JPH0816674B2 - Ultrasonic flaw detector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION A. Field of Industrial Application The present invention relates to an ultrasonic flaw detector for detecting the presence or absence of a defect or the like in a subject using ultrasonic waves.

B. Conventional Technique FIG. 5 is a diagram showing a configuration of a conventional general ultrasonic flaw detector. In the figure, 1 is a water tank filled with water 2, 3 is a subject immersed in the bottom of the water tank 1, and 4 is a probe immersed in the water tank 1 facing the subject 3. Scanning is performed in arbitrary X, Y, and Z directions by a drive mechanism (not shown).

Reference numeral 5 is a pulse wave generator for applying a pulse voltage to the probe 4, 6 is a receiver for receiving and amplifying reflected waves reflected from the surface, inside and bottom of the subject 3 through the probe 4, and 7 is It is an A / D converter for A / D converting the reflected wave signal amplified by the receiver 6.

Reference numeral 8 is a microcomputer for inputting the reflected wave signal A / D converted by the A / D converter 7 and performing processing for displaying as an image and various other data processing, 10 is input image data from the microcomputer 8. An image input / output device that performs image processing and sends an image signal to the monitor 11,
A monitor 11 displays an image of the subject.

This ultrasonic flaw detector detects, for example, the peak value at each scanning point by amplifying the reflected wave from the subject 3 while scanning the probe 4 in the X and Y planes, and based on the peak value, an appropriate value is detected. It displays a two-dimensional image on the X and Y planes at the brightness level, and utilizes that the sound pressure level of the reflected wave correlates with the sound pressure reflectance at the surface, internal interface, and bottom of the subject 3. .

C. Problem to be Solved by the Invention As described above, the ultrasonic flaw detector has a silicon chip 12 used for a semiconductor device as shown in FIG. 3, for example.
It is used to detect defects such as cracks 14 in the. The wiring pattern 13 is formed on the silicon chip 12, and the width thereof is approximately several μm to several tens μm, so that the surface of the silicon chip 12 has minute irregularities.

Therefore, when inspecting the presence or absence of such a defect of the silicon chip by the conventional ultrasonic flaw detector, information of these silicon chip surface shapes is also included in the image obtained based on the reflected wave, There is a problem that defects such as cracks cannot be accurately determined.

A technical problem of the present invention is, for example, to accurately determine an internal defect of a subject having minute irregularities on its surface.

D. Means for Solving the Problem The present invention will be described with reference to FIG. 1 showing an embodiment. According to the present invention, an ultrasonic signal is projected from an ultrasonic probe 4 to a subject 3, It is applied to an ultrasonic flaw detector that detects defects in the subject 3 based on the reflected waves. And an analyzing means 8a for analyzing the frequency component of the two-dimensional original image data corresponding to the inspection region of the subject obtained based on the reflected wave,
Frequency calculating means for calculating a filtering frequency from the frequency components of the normal subject 3 thus obtained
8b and a filtering means 8c for filtering only an arbitrary frequency band from the original image data based on this filtering frequency.
And the subject 3 by the signal from the filtering means 8c.
The above-mentioned technical problem is achieved by detecting the defect of.

E. Action Analysis means 8a performs frequency analysis of original image data obtained by two-dimensionally scanning a normal subject (hereinafter referred to as pseudo subject) 3 and obtains its frequency component. The frequency calculating means 8b calculates a filtering frequency from the obtained frequency components. Next, the original image data obtained from the subject 3 requiring inspection is subjected to frequency analysis by the same procedure to obtain the frequency component. From this frequency component, the filtering means 8c filters only the frequency band equal to or lower than the filtering frequency obtained from the pseudo subject 3. Further, the signal from the filtering means 8c is used to detect the defect, for example, by displaying only an image relating to the defect of the subject 3.

It should be noted that, in the above-mentioned items D and E for explaining the configuration of the present invention, the drawings of the embodiments are used to make the present invention easy to understand, but the present invention is not limited to the embodiments.

F. Embodiment FIG. 1 is a diagram showing the configuration of an embodiment of the present invention.
In the figure, the same parts as those in FIG. 5 are designated by the same reference numerals, and the description thereof will be omitted.

The microcomputer 8 has the following functions added to the conventional device shown in FIG. It is a fast Fourier transform unit (hereinafter, FFT unit) 8a that analyzes the frequency components of the two-dimensional original image data corresponding to the examination region of the subject,
The frequency calculation unit 8b calculates a filtering frequency f from the frequency components of the original image data of the normal subject 3 prepared in advance, and the filter unit 8c filters only a frequency band equal to or lower than the filtering frequency f. Each of these functions 8a, 8b, 8c is realized by the program processing of the microcomputer 8. Reference numeral 9 is a frequency adjustment dial capable of arbitrarily changing the filtering frequency.

 Here, the filtering frequency f will be described.

As described above, the pattern 13 on the silicon chip 12 shown in FIG. 3 is fine, while the crack 14 extending from the surface to the inside may have a size of several μm, but the pattern is usually a pattern.
It is much larger than 12. Therefore, in the original image data obtained by ultrasonic flaw detection of the subject 3 as described above,
A high frequency component due to the unevenness of the fine pattern 13 and a low frequency component due to the large crack 14 are included.

FIG. 4 is a frequency spectrum in an arbitrary direction obtained by performing a two-dimensional FFT on the original image data obtained by scanning the subject 3 having an uneven surface like the silicon chip 12 shown in FIG. Is. In the figure, the frequency f0 indicates the fundamental frequency related to the pattern 13. Here, since the fundamental frequency of the low frequency component related to the crack 14 is lower than the fundamental frequency of the high frequency component related to the pattern 13, the high frequency component can be removed by filtering the spectrum of the original image at a frequency lower than the fundamental frequency f0. . So, in this example,
Approximately 3db from the signal strength at the fundamental frequency f0 of pattern 13
The lower frequency f of the two frequencies of the attenuated spectrum line is set as the filtering frequency for separating the high frequency component due to the pattern 13 and the low frequency component due to the crack 14. That is, the frequency component equal to or lower than the filtering frequency f is regarded as the component related to the defect such as the crack 14.

Next, the operation of the ultrasonic flaw detector thus constructed will be described with reference to FIG.

In step S101, first, the silicon chip 12 having no defects such as cracks is set as the pseudo object 3 in the present apparatus. Then, an ultrasonic wave is projected from the probe 4 to the subject 3 by the pulse signal from the pulse generator 5. The reflected waves reflected by the surface, the inside, and the bottom of the pseudo object 3 are again received by the probe 4 and sent to the receiver 6.
Further, this reflected wave is amplified by the receiver 6, converted into a digital signal by the A / D converter 7, and sent to the microcomputer 8. By scanning the probe 4 in, for example, the X and Y directions and sampling the reflected wave signal, the two-dimensional original image data of the subject 3 is accumulated in the microcomputer 8.

In step S102, the FFT unit 8a uses the original image data h (u,
Two-dimensional fast Fourier transform is performed on v) to obtain image data H (u, v).

here, Represents a two-dimensional fast Fourier transform.

This image data H (u, v) is a two-dimensional frequency spectrum image.

In step S103, it is determined whether the sample 3 is a non-defective pseudo sample or a sample requiring inspection. If it is a non-defective product, the process proceeds to step S107, and if not, the process proceeds to step S104. For example, when this step S103 is executed for the first time, by defining it as a pseudo subject, step S103
Can be determined.

In step S107, the frequency calculator 8b obtains the basic frequency f0 from the image data H (u, v), and further calculates the filtering frequency f based on the basic frequency f0 according to the procedure described above. The filtering frequency f thus obtained is sent to and stored in the filter unit 8c.

Next, the operation of the ultrasonic flaw detector for the subject 3 requiring inspection will be described.

The silicon chip 12 requiring inspection is set in this ultrasonic flaw detector. In step S101, the microcomputer 8 sets the original image data h
(U, v) is accumulated, and the process proceeds to step S102, where the FFT unit 8a performs a fast Fourier transform operation on this original image data, and 2
The three-dimensional frequency spectrum image data H (u, v) is obtained.

In step S103, it is determined whether the subject 3 is a non-defective product or a sample that needs inspection. Since the subject 3 needs to be inspected here, the process proceeds to step S104.

In step S104, the filter unit 8c performs a filtering process of filtering only the frequency component equal to or lower than the filtering frequency f obtained in the non-defective product inspection. This filter is the function L
If (u, v) and the image data after the filter processing is expressed by G (u, v), then G (u, v) = H (u, v) .L (u, v). Next, in Step S105, the FFT unit 8a performs a two-dimensional inverse fast Fourier transform on the image data G (u, v),
Image data g (u, v) is obtained.

here Represents a two-dimensional inverse fast Fourier transform.

Thereafter, in step S106, the microcomputer 8 transfers the data of the image data g (u, v) to the image input / output device 10, performs image processing, and displays it on the monitor 11. The image data g (u, v) displayed here includes only information about defects which is not affected by the unevenness of the surface of the silicon chip 12.

When the monitor image thus obtained is observed and an image showing a defect cannot be accurately obtained, the filtering frequency can be arbitrarily changed by the filtering adjustment dial 9, and the process can be returned to step S104 again for processing. For example, an image filtered by any frequency can be displayed.

In the configuration of the above embodiment, the FFT unit 8a is the analysis means,
The filter unit 8c constitutes each filtering means.

Based on the filtered image data G (u, v) or the image data g (u, v) obtained by further two-dimensional inverse Fourier transforming the image data, the presence or absence of a defect is detected without displaying the image on the monitor 11. Good. In addition to the silicon chip, the present invention can be applied to accurately detect defects not only for silicon chips but also for samples in which the spatial frequency components of internal defects and the spatial frequency components of the upper layer of internal defects are different. . In this case, there may be a sample or flaw detection condition in which the spatial frequency component of the internal defect is high.

G. Effect of the Invention As described above, according to the present invention, the original image data of a non-defective product is subjected to frequency analysis to obtain the filtering frequency from the frequency components related to the surface shape, and the original image data is filtered at this filtering frequency. Since it is possible to extract only the frequency component related to the defect of the subject,
For example, it is possible to accurately determine the defect of an object having an uneven surface.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration of an embodiment of an ultrasonic flaw detector of the present invention, FIG. 2 is a flow chart showing the operation thereof, and FIG. 3 is a sectional view of a silicon chip having an uneven surface and an internal defect. FIG. 4 is a diagram showing an example of a frequency spectrum obtained by frequency-analyzing the silicon chip, and FIG. 5 is a diagram showing a configuration of a conventional ultrasonic flaw detector. 1: Water tank, 2: Water 3: Subject, 4: Probe 5: Pulse generator, 6: Receiver 7: A / D converter 8: Microcomputer 8a: FFT section, 8b: Frequency calculation section 8c: Filter part, 9: Frequency adjustment dial 10: Image input / output device 11: Monitor, 12: Silicon chip 13: Pattern, 14: Crack

Claims (1)

[Claims] 1. An ultrasonic flaw detector for projecting an ultrasonic signal from an ultrasonic probe to a subject and detecting a defect in the subject based on the reflected wave, the ultrasonic flaw detection device comprising: The analyzing means for analyzing the frequency component of the obtained two-dimensional original image data, the frequency calculating means for calculating the filtering frequency from the frequency components of the normal subject obtained by the analyzing means, and the frequency calculating means A filtering means for filtering only an arbitrary frequency band from the original image data on the basis of the obtained filtering frequency, and detecting a defect of an object by a signal from the filtering means. Sonic flaw detector.