JP2715517B2 - Component mounting device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus that picks up a component to be mounted by a mount head and mounts the component at a predetermined position, and in particular, suctions a chip component by a suction nozzle. In addition, the present invention relates to an apparatus suitable for use in a chip placer mounted on a predetermined position on a circuit board.

[Summary of the Invention]

In a component mounter such as a chip placer, the pass / fail judgment of the suction position of the component is made by sensing the thickness of the chip component, and a thread for determining the normal suction using the sensing data at the time of mounting. The instability due to the temperature change and the aging change of the sensor is canceled by applying feedback to the shawl level using a statistical method.

[Conventional technology]

For example, a component mounting apparatus disclosed in Japanese Patent Application Laid-Open Nos. 62-165995 and 62-193199 is used for mounting chip components. In such a device,
A small electronic circuit component is supplied by tape or magazine, and a predetermined component is sucked by a suction nozzle of a mount head, and the mount head mounts the above component at a predetermined position on a circuit board. Like that. Such a component is configured such that its electrode is connected to a connection pattern on a circuit board by a solder dip or solder cream, thereby forming a predetermined electronic circuit.

[Problems to be solved by the invention]

In the component mounting apparatus as described above, a sensor for determining the attitude of the component when the chip component is suctioned by vacuum is required. Conventionally, a method has been used in which the dimension in the thickness direction of a component is measured to determine this attitude, and if the value is within a certain range, it is determined that the component is normal. At this time, the upper and lower limits of the normal range were created when the mount data was created, and did not change thereafter.

However, the output value of the sensor that measures the dimension in the thickness direction of the chip component drifts due to a change in ambient temperature conditions and due to deterioration of the sensor. Therefore, when the sensor is used to some extent, the actual output value of the sensor deviates from the threshold level in the normal range, and erroneous detection occurs. At present, for this measure, the upper and lower limits of the normal range have to be manually changed periodically or suddenly when the temperature changes drastically.

The present invention has been made in view of such a problem, and there is provided a component mounting apparatus that automatically changes an upper limit value and a lower limit value that define a limit of a normal range based on statistical processing. It is intended to provide.

[Means for solving the problem]

The present invention provides a mounting apparatus that picks up a component to be mounted by a mount head, measures a dimension of the component, and compares the dimension with a reference value to detect an abnormality in a pickup attitude. The means for sequentially correcting the reference value by using a statistical method based on the measured value of the dimension of the component in which the abnormality of the picked-up posture is detected by the comparison is provided.

[Action]

Therefore, according to the present invention, the reference value is successively corrected using a statistical method based on the measured value of the component size, so that when the output value of the sensor for measuring the component size drifts, The reference value is corrected according to the drift, and it is possible to remove instability of sensing such as drift.

〔Example〕

1 and 2 show an electronic circuit chip component mounting apparatus according to an embodiment of the present invention, that is, a mount head portion of a chip placer. A nozzle 11 is provided. The suction nozzle 11 suctions the chip component 12 by vacuum suction. The chip components 12 are supplied in a state of being arranged in the concave portions 14 of the tape 13 as shown in FIG.
The suction nozzle 12 is sucked by the suction nozzle 11 and is mounted at a predetermined position.

In order to detect the attitude of the chip component 12 sucked by the suction nozzle 11, a pair of sensors 17, 18 is used. These sensors 17 and 18 are light-emitting elements 1 composed of LEDs arranged so as to face each other.
It detects 9 and 20 lights. The sensors 17, 1
The arrangement of the sensor 17 and the light emitting element 19 is opposite to the arrangement of the sensor 18 and the light emitting element 20 so that 8 is not affected by the light of the light emitting elements 20 and 19, respectively. A pin 21 is provided on a side of the mount head 10, and the pin 21 is detected by the sensor 17.
On the other hand, the sensor 18 directly detects the chip component 12 held by the nozzle 11.

In the above configuration, if the suction nozzle 11 of the mount head 10 sucks the chip component 12, the mount head 10 moves up as shown by the arrow in FIG. Then, the pin 21 of the head 10 crosses the optical path from the light emitting element 19 to the sensor 17, and a pulse signal having a time width as shown in FIG. 3A according to the rising speed of the mount head 10 and the diameter of the pin 21. Is output from the sensor 17. At this time, the light of the light emitting element 20 is blocked by the suction nozzle 11 or the chip component 12. When the mount head 10 is further raised, the chip component 12 passes through the front surface of the sensor 18, and the sensor 18 generates an output as shown in FIG. 3B. This output is an output in the case of a normal suction posture. On the other hand, when the chip component 12 is sucked while lying down, as shown in FIG. 3C, a pulse signal having a timing different from that of the pulse signal in the normal suction posture is generated. Will be.

The computer 22 measures the time until a change in the level of the pulse shown in FIG. 3B or 3C based on the rising or falling of the pulse obtained by the first sensor 17. This time, for example, mount head 10
This is a value representing a value obtained by subtracting the height direction interval between the sensors 17 and 18 from the distance from the pin 21 of the electronic component 12 to the end of the passage of the electronic component 12, and the magnitude of the thickness of the chip component 12 in the moving direction of the mount head 10. Appears. Since the above amount of time differs depending on the ascending speed of the mount head 10, the pulse width shown in FIG. 3A may be read for correction, and each value divided by the time width may be compared. .

It can be considered that the data of the thickness in the height direction of the chip component 12 obtained by the sensors 17 and 18 has a normal distribution as shown in FIG. And in such a normal distribution,
84% between [-6σ, σ] and 16% between [σ, 6σ]
Has an appearance probability of. Therefore, this property can be used in correcting a value used as a criterion for determining a normal value and an abnormal value of the thickness of the chip component 12.

The output value of the sensor 18 at the time of mounting the chip component 12 at the nth time is Cn, the upper limit of the normal range after mounting the chip component 12 at the nth time is Hn, and the lower limit of the normal range is
Ln. The computer 22 makes corrections as shown in FIGS. 5 and 6 based on these data.

I C _{n + 1} > Hn, C _{n + 1} <Ln In this case, the output value of the sensor 18 is out of the normal range. Judge and do not give feedback of reference value. At this time, the chip component 12 is discarded.

II Ln <Cn _{+ 1} <(Hn−Ln) · 5/12 + Ln This range is, as shown in FIG.
12 when C _{n + 1} exists. In this case, the following correction is performed for the lower limit and the upper limit.

Ln _{+ 1} = Ln-(Hn-Ln) · 84 / a Hn _{+ 1} = Hn-(Hn-Ln) · 16 / a III (Hn-Ln) · 5/12 + Ln <Cn _{+ 1} <(Hn −Ln) · 7/12 + Ln This range is for the case where C _{n + 1} is between 5/12 and 7/12 between Ln and Hn, as shown in FIG. 4B. In such a case, the following correction is applied to the lower limit and the upper limit.

Ln _{+ 1} = Ln + (Hn-Ln) · 16 / a Hn _{+ 1} = Hn-(Hn-Ln) · 16 / a IV (Hn-Ln) · 7/12 + Ln <Cn _{+ 1} <Hn This range Is the case where C _{n + 1} is between 7/12 and _{1 above} Ln and Hn as shown in FIG. 4C, and the following corrections are applied to the lower limit and the upper limit, respectively. Add.

Ln _{+ 1} = Ln + (Hn−Ln) · 16 / a Hn + ₁ = Hn + (Hn−Ln) · 84 / a where a is a constant for determining the amount of feedback.

In this way, the computer 22 modifies the upper limit and the lower limit based on the output of the sensor 18. That is, when the output value of the sensor 18 is near the lower limit of the normal range and the appearance probability of the sensor 18 is high as in II (FIG. 4A), both the lower reference value and the upper reference value are lowered. Next, when the appearance probability is high in the center of the normal range as in III (FIG. 4B), the lower limit is raised and the correction is made to lower the upper limit. When the appearance probability is high near the upper limit of the normal range as in IV (FIG. 4C), both the lower limit and the upper limit are increased. Moreover, by using a statistical numerical value for weighting when correcting the upper and lower reference values, Ln converges to −6σ and Hn converges to 6σ. FIG. 6 is a flowchart showing such an operation, and FIG. 5 is a graph showing changes in the upper limit and the lower limit due to the correction. Note that the upper and lower reference values H ₀ and L ₀ at the time of the first mounting in FIG. 5 are set based on the data of the thickness measurement of the chip component 12 performed 20 times prior to the mounting. .

By using such a method, even when a drift occurs in the output value of the sensor 18, the reference value follows the drift, and the posture of the chip component 12 can be detected stably. By thus dynamically changing the reference value and canceling the instability of the sensors 17 and 18 due to a temperature drift or the like, the accuracy of standing adsorption sensing can be improved. Further, it is not necessary to reset the reference value due to the sensor deterioration, which is conventionally required, and the maintainability is improved.

〔The invention's effect〕

As described above, according to the present invention, a reference value is successively corrected using a statistical method based on a measured value of a component size. Therefore, when the output value of the detecting means such as a sensor for measuring the dimension of the component drifts, the reference value is corrected according to the drift. Therefore, even when the detection means drifts, the detection operation can be performed correctly, and the instability of the detection at the time of drift can be removed.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a component mounting apparatus according to an embodiment of the present invention, FIG. 2 is a perspective view of the main part, FIG. 3 is a waveform diagram showing a detection pulse of a sensor, and FIG. FIG. 5 is a graph showing the distribution of the measurement data, FIG. 5 is a graph of the correction of the upper limit value and the lower limit value, and FIG. 6 is a flowchart showing the correction operation by the computer. The names of the main parts in the drawings are as follows. 10 Mount head 11 Suction nozzle 12 Chip components 17, 18 Sensor 21 Pin 22 Computer

Claims (1)

(57) [Claims] 1. A mounting apparatus for picking up a part to be mounted by a mount head, measuring a dimension of the part, and comparing the dimension with a reference value to detect an abnormality in a pickup attitude. A component mounting means for sequentially correcting the reference value by using a statistical method based on a measured value of the dimension of the component in which an abnormality in the picked-up posture is detected by comparison with the component mount. apparatus.