JPS59128435A - Automatic inspecting system - Google Patents

Abstract

PURPOSE:To perform highly accurate automatic inspection at a high speed, by performing the frequency conversion of the sensor output with respect to a sample, which is located in the detecting range of a sensor, memorizing the maximum and minimum values within the pulse width, and comparing the values with a reference value after the sample has passed the detecting range. CONSTITUTION:A microcomputer 31 in a controller 26 corresponds to a position sensor 25. As long as a conveyed sample is located in the detecting range of a sensor, which receives reflected light, the pulse width of a pulse corresponding to the detected output of the sensor, through a voltage-frequency converter 28A, a signal insulator 29A such as a photocoupler, and an interval timer 30A, is received by the computer 31, and the maximum and the minimum values are stored. After the sample has passed the detecting range of the sensor, the computer 31 compares the stored contents with the similar reference pulse width through the sensor, which receives the reflected light from a reference surface, and the quality of the sample is judged. Therefore, the comparison with the reference value is not performed for every pulse but only once, and highly accurate automatic inspection can be performed at a high speed.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention 1 is an automatic inspection method in which an analog signal detected by a sensor is compared with a reference value while the specimen (tested object) passes through the sensor detection range to determine the quality of the specimen. Regarding.

In the inspection process during the receipt of raw materials in the manufacturing process and the shipping of products, the state of the sample that is continuously moving on a conveyor such as a belt conveyor is detected by some method, and the detected value is compared with the reference state. An automatic testing device is prepared that compares all samples and determines whether the sample is good or bad.

This type of automatic inspection device generally has the configuration shown in FIG. A sensor J detects the state of the sample J as an analog signal for the sample J transferred above the transport load/above, and detects that the sample J is transferred within the detection range of the sensor 3. It includes a position sensor and a controller S that receives detection signals from these sensors J and I and determines whether the specimen 2 is good or bad. After fully confirming that the sample is within the detection range of the sensor 3 using the detection signal from the controller sB position sensor 4, the analog signal from the sensor 3 is input to determine whether the sample is good or bad.

If a microcomputer is used as the control center for the controller S of such automatic inspection equipment, all the signals that can be processed by the microcomputer are digital signals, so it is necessary to have a circuit that converts the analog signal from the sensor 3 into a digital signal. It will be configured as follows. In general, when signals from various processes are input to the measurement side A1 equipment (C), analog signals are input to the measurement control equipment to avoid noise, etc., so as to insulate the signals from the process into the measurement control equipment. Input via isolator.

As mentioned above, when a microcomputer inputs an external analog signal without signal isolation, there is a method for signal isolation shown in FIG. 2 that is compact, low cost, and has a simple configuration. The analog signal from the outside is converted into a frequency signal by the voltage-frequency conversion circuit B, and this frequency signal is taken into the interval timer via the signal isolator 7 that electrically insulates it, such as by optical signal conversion using a photocoupler. g measures the frequency of the frequency signal, and from this measurement signal the microcomputer 2 calculates a numerical value corresponding to the input analog signal voltage.

As shown in FIG. 3(a), the voltage-frequency conversion circuit 6tI'i integrates the analog signals vI and I with a time constant defined by a capacitor C8 and a resistor P to perform a constant current process. The integrator // charges the integrating capacitor c1 '1 with the constant current source 10, and the output voltage of the integrator l/. a comparator/2 that detects that the signal has reached zero, and a one that is triggered by the detection output of the comparator/2 to generate a constant width pulse as the frequency conversion output fO, and connects the constant current source 10 to the integrator//. A shot circuit/3 is provided.

In this configuration, the input of comparator/2 is shown in the waveform diagram shown in FIG. 3(b). When T reaches zero, the one-shot circuit operates, and its output fol''l continues for a certain period of time T.
At the same time, the constant current source 10 is connected to the integrator/
From this time, the capacitor C1 of the integrator // starts charging, and the time T of the one-shot circuit /3 starts.
During this period, the integrator output vc rises with a slope of vIyaw=(Eng.-R)/Ct. As the time T elapses, the output fO becomes low level, the constant current source 70 is disconnected, and from this point on, the capacitor c1 begins to discharge with a slope of . When becomes zero, the one-shot circuit 13 again goes into a repetitive operation in which it goes high for a time T.

If the high level time T and low level time t of the output fo are constant, then if the analog input V, is constant, the charging voltage and discharging voltage of the capacitor c1 are the same, so ((Eng.,) / C
I) Then, the analog human power V + N k can be calculated from the calculation using the above equation (1) using T. At this time, it is necessary to isolate the signal at the frequency signal stage of the output fo, and the signal isolation at the stage of the analog signal. The circuit can be simplified compared to the above.

By the way, when comparing the external analog signal v with the input reference value V in the automatic inspection equipment #,
Generally, the operation of inputting V and vse respectively and comparing and determining is repeated, but when inputting an analog signal using the circuit system shown in Figure 2 above, the time t'l in equation (1) is measured. In addition to the time required, the time required for calculation based on equation (1) is required, and the analog signals v, l v,
Therefore, it is not possible to shorten the repetition period of the comparison and discrimination operation, and there is a problem that high-speed measurement cannot be performed in such an automatic inspection system.

The present invention continuously captures only the voltage-frequency converted pulse width signal while the sample is within the sensor detection range, and stores only the maximum or minimum value of the pulse width. The object of the present invention is to provide an automatic inspection method that enables accurate high-floor measurement by calculating an input value and inputting a reference value immediately after passing through a detection range and comparing the two.

The present invention measures the pulse width of the frequency signal corresponding to the analog input voltage by voltage-frequency conversion, frequency signal isolation, and pulse width detection using an interval timer to input the analog signal for detecting the state of the specimen, and performs a predetermined calculation. The input voltage is calculated based on a formula, but there is a difference from conventional devices in how analog signals are compared and judged. That is, when all analog signals are input, the pulse width of the frequency signal corresponding to the analog signal is measured, and the input voltage is measured based on this f'Lk.
In the present invention, the position sensor detects that the sample is within the detection range of the sensor, and during this time only the pulse width is continuously measured, and only the maximum or minimum value of these measured values is stored. . This maximum value.

The minimum value is the maximum value of the conversion pulse width from equation (1) described in ifT.

Corresponding to the minimum value, calculating the analog signal voltage based on this maximum value or minimum value means calculating the minimum value or maximum value of the detected value of the state of the specimen, and this calculated value and input By comparing standard values, determining whether a specimen is good or bad can be done by comparing one value representative of the condition of the specimen once, compared to comparing the detected values of each part of the specimen multiple times. This means that the number of determinations can be significantly reduced.

In general, the time the sample remains in one area horizontally to the sensor is short, and the reference value does not change significantly during that time.
Even in the above-mentioned test according to the present invention, the state of the specimen can be accurately determined, and there is no need to calculate the input voltage each time the magnitude of the reference value and the analog signal from the specimen is determined, and the micro The processing time by the computer can be reduced, and the time required for the specimen to pass through the detection range of the sensor can be shortened, allowing for high-speed measurement or improved accuracy.

The difference between the method of the present invention and the conventional method is shown in a flowchart as shown in FIG. Detection analog value of specimen condition■,
If the conversion pulse widths for the detection reference values V, , and the respective values VP and Vs are tp and is, then in the conventional method shown in FIG. ．． , 1. measurement (step S,
2,54t) and calculation of vP + v (step S, ?
. The quality of the specimen was judged 101 (steps s, r) based on the ratio axis result of vs.

In contrast, in the present invention, as shown in FIG. 4(b), when the specimen enters the detection range (step S // ), 1
, measurement (step S/!) followed by t.

Step S: Extract the maximum value (or minimum value) of the measured values, store and update the maximum value (Step S/3), and repeat this operation until the specimen is out of the detection range.
/&), Measurement of t for the first time when the sample is out of the detection range (Step S/!;), Calculation of Vi (Step S
l&), V, (step El /1) and v.

A comparative judgment is made between and V (step s/8). As a result, if the amount of time that the analyte remains within the detection range of the sensor is the same, the present invention can perform more comparisons of fd analog signals, that is, more comparisons of pulse widths corresponding to the analog signals than in conventional methods. Accurate measurements can be made to detect even small abnormalities in the specimen12, and the method of the present invention can increase the movement speed of the specimen to obtain the same accuracy.

FIG. 5 is an apparatus configuration diagram showing an embodiment of the present invention, in which a specimen is continuously transferred on the transport device and is removed if there is a scratch on the surface of the specimen. The case where it is applied to the intended automatic inspection equipment is shown. AC-excited light source lamp 2/ irradiates the detection surface of the specimen 2 with light when the specimen 2 is transported within a predetermined range, and a reference object 72 having a comparison reference surface fixed in the vicinity of the detection range of the specimen. irradiate with light. The light sensor ring is a light source lamp, and the light from 2/ is the sample 2.
The light diffusely reflected by the scratches on the surface is captured and an electrical signal corresponding to the intensity of the diffusely reflected light is obtained. Similarly, the optical sensor 2da detects diffusely reflected light due to the surface roughness of the reference object 22 as an electrical signal. The position sensor 5 detects whether the specimen is within the detection range of the optical sensor ring. The controller 26 determines whether the specimen is good or not (strength of scratches) from the analog signal of the optical sensor 23゜J on the condition that the specimen is within the detection range from the detection signal of the position sensor Δ, and if the specimen is defective. controls the entire drive of the remover 27 to cause the remover 27 to remove defective samples. In addition, the light source lamp-27
The ljJ magnet is used to eliminate the shadow of external light (for example, illumination light) on the optical input of the optical sensor), 3.2'l,
The sense amplifier in the optical sensor υ1w extracts only the excitation frequency component of the light source lamp 27 from the detection signal, and outputs it to the controller 26 as an amplitude indicating the presence or absence of scratches at a DC analog level.

The controller 26 has the configuration shown in FIG.

The detection signal (■) of the optical sensor ring and the detection signal (■) of the optical sensor 1tl- are voltage-frequency conversion circuits 21A and 211, respectively.
These converted signals are insulated by signal isolators 29 and 29B, respectively, and sent to interval timers 30A and 30B.
The width of the curve is detected. interval timer 30
A.

? The rain detection signal from the OB is input to the microcomputer 71 on the condition that the position sensor ivy is detecting a specimen. Eliminator 27 according to
Give operation commands to.

In such a configuration, the detection signals V, , V, from the surface of the specimen and the surface of the reference object, and the conversion nodal width t,.

t and have the relationships shown in the following equations (2) and (3), respectively.

7・='τso7 °−=°(2) v,=□ ・・・・・・(3) 1 + tJT@ Here, Ks + Ts I Kp and Tp are constants of each analog input means. For example, K, ,'rp=lOV.

If Tp = lOOμ8, when VP = I V (7), K t p = 900 μs and V, )LV (7
) when 't'i tp <900 μs. Therefore, the microcomputer 11 uses the interval timer 30
The time required for two people to use Af is about 1 millisecond or less. For kneading, from equation (2), v
It is normal for it to take about 10 millimeters to calculate pe according to the program on the microcomputer 71. Taking these times into consideration, Figure 4 (a)
The time required for one measurement using the conventional method shown in ) is approximately (1 ms + LOms) x 2 = 22 ms, and the input period of the analog signal cannot be shortened beyond this time. If the time that the specimen is in the detection range of the sensor is 50 milliseconds, the state of the specimen can only be inspected twice (at two locations).In contrast, in the method of the present invention, As shown in FIG. 4(b), Fl t P while the specimen is within the detection range of the sensor.
Only the minimum value (or maximum value) between them is memorized and updated, and when the sample goes out of the detection range, 1. V calculated from and t.

Since the calculated values of (1) and (2) based on the input are compared for the first time, the input period of the analog signal V, becomes approximately equal to the time for measuring tre, and can be shortened to, for example, about 1 millisecond in the above example. In other words, what is the state of the specimen during the 50 milliseconds before it leaves the detection range? This means that the state can be detected over the entire surface of the detection surface by applying a strange amount of force about 50 times, and the detection accuracy can be greatly improved.If the detection accuracy is the same as before, it will be possible to measure at least 20 times faster. means.

Note that the time width signal t corresponding to the input of the analog signal vP
, to update the memory of not only one minimum value (or maximum value) but also several values close to the minimum value, and when the minimum value is extremely smaller than other memory candy, Disregarding C as the minimum value due to noise and adopting the next smallest value or average value as the minimum value increases the influence of disturbances such as sudden noises such as lamp light and spike noise on the power line. Enables highly reliable inspection without removing

As described above, according to the present invention, while the specimen is within the sensor detection range, only the maximum value or minimum value of the pulse width signal corresponding to the detection signal of the sensor is updated in the entire memory history, and the specimen is within the sensor detection range. When the sample is removed, the analog signal level is calculated, the reference value is input, and the two are compared to determine whether the sample is good or bad.This has the effect of improving the detection accuracy of the sample condition and enabling high-speed measurement. .

[Brief explanation of drawings]

FIG. 1 is a block diagram of a conventional automatic inspection device, FIG. 2 is a block diagram of a controller in FIG. 1, and FIG. 3 is a block diagram (a) of a voltage-frequency converter and a waveform diagram (b) of essential parts. FIG. 4 is a conventional flowchart showing signal processing in a controller (a) and a flowchart of the present invention) (b),
FIG. 5 is a configuration diagram showing one embodiment of the present invention, and FIG.
It is a block diagram of the controller in the figure. /...transport device, λ...sample, t...voltage-frequency converter, n...signal isolator, g...interval timer, ? ...Microcomputer, IO...Constant current source, /l...Integrator 5, /12...Comparator, /3...One-shot circuit 1, 2/...Light source lamp, U...・Reference object 1. ZJ, 7.4'...
Optical sensor, JS...Position sensor, 2A...Controller 1.27...Exclusion impurity, JA, -zg13...
・Voltage-frequency converter! 9A.

Claims (1)

[Claims] A sensor that detects the state of the inspected object being transported and outputs an analog signal, a position sensor that detects whether the inspected object is within the detection range of the sensor, and a position sensor that detects whether the inspected object is within the detection range of the sensor. a controller that compares the detection signal of the sensor with a predetermined reference value to determine whether the object to be inspected is good or bad when the object is within the detection range of the sensor, and the controller converts the analog signal from the sensor to a corresponding frequency. The automatic inspection device converts the analog signal into a pulse signal with While the object is within the detection range, the width signal of the pulse signal is continuously input and its maximum or minimum value is memorized and updated, so that the object to be inspected is outside the detection range of the sensor.
1. An automatic inspection method characterized in that when an object is inspected, an analog signal level is calculated from the maximum value or the minimum value, and the quality of the object to be examined is determined by comparing the calculated value with the reference value.