JPH01192675A - Poor bobbin detecting method for winder - Google Patents

Abstract

PURPOSE:To execute accurate discrimination of whether a bobbin is poor, by a method wherein a yarn is pulled off from an exchanged new bobbin, and in a section where a first winding speed is specified, a signal from a yarn unevenness detector is sampled a plurality of times. CONSTITUTION:A yarn is pulled out from an exchanged new bobbin and is wound after passage of it through a yarn unevenness detector 8. In a section where a first winding speed is specified, a signal from the detector 8 is sampled a plurality of times, and an accurate digital value from which a fluctuation in a winding speed is removed is obtained. In sampling only in a section where a first winding speed is specified, if a first pulled off part has a fluctuation in the size of some yarn, a following part has the same fluctuation in the size of a yarn, and therefore by checking the fluctuation in size of a first yarn, a poor bobbin can be discriminated without needing checking a bobbin for an overall length. Based on the sampling data, Fourier computation is effected, its result is compared with a comparator 27, and it is decided whether unevenness in a period exceeding a reference value occurs.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Field of Application J] The present invention relates to a defective bobbin detection method useful for automatic winders and the like.

[Prior Art] An automatic winder used for winding bobbins after spinning consists of a single winder with multiple winding units arranged side by side, and the bobbin supplied to each unit is pulled out at high speed. The yarn is wound onto a package rotated by a drive drum, and when the wooden bobbin runs out of yarn, a new bobbin is supplied, the yarn is tied, and winding is performed again. This automatic winder is equipped with a slub catcher for detecting yarn defects, and the slub catcher detects defects in thick yarn, thin yarn, slabs, etc. contained in the system while winding the bobbin. If a defective part with a value exceeding a certain set value is detected, the thread is cut.

However, defects such as fluctuations in yarn thickness that occur in the spinning machine itself are not detected during winding. Such fluctuations in yarn thickness can include eccentricity or deformation of the rollers in the spinning machine, defects in the drive system, etc. There are periodic fluctuations in thread thickness caused by thread thickness, and non-periodic unevenness caused by abrasion of the apron surface in spinning machines. When woven, unevenness in thread thickness can be a serious drawback depending on the type of fabric, such as defects such as moire patterns, which can significantly reduce the commercial value of the fabric.

Therefore, in general, some of the wound bobbins are taken out, and the threads wound on the bobbins are evaluated for thread unevenness by testing equipment such as an Ustamura tester and a spectrogram installed at separate locations. Based on this information, the yarn can be tested and the defects of the spinning machine that spun the yarn can be estimated.

However, with this detection method, manual sampling is labor-intensive, inspection using test equipment such as the Ustamura tester and spectrograph takes a long time, and the accuracy is low, resulting in a high level of final reliability. Since the winder is in continuous operation until the analysis results are obtained, a large amount of yarn with defects and unevenness is wound onto the package during that time.

This problem is regarded as a problem on the spinning machine side, and methods disclosed in Japanese Patent Publication No. 52219/1982 and Japanese Patent Publication No. 17929/1988 are available as methods for solving the variation in yarn thickness. That is, by digitizing the analog signal emitted from the yarn unevenness detector directly attached to the spinning machine, and having the digitalized signal be processed in real time by a computer, yarn unevenness signals can be analyzed with high precision in an extremely short period of time. The thread unevenness signals of multiple spindles are sequentially input into one analysis device via a multiplexer.
A system is disclosed in which yarn unevenness is evaluated one spindle at a time.

[Problems to be Solved by the Invention] However, winders have special problems that are different from spinning machines. That is, in a spinning machine, the spinning speed is constant, but in a winder, the rotation speed of the drive drum is constantly changed to prevent so-called ribbon winding, and the time axis varies. For this reason, the above-mentioned analysis device has not yet been installed.

Therefore, in current winders, among the yarn defects,
There is a problem in that slight defects near the set value, such as the above-mentioned fluctuation in thread thickness, are overlooked, and when the set value is made severe, the number of thread breakages increases.

The present invention aims to solve the above-mentioned problems, and makes it possible to detect slight defects such as variations in yarn thickness in a winder whose winding speed is variable, and to determine whether a bobbin is a defective bobbin or not. An object of the present invention is to provide a method for detecting a defective bobbin that can accurately identify the defective bobbin.

[Means for Solving the Problems] The method for detecting a defective bobbin of the present invention is to draw yarn from a new bobbin that has been replaced, pass it through a yarn unevenness detector, and wind it in a winder whose winding speed is variable. During the initial constant winding speed section, the signal obtained from the yarn unevenness detector is sampled multiple times, each converted to a digital value and stored, and a computer performs Fourier calculations based on these sampling data. The comparison means compares the result with a reference value, and determines whether there is periodic irregularity exceeding the reference value.

[Operation] The thread is drawn out from the new bobbin that has been replaced, passed through the thread unevenness detector, and wound up. Although the winding speed of this winder is not always constant to prevent ribbon winding, there are several sections where the winding speed is constant. During the initial period where the winding speed is constant, the signal obtained from the yarn unevenness detector is sampled multiple times to obtain a correct digital value from which fluctuations in the winding speed have been removed. The reason for sampling only in the initial constant winding speed section is that if the first drawn-out section has a variation in the thickness of the thread, the subsequent parts will also have the same variation in thread thickness. By checking the yarn length fluctuation, it is possible to determine whether or not the bobbin has a defective bobbin, without having to check the total thread length of the bobbin. By comparing the result with a reference value by the comparison means, it is determined whether or not there is a cycle irregularity exceeding the reference value.

The yarn unevenness detector may be an existing detector such as a slab catcher already installed in the winder, or it may be a separate detector.

[Example] Below, I5! ! The present invention will be explained based on the illustrated embodiment.

FIG. 2 schematically shows an automatic winder to which the present invention is applied, in which a support shaft 2 and a suction pipe 3 are installed between each side frame 1.

Reference numeral 4 denotes a winding unit, which is rotatably supported on the support shaft 2, and is also placed on the suction pipe 3 and fixed to an appropriate amount while the automatic winder is in operation. Incidentally, the suction pipe 3 is connected to a blower (not shown), and a suction air current is constantly applied to the suction pipe 3.

Rewinding the yarn from the bobbin 15 to the package P in the winding unit 4
5, the yarn Y passes through a guide 6, and is given an appropriate tension by a tensor 7, and a yarn unevenness detector (slab catcher) that also serves as yarn cutting and yarn running and loading when yarn unevenness exceeding a predetermined standard such as a slab is detected. 8 and is rolled up onto a package P rotated by a grooved drum 9.

The rotational speed of the grooved drum 9 can be controlled steplessly by an inverter (frequency variable device) 22 shown in FIG. 22.

The rotation control of the grooved drum 9 during winding is performed as a combination of a constant speed control section and a variable speed control section in which the speed is varied so that ribbon winding does not occur on the package P.

The yarn unevenness detector 8 is an electrostatic type or a photoelectric type consisting of a light emitting diode and a phototransistor, and is highly sensitive and highly responsive. When the displacement is detected, the cutter 8b installed near the yarn unevenness detector 8 operates to cut the running yarn Y.
Winding is stopped.

When winding is stopped due to slab detection, the first yarn guide suction arm 10 connected to the spool pipe 3 pulls out the yarn YP on the package P side, and the second yarn guide also connected to the spool pipe 3 pulls out the yarn YP on the package P side. The spool arm 11 pulls out the thread YB from the bobbin 15 and guides it to the thread ai* position 12 located at a position away from the normal thread travel path.
The yarn splicing device 12 connects the yarn YP and the yarn Y by a spooling action.
The ends of the untwisted yarns YP and YB are untwisted, and the ends of the untwisted yarns YP and YB are polymerized by compressed fluid (in this case, compressed air) supplied to the yarn splicing box 16 from the pipe 13 in a separate route via the conduit 14. After the yarn splicing is performed and the yarn is spliced by the yarn splicing device 12, rewinding of the yarn is continued.

FIG. 1 shows the configuration of a control device including a microcomputer 20 provided in the front winding unit 4.

The electric signal S from the yarn unevenness detector 8 is used for cutting the yarn as described above, and is also input to the yarn unevenness signal processing means 23 of the control device shown in FIG. A period in which the rotational speed of the drum 9 is constant, particularly in the first constant speed period after bobbin replacement, is sampled multiple times and digitally analyzed.

That is, in FIG. 1, the electric signal S from the yarn unevenness detector 8 is amplified by the amplifier 24 to a voltage level most suitable for A/D conversion, and then converted into an analog signal by the A/D converter 25. converted into a digital signal.

The A/D converter 25 is an accurate oscillator that creates a data sampling time determined according to the frequency bandwidth to be analyzed in the first constant rotational speed section of the grooved drum 9 after replacing the bobbin. , samples the input signal multiple times and converts each into a digital signal. For this sampling, the microcomputer 20 performs A/D conversion to determine that the rotational speed of the grooved drum 9 is constant during a predetermined interval synchronized with the speed command given to the inverter 22, that is, the first interval after replacing the bobbin. to the device 25.

Each sampling data obtained by the A/D converter 25 is
It is stored in the memory RAM of the microcomputer 20.

Each of the stored sampling data for a predetermined number of times is first weighted by Landau, and then sent to a Fourier transformer 26 where it is subjected to a fast Fourier operation (FFT operation).
The calculated results are vector-combined into a power spectrum and output as a power spectrum of each frequency component.

The signal output as a power spectrum of each frequency component is processed by an output processing circuit to be suitable for analysis, such as by cutting off signals related to a predetermined frequency, for example, a frequency of 50 Hz or higher, and the processed signal is sent to a microcomputer 20. The data is stored in the memory RAM of the computer, and in some cases, outputs from the Fourier transformer 26, is input to a D/A converter, is converted into an analog value, and is displayed as a graph on a display device (not shown).

The data resulting from the fast Fourier calculation thus obtained is read out from the memory RAM of the microcomputer 20 and compared with the set value by the comparator 27 to evaluate the yarn. In the case of a defective bobbin with cycle irregularities exceeding a reference value, winding is stopped by cutting the thread, and the bobbin processing device 30 replaces the bobbin with a new bobbin.

The reference value here can be set much more severely than the setting value in the case of slab detection described above, since the sampling data is correct data that is not affected by fluctuations in the yarn winding speed. .

The microcomputer 20 basically consists of a CPU which is a processing unit, a ROM and a RAM as memories, and an RO
The programs shown in FIGS. 3(a) and 3(b) are written in M, and the CPU reads data signals etc. from the thread unevenness signal processing means 23, or transfers data signals between the thread unevenness signal processing means 23 and the RAM according to the program. It sends and receives data, performs arithmetic processing, and outputs the processed data as necessary.

Here, FIG. 3(a) shows a main routine for defective bobbin detection processing, and FIG. 3(b) shows a timer interrupt routine for sampling which is interrupted at fixed time intervals.

First, in the main routine shown in FIG. 13(a), the bobbin processing device 30 is instructed to replace the bobbin and a new bobbin is set (step ■).Next, the sampling start flag is set (step ■) for a predetermined number of times ( Wait for the sampling data (n times) to be collected (step ■).

On the other hand, in the timer interrupt routine of FIG. 3(b),
Wait until the sampling start flag is set (step ■)
, When the sampling start flag is set, check whether the winding speed is within a constant range (step ■
) If the 0 winding speed is constant, perform data sampling (step o) and return (step 0)
, by repeating the same steps (step ■～■),
When sampling has been performed a predetermined number of times (n times) within a constant winding speed section (step o), a sampling end flag is set, a sampling start flag is reset,
Return (step @.

O).

In this way, when the sampling data for a predetermined number of times (n times) is collected, the main routine in FIG. Perform calculation (step ■) 0 Then, based on this FFT calculation result,
Determine whether or not the thread of the bobbin has a periodic unevenness exceeding a reference value (step ■). If the result of 0 determination is that there is a one-period unevenness, instruct the bobbin processing device 30 to replace the bobbin (step ■). Set the correct bobbin and exit the defective bobbin detection process.

In other words, in this defective bobbin detection flow, after the bobbin is replaced by the bobbin processing device 30, the yarn thickness is determined only in the first constant yarn speed section when the yarn is pulled out from the bobbin and begins to be wound. The thread thickness is not checked in subsequent sections of the same bobbin where the thread speed is constant. Since the subsequent sections have the same variation in thread thickness, it is sufficient to check the first drawn-out section.

In the embodiment, when it is determined that the bobbin is defective, it is replaced with a new bobbin, but it is also possible to generate an alarm sound or display an abnormality display.

[Effect of the invention 1 As described above, the method for detecting a defective bobbin of the present invention is as follows:
The thread is pulled out from the replaced new bobbin, passed through the thread unevenness detector, and wound up, and the signal obtained from the thread unevenness detector is sampled multiple times during the initial constant winding speed section. For this reason, correct digital values are obtained by removing fluctuations in the winding speed.2 By performing Fourier calculations on a computer based on these sampling data and comparing the results of the Fourier calculations with the standard values using a comparison means, the standard value can be determined. It is possible to easily determine whether there is periodic irregularity exceeding the value. In other words, the settings for determining whether a bobbin is a defective bobbin can be set severely, and the thread from the defective bobbin, which varies in thread thickness, may be mixed with the correct thread from another bobbin and wound into the package. This inconvenience is avoided.

In addition, since it is immediately determined whether the thread is from a defective bobbin during the first constant winding speed section, the thread from the bobbin will not be wound around the package, and measures such as replacing the defective bobbin can be taken. can be done quickly.

Therefore, in a system in which an automatic winder is directly connected to a spinning machine and winds a wide variety of yarns using one winder, even when trays with various bobbins inserted are transported randomly along a common conveyance path, it is possible to This method is particularly suitable for selecting only defective bobbins.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a defective bobbin detection device according to the method of the present invention, and FIG. 2 is a schematic diagram of a winder to which the present invention is applied.
FIGS. 3(a) and 3(b) are flowcharts showing examples of programs input to the electronic computer. In the figure, 8 is a thread unevenness detector, 8b is a cutter, 10.11
12 is a sakushi arm, 12 is a yarn splicing device, 20 is a microcomputer, 22 is an inverter, 23 is a yarn unevenness signal processing means, 25 is an A/D converter, 26 is a Fourier transformer,
27 is a comparator, and 30 is a bobbin processing device. Patent Applicant: Murata Machinery Co., Ltd. Representative Patent Attorney: Nobuo Tsunatani

Claims (1)

[Claims] 1. In a winder whose winding speed is variable, thread is pulled out from a new bobbin that has been replaced, passed through a thread unevenness detector, and wound.In the first section where the winding speed is constant, the thread is obtained from the thread unevenness detector. A signal is sampled multiple times, each converted into a digital value and stored, a computer performs a Fourier operation based on these sampling data, the result of the Fourier operation is compared with a reference value by a comparing means, and a period exceeding the reference value is determined. A method for detecting a defective bobbin in a winder, the method comprising determining whether there is unevenness.