JPS63228285A - How to check the number of holes on a board - Google Patents

Abstract

PURPOSE:To correctly detect a failure such as the partial jamming of a hole by dividing a hole dimension signal into different set hole dimension divisions of plural steps and respectively inspecting the number of the holes. CONSTITUTION:A substrate 1 to be inspected is travelled at a substantially constant speed, the hole dimension signal of the cross direction of the substrate to be inspected is generated by one dimensional image sensor constituting of a light source 21 disposed orthogonally with the traveling direction, a rod lens 22, a CCD sensor 23 to calculate the number of the holes of the substrate to be inspected based on the hole dimension signal. Then, the hole dimension signal is divided by a sorting circuit group 50 constituted of the three sorting circuits 51-53 of similar constitutions having the previously set different set hole dimension divisions and the number of the holes of the substrate to be inspected for every set hole dimension division is simultaneously calculated. Thereby, the failure such as the partial jamming can be discriminated.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a method for inspecting whether a predetermined number of holes have been drilled in a substrate.

1. During the process of drilling holes for mounting components on a board, the holes may become clogged with chips, but if these are sent directly to the next process, an automatic component mounting machine, the legs of the components may break. etc. may occur. For this reason, visual inspections have been performed on each board individually, but in recent years, hole diameters have become smaller due to high-density packaging and miniaturization of components, making it difficult to find defective holes by visual inspection. To solve this problem, it is possible to use optoelectronic technology instead of human eyes, which will be explained below.

In FIG. 7, reference numeral 10 denotes a conveying device, which consists of horizontally spaced conveyors 11, 12 and their drive motors (not shown).
A substrate 1 is supplied from the left onto the substrate 1 and is run rightward at a substantially constant speed. A contact type one-dimensional image sensor 20 is placed between the two conveyors 11 and 12, and a light source 21 is placed opposite to the substrate 1 in its width direction.
It consists of a rod lens 22, a CCD sensor 23, and a CC
Each pixel of the D sensor 23 is scanned at predetermined intervals to extract the output change occurring in the pixel at the position facing the hole in the substrate 1, and the next binarization circuit 30 outputs a high level (opposed to the hole),
Conversion to a low-level (not facing the hole) hole dimension signal is performed. FIG. 8(A) shows the relationship between the scanning of the CCD sensor 23 and the hole size signal. When the hole 2 of the substrate 1 is scanned from A to E, hole 2
Binarized signals of ae corresponding to each A-E, which are at a high level, are formed for each position facing the A-E. However, substrate 1
In order to detect the upper hole without omission, the traveling speed of the substrate may be set so that at least one scan is performed when the hole with the smallest diameter on the substrate faces the CCD sensor 23. However, In this case, the scanning circuit will differ depending on the hole diameter, and simply counting the number of high-level holes in the hole size signal will not match the number of holes. (high levels of times will be detected). Therefore, we devised a counting circuit 41 having a function of determining the number of holes, which will be explained with reference to FIG.
A first counter 413 that counts the hole size signal from the CCD sensor 23, a 1-line delay circuit 411 that delays the hole size signal by one line of the CCD sensor 23, and a hole size signal from the delayed hole size signal and the binarization circuit 30. an AND circuit 412, a second counter 414 that counts the output of the AND circuit 412, and a subtracter 415 that calculates and accumulates the difference between the counts of the first and second counters 413 and 414. Become. In this case, the first counter 413 counts the hole size signals for each scan ("3" in the example of FIG. 8), and the second counter 414 counts the hole size signals for each scan. The output is counted (“2” in the example of FIG. 8),
The subtractor 415 calculates the difference between the numerical values of both needles, and a subtraction result of "1" corresponding to the number of holes is obtained for each hole, regardless of the scanning circuit on the hole. This is then accumulated for each hole. Therefore, the number of holes is determined from the output of the subtracter 415 in the counting circuit 41 while the front end of the board 1 passes the position of the image sensor 20, and by comparing it with the number of holes of a normal board, the number of holes to be inspected is determined. It becomes clear whether the board passes or fails.

Problems to be Solved by the Invention By the way, when examining the state of hole clogging in substrates, not only the holes are completely clogged, but also there are cases in which only a portion of the hole is clogged. In such a case, there is a risk that when light passes through the non-clogged portion, it may be detected as a hole. For example, as shown in FIG. 8, if the middle part of the hole 2 is clogged with chips 3 and there are non-clogged parts on both sides, the scans B', C', and D' by the one-dimensional image sensor are When the processing is performed on the upstream side, above the chip 3, and downstream side of The holes are generated in the order of the first one and the second one having a high level, and as a result, the counting circuit 41 counts this as the number of holes "2". Therefore, in this case, if there are other clogged holes, the number of detected holes on the entire board will match the normal value, resulting in erroneous judgment.

As an attempt to solve this problem, it is conceivable to disable counting when the time width of the high level of the hole size signal is smaller than a set value. In other words, set this value to be slightly smaller than the diameter of the minimum hole (and if a part of the hole near the center is clogged, do not count it as a hole. However, in a hole with a diameter sufficiently larger than the set value, even if the center area is clogged as shown in Figure 9 (a), the chord length in other parts will be larger than the set value, and eventually the hole will become clogged. The above problems are not resolved.

In order to solve the above-mentioned problem, the present invention makes the substrate to be inspected travel at a substantially constant speed, and uses a one-dimensional image sensor disposed in a direction perpendicular to the traveling direction of the substrate to be inspected. In a board hole number inspection method that generates a hole size signal in the width direction of the board and calculates the number of holes on the board to be inspected based on the hole size signal, the hole size signal is set in advance and a plurality of appropriately different set hole sizes are used. The number of holes in the board to be inspected for each set hole size classification is calculated at the same time. In this case, as shown in FIG. 9(a), there is a hole in the middle that is clogged, and it is determined that there are two holes.
And even if a hole with a small diameter corresponding to the chord length is clogged and the total number of holes with that small diameter matches the normal value, the set hole size classification of the clogged hole is determined as 2. The number of holes is not detected, and this result indicates that there are clogged holes.

Embodiment In FIG. 1 showing an example of an apparatus for carrying out the method of the present invention, a conveying device 10 consisting of belt conveyors 11, 12, etc. with the same numbers as in FIG. 7, a light source 21, a rod lens 22,
a one-dimensional image sensor 20 consisting of a CCD sensor 23;
A binarization circuit 30 that binarizes the output of the image sensor 20 into high and low levels and forms a hole size signal that is at a high level during the scanning time of pixels facing the hole is the same as that shown in FIG. Works similarly. However, in the present invention, since the classification is performed based on the hole size itself, the upper limit value V of the traveling speed of the substrate 1 is determined as follows so that the diameter portion of each hole is also scanned. That is, @2
Considering two scans across the diameter of hole 2 with radius r as shown in the figure, after scan A is performed,
By the next scan B, the substrate has traveled a distance Δy. Therefore, the traveling speed V of the substrate 1 is expressed as ■=Δy/l, where t is the scanning time of one line. At this time, in scans A and Bi, the hole size is detected with an error of ΔX with respect to the diameter, and the Δy is subject to the following restrictions due to the allowable error ΔX.

In this embodiment, the length of one pixel of the image sensor 20 is 62.5μ, which is the allowable error Δ×, and based on that and the scanning time of one line 2 + 11'3ee,
From the above relational expression, the upper limit of the traveling speed ■ is determined by the hole diameter r=o, 2
In the case of mm, about 4 ugliness/ll1in, r=IIIl111
In this case, it will be about 107 Ioin.

Next, 50 is a classification circuit group consisting of three similarly configured classification circuits 51 to 53.
1, as shown in Figure @3, an AND circuit 511 into which the hole size signal from the binarization circuit 30 and the scanning clock pulse from the image sensor 20 are introduced, and a counter 512 which counts the lock pulses through the AND circuit 511. , the counted value is compared with the set hole size classification value set in the setter 514, and while the former is large, the 7 lip-flop circuit 515
It consists of a comparator 513 that sends an output to the hole size signal, and a control section (not shown) that restores the counter 512 and flip 70 knob 515 to their initial state when the hole size signal changes from high to low level. In this case, while the hole size signal is at a high level, scanning clock pulses are counted and converted into a count value corresponding to the hole size. Start-up is performed, and then the output of the flip 70 tube circuit 515 is kept at a high level while the hole size signal changes from high to low level, and then the initial state low level is restored6. The signal is sent to a counting circuit 41 shown in FIG. Similarly, the other classification circuits 52 and 53 only have different hole size classification values set in their setters, and the outputs of the flip-flop circuits that become high level while the hole size signal exceeds the respective hole size classification values correspond to the corresponding ones. The signal is sent to the counting circuits 42 and 43 of the counting circuit group 40.

FIG. 5 shows the above relationship in waveforms, and shows the set hole size classification values of each of the classification circuits 51 to 53 as S, -S,
In contrast, the hole size signal from the binarization circuit 30 is as shown in the figure. If the fourth is smaller than 82 and the fourth is larger than 83, the classification circuit 51
In this case, since the count value of the counter 512 exceeds SI in any high level portion, an output is generated, and in the same way, in the classification circuits 52 and 53, the hole size signal,
Signals are generated at the first, fourth and fourth high level portions, respectively. Note that an appropriate value smaller than the detection minimum hole diameter is selected as the classification value S1, and S2. S is set to be smaller than the other detection hole diameters and larger than the detection hole diameters below it, respectively.

As described above, each of the classification circuits 51 to S3 performs sorting according to the width of the high level of the hole size signal, and its output is sent to each of the counting circuits 41 to 43 of the corresponding counting circuit group 40.

Each of the counting circuits 41 to 43 is exactly the same as that explained based on FIG. .

Now, three holes 2 are made on the board 1 as shown in Fig. 6 (a).
．． 2' and 2'' (assuming that the position of 2' is on the downstream side with respect to 2 and 2''), scanning is performed at positions A'' to F'', and as shown in (a), each scan When outputs a" to f" are formed from the classification circuit 51 in response to
First, when inputting b'', the count value 41 of the first counter 413
3 is "2", the count value of the second counter 414 is rOJ, and the count value of the subtractor 415 is "2-1".

Next, when inputting “c”, each count value is “3”,
The count value of the subtractor 415 is "2", and the difference between the two values is "2".
1'' and the above-mentioned subtraction result ``2'', resulting in ``3''. Next, when inputting ``d'', each count value is ``3'', ``3'' and the count value of the subtracter remains the previous value r3J, and when inputting <e'', the count value is rlJJIJ and the count value of the subtracter remains the same. The numerical value remains the same as "3" from the previous time, the count value when inputting "f" is "rOJ 汀O", and the subtraction result is kept as "3", thereby changing the set dimension classification value of the classification circuit 51. It turns out that the largest number to exceed is 3.

Hereinafter, each counting circuit 42.43 counts the number of holes exceeding the set hole size classification value of each corresponding classification circuit 52.53, and compares this with the large number of holes in each corresponding classification of the normal board. This will determine whether it is normal or defective.

That is, hole diameters d, , d2 . If there are n1+n2+ns holes in d, respectively, the classification circuits 51-
When the setting value of each setter 53 is selected to be slightly smaller than the respective hole diameter, the count value N1 of the counting circuit 41 becomes n 1 +
If it is larger or smaller than n2+n, it indicates that there is a defective hole, and even if N matches the normal value of n1+n2+ns, if the count value N2 of the counting circuit 42 is different from n2+n3, there is a defective hole. Yes, and also N1°N2
Even if the count value N3 of the counting circuit 43 is different from the normal value n, it indicates that there is a defective hole, and any two of the count values of the counting circuits 41 to 43 are different from the normal value n. If it is normal, it will also identify which diameter hole is defective.

In the above explanation, the case where the hole size signals are distributed into three stages is exemplified, but the number may be appropriately divided into a plurality of stages depending on the number of hole diameters of the substrate, etc.

The effects of the invention are as described above, and the present invention distributes hole size signals into multiple stages of different set hole size classifications and inspects the number of holes in each of them, so defects such as partial clogging of holes can be accurately detected. can be detected.

[Brief explanation of the drawing]

Fig. 1 is a structural diagram including a partial block diagram showing an example of an apparatus for implementing the present invention, Fig. 2 is a geometric relationship diagram showing the relationship between board running speed and hole dimension error, and Fig. 3 is classification circuit,
FIG. 4 is a block diagram of each counting circuit, the ms diagram is a classification circuit, FIG. 6 is a waveform diagram explaining the operation of each counting circuit, and FIG. 7 is a partial block diagram showing a conventional device. The block diagram shown in FIG. 8.9 is a waveform diagram showing hole scanning and hole size signals.

Claims (1)

[Claims] 1. Run the board to be inspected at a substantially constant speed, generate a hole dimension signal in the width direction of the board to be inspected using a one-dimensional image sensor arranged perpendicular to the running direction, and then move the board to be inspected based on the hole dimension signal. In a board hole number inspection method that calculates the number of holes in a board, the hole size signal is compared and distributed to a plurality of appropriately preset hole size classifications, and the hole size of the board to be inspected is determined for each set hole size classification. A method for inspecting the number of holes on a board, characterized by calculating the number of holes at the same time.