving element pair, wherein signals corresponding to a difference between odd-numbered and even-numbered differential amplifier circuits are individually added. 5. The defect detection optical system according to claim 4, wherein the detection signal is generated by inverting and synthesizing a signal corresponding to a difference between the odd number and the even number. 6. A disk according to claim 1, wherein said disk is a magnetic disk or a glass substrate thereof, and said light projecting system is focused on a surface of said disk in a main scanning direction; 5. The defect detection optical system according to claim 4, wherein focusing is performed on a front side in a direction perpendicular to the direction. 7. A defect inspection apparatus which scans a surface of a face plate with a light beam, receives light reflected from the surface by the light beam by a light receiver, and the light receiver generates a signal for defect detection. A light projecting system that irradiates the light beam having a width in a direction perpendicular to the main scanning direction to relatively scan the face plate, and receives light reflected from the face plate and arranged in the perpendicular direction. A light receiving system having a plurality of light receiving elements (where n is an integer of 2 or more), a light receiving system for forming an image of the scanning position of the face plate on the n light receiving elements, and the light receiving system inserted into the light receiving system; A stripe pattern filter in which the relationship of transmission or light blocking is substantially reversed between light receiving elements adjacent to the right and left sides from the center of the light receiving surface of the element, and a detection signal corresponding to the difference in the amount of light received by the adjacent light receiving element is defective. For detection A defect detection circuit that generates a signal, and wherein the reflected light from the surface when there is no defect is substantially applied to the central portion. 8. The detection circuit according to claim 1, wherein n is 4 or more,
An average value calculation circuit that calculates an average value of the levels of the respective electric signals obtained from the light receiving elements adjacent to the front and rear thereof in the arrangement direction of the light receiving elements, and the level of the electric signal from the light receiving elements of the self. 8. The defect inspection apparatus according to claim 7, wherein a difference from a level calculated by the average value calculation circuit is used as the detection signal. 9. The light receiving device according to claim 1, wherein the face plate is a disk, and the fringe pattern filter has alternately transmissive and light-shielding fringe patterns corresponding to the width of the light receiving elements in the arrangement direction, and is substantially at the center of each of the light receiving elements. Where the left and right patterns are shifted in the array direction by the width of the light receiving elements in the array direction, and the detection circuit receives an electric signal from the adjacent light receiving element and outputs a signal according to the difference, 8. The defect inspection apparatus according to claim 7, comprising a plurality of differential amplifier circuits, receiving a signal corresponding to the difference from these differential amplifier circuits, and generating the detection signal obtained by adding a signal according to the difference. .