JPH06258176A - Automatic measuring system for functional inspection - Google Patents

Abstract

PURPOSE:To perform the automatic measurement for the noise amount, abnormal sound and the vibrating amount of a motor itself with the same system without judgment based on skilled inspeting person's experience and intuition in the conventional method. CONSTITUTION:An acceleration-type vibration sensor 23 is brought into contact with a bearing part 21 and a core part 22 of a motor to be measured 20 under the constant pressure, and vibration data are sampled. The sampled data are divided into a processing system for judging noise and abnormal sound (a) and a processing system for judging vibration (b). The data are transferred into a personal computer 31 through a 1/3 octave filter 28, a smoothing circuit 29 and an A/D converter 30. The data are judged based on the judging reference for quantity determination. The data are transferred into a displacement vibration meter 25 and a comparator 26 and judged. The integrated judged result of both data is obtained based on the judged result of the noise and the abnormal sound and the judged result of the vibration.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a noise amount inspection, an abnormal noise inspection and a vibration inspection of a motor which is judged to be a good product or a defective product based on the experience and intuition of a skilled worker in the manufacturing industry. The present invention relates to automatic measurement of a sensory test of a motor by determining whether or not the sensor is activated.

[0002]

2. Description of the Related Art In recent years, automated measuring technology has been developed in the manufacturing industry, and the sensory test of the above-mentioned contents also uses a commercially available measuring instrument to measure the noise amount and vibration amount of a motor or rotating object. Is automatically measured.

It is well known that what is conventionally judged by the amount of noise can be measured without being affected by the surroundings by using a vibration sensor. A method of measuring the noise amount of the measured motor collected by the vibration sensor will be described with reference to FIGS. 4 to 6.

FIG. 4 shows a block diagram of a conventional method, in which the noise amount of a relay is automatically measured. FIG. 5 also shows a block diagram of the conventional method, in which the noise amount of the measured motor is automatically measured. In Fig. 6, the automatically measured data is the frequency on the horizontal axis,
The vertical axis is the output voltage and is a graph.

As shown in FIG. 4, the relay 1 has an external power source 2
The contact point 3 operates by being energized by, and the output level of the impact sound when the contact 3 operates and the output level of the electromagnetic sound generated by the relay coil 8 when the external power source 2 is energized are collected by the vibration sensor 4 and charged. Amplified by amplifier 5,
Further, the comparator 6 makes a determination based on the level setting values of the non-defective product and the defective product.

Those exceeding the set value are displayed or output as defective products, and those within the set value are displayed or output as non-defective products. This automatic measurement is characterized by a relatively simple structure and is inexpensive.

In FIG. 5, the motor to be measured 10 is rotated by being supplied with power from the outside. The amount of noise is collected by the vibration sensor 11, frequency-analyzed by the FFT analyzer 13 via the charge amplifier 12, and the result is compared with the standard value set in advance on the personal computer 14 to display the judgment of good product or defective product. Is output.

FIG. 6 is a graph showing the frequency-analyzed data by the FFT analyzer 13 with the horizontal axis representing the frequency and the vertical axis representing the output voltage. The area d between arbitrary frequencies or the arbitrary frequency is shown. The peak value e is the comparison target.

[0009]

However, in the conventional automatic measurement system of the sensory test using the vibration sensor as described above, the noise of the relay coil shown in FIG. In the measurement system, the amount of noise collected and amplified by the vibration sensor is evaluated and evaluated based on the mere high and low levels of sound. For example, an abnormal sound of the electromagnetic noise of the relay coil 8, that is, an abnormality such as chattering or humming. It is difficult to determine the proper sound.

In the motor automatic measuring system shown in FIG.
The frequency is analyzed by the FT analyzer 13,
As can be seen from the graph of FIG. 6, since the judgment is made by comparing the areas of the standard values in which the levels of the non-defective product or the defective product are set, the total area is the standard value even if it is a part of the non-defective product. If there are more, it is determined to be defective.

In such a conventional measuring system, since the noise level of the sound and the area between arbitrary frequencies are used for the determination as the whole sound, the factor analysis of the sound itself is not performed, and the simply set non-defective product is used. Or it is only judged and evaluated by comparison with the limit value of defective products.

In addition to the high noise level, some of the sound defects include mechanical sliding noise, such as noise caused by contact with the bearing portion,
There is a sound when the rotor comes into contact with other parts, such as an outer case, a coil, and a stator.The electrical sound is a sound that is generated by the imbalance of the air gap between the rotor and the stator, and the whole motor. There are vibrations that may be affected and beats that are electromagnetic sounds from the core due to excessive electrical loading, and it is also necessary to determine whether or not the sound is abnormal due to the height of the sound or the strength of the ears. The determination of such an abnormal sound has a problem that the determination and evaluation are difficult with the above-described automatic measurement system.

The present invention has a data measuring unit and a processing unit for quantitatively determining whether the motor under test is a good product or a defective product, and automatically measures the noise amount, abnormal sound and vibration amount of the measured motor by the same system. It is intended to provide a system for doing so. It is another object of the present invention to provide an automatic measurement system provided with a judgment standard for quantitatively judging whether the motor under test is a good product or a defective product.

[0014]

In order to achieve the above object, the sensory test automatic measuring system of the present invention collects data from a portion of the motor to be measured which emits sound and vibration, that is, from a bearing portion of the motor to be measured. An acceleration type vibration sensor for detecting mechanical and electrical noise amount, abnormal noise and vibration amount from the motor core is provided, and a mixing amplifier for synthesizing data to process a plurality of data collected from these sensors, A bandpass filter of 1/3 octave that divides each of the synthesized data into a required frequency band, a smoothing circuit for converting the data into an effective value and a logarithm, and a data processing unit including an A / D conversion unit are provided. The amount of noise and abnormal noise and vibration of the motor to be measured are automatically measured by the data of each bandpass filter.

Further, the sensory inspection automatic measuring system of the present invention is provided with a judgment standard for quantitatively judging whether the motor under test is a good product or a defective product. Fold the characteristics and compare the data collected from the acceleration type vibration sensor in contact with the sound and vibration of the measured motor with each sampling taken for each band of the 1/3 octave filter The results are optimally evaluated in a vertical and horizontal matrix, and a quantitative judgment is derived.

[0016]

According to the present invention, with the above-described configuration, the data collected from the acceleration type vibration sensor is comprehensively determined by the original quantification determination criteria for the noise amount and abnormal noise of the measured motor bearing portion and the vibration amount of the motor core. Therefore, not only the conventional determination and evaluation at high and low levels of sound but also the sound and vibration of the motor alone can be determined and evaluated by the same system.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the overall flow of an automatic measurement system according to an embodiment of the present invention.
In FIG. 1, the acceleration type vibration sensor 23 is brought into contact with the bearing portion 21 and the core portion 22 of the motor 20 to be measured under a constant pressure, and the noise amount of the motor, the abnormal sound and the vibration amount are collected as vibration data. The collected data is divided into a processing system that processes vibration data to determine noise and abnormal noise of the motor and a processing system that determines vibration of the motor from the vibration data.

The processing system for determining the noise and abnormal noise of the motor is 1 from the charge amplifier 24 and the mixing amplifier 27.
/ 3 octave filter 28 and smoothing circuit 29, A /
Data is input to the personal computer 31 via the D conversion device 30. The personal computer 31 makes a determination based on the quantification determination standard for the motor to be measured as shown in FIG.
To the sequencer 32. On the other hand, the processing system for determining vibration is the displacement vibrometer 25 through the charge amplifier 24.
And the judgment result is inputted to the sequencer 32 by the comparator 26.

The operation of the above configuration will be described. When power (not shown) is supplied to the measured motor 20, the measured motor 20 is energized and rotates. At this time, the bearing and the motor core of the measured motor 20 are energized to generate a mechanical and electrical sound. This sound is collected by the acceleration type vibration sensor 23.

The noise amount and abnormal sound of the motor 20 to be measured are collected as two vibration data by the acceleration type vibration sensor 23 which is brought into contact with the bearing portions 21 at both ends of the motor shaft at an arbitrary constant pressure in order to detect them. , The vibration data respectively amplified by the charge amplifier 24 are input to the mixing amplifier 27, where the vibration data 2 collected by the bearing unit 21
The two vibration data are combined into one and output. The mixing amplifier 27 realizes weight reduction of the system by synthesizing a plurality of data. That is, each device in the process of processing data is prevented from being burdened by the number of sensors collecting the data, and the complexity and cost of the system are suppressed.

The synthesized vibration data is 15 pieces of bandpass filters from 500 Hz to 12,500 Hz of the 1/3 octave filter 28 and one piece that has been audibly corrected (corrected so as to be easily heard by the human ear). , Through the smoothing circuit 29 for converting the effective value, and further, the conversion from analog to digital and the processing by the personal computer 31 are simplified, and the same numerical value as the normal evaluation. The data is transferred from the I / O interface (not shown) to the personal computer 31 via the A / D conversion device 30 for performing logarithmic conversion that can be determined by. Here, by using the quantitative determination criteria shown in FIG. 2, it is possible to determine whether the product is a good product or a defective product.

The vibration amount of the motor 20 to be measured is the acceleration type vibration sensor 2 which is brought into contact with the side surface of the motor core at an arbitrary constant pressure.
Collected as vibration data by 3 and charge amplifier 2
Amplified by 4 and passed through a displacement vibrometer 25 and a comparator 26
Is determined as the amount of vibration of the motor. Since the vibration of the motor core is based on the power supply frequency, it is necessary to determine it in the low frequency range, and the wide range frequency characteristics that are characteristic of the acceleration type vibration sensor are inaccurate in this low frequency range. A displacement vibrometer 25 is used so that the acceleration type vibration data is square-integrated so as to be represented by a displacement amount.

The comparator 26 can easily set the upper limit to the standard (the lower limit is not necessary because it is a non-defective product) by converting the displacement amount into a voltage, and can judge the non-defective product and the defective product by comparison. This motor core vibration amount determination signal is
It is input to the sequencer 32 that controls the mechanical system of the system together with the output signal output from the personal computer 31 as a judgment signal of the noise amount and abnormal sound of the motor.

As described above, when both the determination result of the motor noise amount and the abnormal sound and the determination result of the vibration amount of the motor itself are the non-defective item, the sequencer 32 outputs the non-defective item as a comprehensive determination of this system. The system waits for the mechanical mechanism to be released and the next measured motor to be received. Also,
If one of the judgment results of the motor noise amount and abnormal sound is defective, the sequencer outputs the defective product as a comprehensive judgment of this system, and then the system and mechanical mechanism are released, and the next measured object is measured. Waiting for motor acceptance.

In FIG. 2, the quantification criterion is 1/3.
For each band pass filter of the octave filter, the standard value 35 for the measured motor is compared with the acquired data, and is further aggregated for each band pass filter to give a result as a horizontal comparison result 36. This comparison result 36 is the standard value 35.
Since it is a comparison between the captured data and the captured data, the number of OKs is totaled.

Since the standard value 35 is defined as the non-defective range of the measured motor of each bandpass filter, the numerical value which is the base of the standard value is calculated by obtaining the process capability from a population of non-defective products of the same type as the measured motor. Calculate. The standard value 35 is the average bar X of the population, and the upper limit process capability, that is, the standard deviation σ is 3
It is set by adding the multiplied value and multiplying the margin ratio 43 by α as necessary. That is, standard value 35 is bar X
It can be expressed by + 3σ × α, and σ is usually a real number.
By using α of the margin ratio 43 for a frequency pair having a characteristic of noise or abnormal sound, it is possible to set a reference value as an absolute determination condition. That is, σ can be made by using a real number of 1 or less in the case of tight judgment, and α can be made by using a real number of 1 or more in the case of loosening judgment.

Further, in the horizontal condition 37, for the judgment of tightness or looseness, all the taken-in data must be judged as non-defective products as a result of comparing the standard value and the measured motor data. Setting or tightening or loosening settings for the frequency band in which the levels of non-defective products and non-defective products are mixed is set to be tight or loose. 37, vertical tallying and conditions 39, 40.

FIG. 3 is a graph showing characteristics of each motor. Using this system, this graph collects data on good and bad motors of each model and graphs each bandpass filter. As an example, an excellent product 45 whose level is far lower than the standard value even with a good product motor, a high noise amount 46 even with a defective product, or a shearing sound 47 which is a mechanical sliding sound, or an electrical product. Characteristic of abnormal sound such as roaring sound 48 is found in each frequency band.

Whether all the motors have such characteristics has an influence on the materials used, the number of revolutions, and the characteristics such as torque. As shown in this graph, by grasping and using the process capability of the motor to be measured, such as how to use the setting of the margin ratio 43 and setting of the criterion for tightening or loosening in the characteristic frequency band. It is possible to reliably determine each defective product as defective. With respect to the abnormal sound, it is possible to take a cause and countermeasure depending on which frequency band has the characteristic.

For the horizontal determination of the determination criteria, the standard values of 16 of each bandpass filter 38 and the acquired data are compared, and the OK, that is, the number of non-defective products is compared with the horizontal condition 37. O as a result of comparison
Count the number of K. This OK, that is, the setting of the threshold value can be input externally from the keyboard of the personal computer.

The fetching cycle is set so that the fetched data is not collected in synchronism with the number of rotations of the motor, and the accuracy of the judgment improves as the fetched data increases, but the processing time and productivity are improved. In other words, it goes without saying that the optimum sampling quantity can be set based on the production takt time. Each of the 15 band-pass filters has a characteristic at each frequency of the motor to be measured, that is, enables individual evaluation for each frequency, and the individual evaluation is performed by vertical and horizontal matrix calculation as shown in FIG. The optimum condition setting and the determination result are obtained.

Further, the purpose of the audibility-corrected A range is to quantify the total noise amount of the motor to be measured in a human audible frequency band as a judgment condition other than the individual evaluation for each frequency. .

The 15 band-pass filters from 500 Hz to 12,500 Hz in FIG. 2 mainly evaluate the abnormal sound and the noise amount in each frequency band in a narrow range and emphasize the characteristics in each frequency band. The primary purpose is to derive a quantification determination by matrix calculation, and the subsequent A range filter is capable of making a relative evaluation of the overall noise amount and abnormal sound within the audible range of humans.

For the vertical judgment of the judgment standard, the standard values and the sampling data of 16 of each bandpass filter 38 are compared, and as a result, OK.
The total number of samples, and OK for each sampling
The number of OKs tabulated under the condition 40 for the number of is recounted. As a result, the vertical OK number 42 is determined as the condition setting for the sampling number. The same applies to the horizontal counting and determination.

If the determination results of the number 41 of horizontal OKs and the number 42 of vertical OKs are both OK, the sound of the measured motor is judged OK, that is, the non-defective product is judged. The determination of the sound of the measurement motor is NG, that is, the defective product determination is performed.

As described above, according to the sensory test automatic measuring system of the present embodiment, the collected data from the motor 20 to be measured is divided into 16 bandpass filters by the 1/3 octave filter 28, and each bandpass filter is divided. The characteristics of the noise amount and abnormal sound of the motor to be measured can be represented by the data of 1. and the good product or the bad product can be easily determined by comparison with the quantification determination standard.

By converting the vibration data from the sensor, which is in contact with the motor core with a constant pressure, from the acceleration to the displacement based on the characteristics of the sensor, it is possible to determine whether the vibration amount of the motor is good or bad in the same system. It can be carried out.

The quantification criterion is created for each target motor to be measured and registered in the personal computer 31 for each target model, so that model switching and production command correspondence can be facilitated. . Needless to say, the determination standard can be shared for the measured motors that have similar motor characteristics and used materials and that have similar bearing portions.

[0039]

As described above, according to the present invention, the characteristics of the data are expressed by dividing the data from the acceleration type vibration sensor for each band pass filter by the 1/3 octave filter, and the characteristics of the data. Can be determined as a good product or a defective product by the quantification criterion.

For those which cannot be judged by the pattern matching method or the above-mentioned conventional method, the limit non-defective level for each model is set by using such a system and the quantification judgment standard, and further, the mechanical sliding noise or the electric sound is generated. Various defective products can be determined by adding an abnormal sound level such as a target beat sound to the characteristic frequency band.

Since the noise amount and the abnormal noise of the motor and the vibration amount of the motor itself can be automatically measured by the same system, the judgment based on the experience and intuition of the conventional inspector is not required, and the quantitative sensory test is performed. It is possible to provide an automatic measurement system that can perform automatic measurement and can be used offline and online, and at the same time, save labor.

[Brief description of drawings]

FIG. 1 is a block diagram showing an overall flow of a sensory test automatic measurement system according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating a quantification determination standard of a sensory test automatic measurement system according to an embodiment of the present invention.

FIG. 3 is a graph showing characteristics of each measured motor for each frequency band.

FIG. 4 is a block diagram showing a conventional relay automatic measurement system.

FIG. 5 is a block diagram showing a conventional motor noise automatic measurement system.

FIG. 6 is a diagram showing an output waveform of an FFT analyzer of a measured motor by conventional motor noise automatic measurement.

[Explanation of symbols]

 21 Bearing part # 22 Motor core part 23 Acceleration type vibration sensor 24 Charge amplifier 25 Displacement vibrometer 26 Comparator 27 Mixing amplifier 28 1/3 octave filter 29 Smoothing circuit 30 A / D converter

Claims (2)

[Claims] 1. A mechanical device from a bearing of a motor to be measured,
A contact portion having an acceleration type vibration sensor for detecting an electric noise amount, an abnormal sound and a vibration amount from a motor core, and a mixing amplifier for synthesizing data for processing a plurality of data collected from the sensor , A filter section for dividing the synthesized data into necessary frequency bands, a smoothing circuit section for performing effective value conversion and logarithmic conversion, A
A sensory test automatic measurement system comprising a data processing unit including a / D conversion unit, and configured to quantitatively and automatically measure the noise amount and abnormal noise and vibration amount of a measured motor in the same system. 2. A determination criterion for quantitatively determining the noise amount and abnormal noise of the motor to be measured, the criterion including the process capability of a good product and the feature of a defective product, and Equipped with means for quantitatively determining results by matrix calculation by weighting the data of the measured motor collected from the acceleration type vibration sensor that is in contact with the place where vibration is generated, by the acquisition period, the number of samplings, and each frequency band Sensory test automatic measurement system.