JPH012752A - Spring manufacturing device and method - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a spring manufacturing apparatus and method for manufacturing a spring with a predetermined free length.

[Prior Art] Conventionally, when manufacturing a spring using this type of apparatus, it is common to set parameters related to spring manufacturing.

However, even if these various parameter data related to spring manufacturing are set, springs with the same free length are not always manufactured, and a certain degree of variation usually occurs.

The reasons for this are mainly due to changes in the material of the wire, wire characteristics such as non-uniformity of its cross-sectional shape (diameter, etc.), and changes in the environment such as temperature during manufacturing. In particular, the wire properties not only change between different wire lots, but also vary slightly within a single lot.

If the wire properties and temperature change as described above, the elastic force of the wire itself will change.

For example, even if a spring is manufactured by forcibly winding a wire around a winding shaft and rotating the winding shaft eight times in the axial direction, the wire itself will slightly return to its original state due to its elastic force. It is not possible to manufacture a spring with the same pitch and number of turns.Considering that the elastic force of the wire rod changes from time to time when manufacturing a spring, it is difficult to manufacture a spring with a constant free length.・ Therefore, if a spring that is within the allowable range for the target free length is considered to be a good product, what is the ratio of the number of good springs to the total number of springs manufactured (defective product rate)? The problem is how to increase the yield, or in other words, how to improve the yield, but in the past, there were many aspects that relied on the intuition and skill of the workers, and no solid technology had been established. Ta.

[Problem A to be Solved by the Invention] The present invention has been made in view of the above-mentioned prior art,
It is an object of the present invention to provide a spring manufacturing apparatus and method that can manufacture springs with the best yield rate when manufacturing springs with a desired free length.

Another object of the present invention is to provide a device for manufacturing a spring by forcibly bringing a wire into contact with a coiling point, which can manufacture a large amount of springs with a desired free length with simple operations.

[Means for Solving the Problem] In order to achieve this problem, the present invention includes the configuration shown below.

That is, a setting means for setting a control amount related to the target spring free length, an adjustment means for adjusting the control amount set by the setting means, and a plurality of springs for each of the various control amounts adjusted by the adjustment means. a sampling means for manufacturing a spring, a detection means for detecting the free length of each spring manufactured by the sampling means, and an optimum control method based on the free length of each spring for each control amount detected by the detection means. and analysis means for analyzing the quantity.

Further, another present invention that achieves the above-mentioned object includes the following configuration.

In other words, a coiling point that is located in the direction in which the wire is sent out and forcibly bends the wire in a predetermined direction by contacting the wire, and a coiling point that bends the wire continuously by the coiling point. extrusion means for extruding to create a coil pitch while reciprocating in a direction substantially perpendicular to the surface where the wire is being pushed; a cutting means for cutting the wire in synchronization with one reciprocation of the extrusion movement by the extrusion means; a setting means for setting the length; a detection means for detecting the amount of difference between the free length of the manufactured spring and the target free length; and a plurality of functions for converting the detected amount of the difference into a feedback amount; adjustment means for adjusting the extrusion amount of the extrusion means in accordance with the feedback amount determined by the function; a plurality of sampling means for sampling and manufacturing a predetermined number of springs for each different function; and springs manufactured based on each function. and analysis means for analyzing an optimal function using a distribution based on the free length of .

Further, a spring manufacturing method for achieving the object of the present invention is as follows:
It includes the steps shown below.

In other words, there is a feedback process in which a feedback amount is determined by a predetermined function based on the amount of difference between the free length of the spring manufactured immediately before and the target free length, and the determined amount is fed back to the spring manufacturing process. a sampling step in which a predetermined number of springs are manufactured as samples at each sampling step; a function changing step in which a function for calculating the feedback amount is changed each time a predetermined number of springs are manufactured in the sampling step; The method includes an analysis step of analyzing an optimal function based on a distribution based on the free length of each spring, and a spring manufacturing step of manufacturing a spring based on the optimal function obtained in the analysis step.

[Function] In the configuration of the present invention, each time the adjustment means adjusts the control amount related to spring manufacturing, the sampling means manufactures a plurality of springs, and the analysis means analyzes the optimum control amount. This is an attempt to analyze the optimum control amount and manufacture springs with the best yield rate.

In addition, in a device that manufactures springs by forcibly bringing a wire into contact with a coiling point, multiple functions that define the amount of feedback are prepared, and multiple springs are sampled and manufactured for each function using sampling means. go. Then, the analysis means analyzes the optimal function based on the distribution for the target free length obtained by each sampling means. With this, a large amount of springs with a desired free length can be manufactured with simple operations.

In addition, in the process of the spring manufacturing method of the present invention, a plurality of springs are sampled and manufactured according to the amount of feedback determined by each function, and the optimal function is determined from the distribution of the target free length obtained in each sampling process. This is to analyze.

This is intended to produce springs with the best yield rate.

[Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a schematic diagram of the configuration of a spring manufacturing apparatus according to an embodiment.

In the figure, 1 is a microprocessor (hereinafter referred to as CPU) that controls the entire device. The CPUI is as shown in Figure 7 and as shown in Figure 8.
Processing is performed according to the flowchart shown in the figure, and a program related to this flowchart is stored in the ROM 1a. Further, the RAM 1b is used as a work area for the CPU. Further, 2 is a keyboard for setting parameters related to spring manufacturing (for example, the allowable range of free length, etc.), and 3 is a display section for setting parameters or displaying various graphs based on the spring free length measured during spring manufacturing. It is. The graphs and the like displayed on the display section 3 can also be printed using the printer 4. Reference numeral 5 denotes an oblong device, which detects the distance between the tip of a spring manufactured by a spring manufacturing mechanism section 6 (details will be described later) and a detecting section 5a of the oblong device 5. Specifically, the oblong device 5 employs a capacitance detection method, and can be realized by the circuit shown in FIG.

That is, since the charge capacity changes depending on the distance between the tip of the spring and the detection section 5a, this is made to correspond to the variable capacitor 55. And 0UTA,
By detecting the potential of 0UTa, the capacitance of the variable capacitor 55 can be calculated, and as a result, the distance between the detection section 5a and the tip of the spring can be detected. In addition, capacitor 51~
The charge capacity of 52 and the resistance value of resistors 53 and 54 are known, and the AC voltage generator 56 generates E-sin(Llt(
A voltage of 0≦ω×π) is generated. Therefore, if the length analyzer 5 is fixed in advance, the variation ΔL with respect to the desired free length
It becomes possible to detect the amount of Note that the circuit shown in FIG. 3 is an example, and the present invention is not limited to this.

Further, the CPUI determines from the free length of the manufactured spring whether the spring is a good product within an allowable range, or whether it is shorter or longer than the allowable range. The sorter 7 receives a solenoid drive signal corresponding to this judgment result from the CPU 5, and sorts out springs that are within the permissible range and springs that are not.

The specific structure of the sorter 7 is shown in FIG. Shutter 73.
74 is rotated by the drive of solenoids 71 and 72,
When the levels of the solenoid drive signals output from the CPU 1 are both "0", each shutter is at the position shown by the solid line due to the action of a spring (not shown).

Now, the spring whose free length has been detected by the length detector 5 is cut by the cutter 27 and falls down the common passage 70. At this time, the CPUI uses the solenoid 71.7 based on free length detection.
2 drive signals are output. For example, when it is determined that the free length of the manufactured spring is shorter than the allowable range, the CP
LI 1 outputs a signal that drives only the solenoid 71. Then, the shutter 73 rotates to the position indicated by the broken line 73° in the figure, and the spring that has fallen through the common passage 70 passes through the branch passage 76.

Next, the configuration of the spring manufacturing mechanism section 6 and its operating principle will be explained with reference to FIGS. 1 and 2 (A) and (B).

A first gear 26a and a first feed roller 20a are coaxially supported on the drive shaft of the motor 25. Further, a second gear 26b meshes with the first gear 26a. A second feed roller 20b is fixed coaxially to the second gear 26b. Here, the first and second feed rollers 20a and 20b are configured to sandwich the wire 100 between them and feed the wire 100 toward the point 22 according to their rotation. There is. That is, in FIG. 2(A), the first and second feed rollers 20a and 20b are rotated in the direction shown by the clockwise rotational drive of the motor 25, so that the wire 100 is moved through the guide 21. It is sent out in the direction of point 22.

In addition, a guide groove is provided on the contact surface of the point 22 mentioned above.
is formed. This groove is formed by the wire 10 that abuts here.
0 is tilted so that it is forcibly bent downward as shown.

On the other hand, in addition to the motor 25 described above, a motor 32 is provided to rotate the drive shaft once each time one spring is manufactured to form a pitch. A cam 33 is attached to the drive shaft of this motor 32. A driven member 30 is provided in contact with the cam surface of the cam 33. The driven member 30 reciprocates once in response to one rotation of the cam 33 along a direction transverse to the feeding direction of the wire rod 100 while its axial rotation is restricted by the guide 31 .

An extrusion rod 29 is screwed onto the driven member 30 so as to be movable forward and backward along the direction in which the extrusion rod 29 extends. A pitch tool 23 is attached to the tip of this rod 29 so as to move forward and backward without rotating by a guide 28. Further, FIG. 1 shows a state in which the short diameter portion of the cam 33 is in contact with the driven member 3o, and the pitch tool 2
3 is located at a position that does not substantially form the pitch of the spring. Then, as the cam 33 rotates and the contact position of the driven member 30 changes from the short diameter portion to the long diameter portion, the pitch tool 23 gradually crosses the running path of the wire 1000 and is wound by the groove at the point 22. Push the wire rod part away,
This will form the pitch described above. This state is the second
Figure (A).

Shown in (B).

Note that the wire rod 100 immediately after being bent at the point 22
is cut by a cutter 27 driven in synchronization with one revolution of the motor 32.

By the way, the free length of the spring can be estimated based on the rotational speed of the motor 32 relative to the motor 25, the pitch of the large spring, and the number of turns of the wire 100. However, it is difficult to manufacture springs with exactly the same free length. Because the second
As shown in Figure (B), even when the pitch tool 23 is made to protrude by the protrusion amount LP, the spring pitch P is not necessarily 2LP because the elastic force of the wire changes from time to time.
This is because it fluctuates. Therefore, it is necessary to finely adjust the protrusion amount LP of the pitch tool 23 shown in FIG. 2(B). In order to finely adjust the protrusion amounts t and p, in the embodiment, the rod 29 is rotated, and the driven member 3 of the rod 29 is rotated.
By changing the insertion amount to 0, the driven member 3
The length from the point of contact between the pitch tool 23 and the cam 33 to the tip of the pitch tool 23 is finely adjusted.

In order to realize this, in the embodiment, the worm wheel 36 and its locking member 34, and further the worm wheel 36
A motor 9 and the like are provided to drive the rotation of the motor. The relationship between these components will be explained below.

A worm wheel 36, which is slidably passed through the rod 29 and rotates 155 times during the rod 29, is restricted from moving in the axial direction by a locking member 34. Further, a worm screw 37 that is pivotally supported by the drive shaft of the stepping motor 9 is screwed into the worm wheel 36 . Therefore, by rotating the drive shaft of the stepping motor 9 in a desired direction and as much as necessary, it is possible to finely adjust the protrusion amount LP of the pitch tool 23 described above. This stepping motor 9 is driven by the driver 8, but the direction and amount of rotation of the worm screw 37 are controlled by cput.

Therefore, the problem is how to determine the control amount C that defines the protrusion amount Lp of the pitch tool 23.

That is, when a spring that is longer than the target free length by Δ is manufactured, the amount of feedback (=C×ΔL) is calculated in order to reduce the protrusion fU L P of the pitch tool. Then, the stepping motor 9 is driven according to the value to finely adjust the protrusion amounts t and p of the pitch tool.

For example, if control i (feedback ratio) C is set to 0.01 and a spring with a target free length of +0.05 mm is manufactured, the amount of feedback is 5.0×1O−4
becomes. Then, the drive shaft of the stepping motor 9 is rotated by an amount corresponding to this value, thereby shortening the length from the tip of the pitch tool 23 to the end of the driven member 30. In other words, the pitch tool protrusion TfE L p
Make smaller.

Note that, on the contrary, when ΔL is negative, a corresponding feedback amount is calculated to increase the amount of protrusion of the pitch tool 23.

However, as explained above, since the elastic force of the wire 100 changes from time to time, it is impossible to determine the control RC that satisfies all of these factors.

Therefore, in this embodiment, statistics are collected and analyzed in order to determine the optimum control amount C before manufacturing a spring with the desired free length.

The specific processing contents will be explained below.

First, N springs are manufactured using a function of the control amount C0 as an initial value (hereinafter referred to as sampling manufacturing). Differences with respect to the target free length detected during this sampling manufacturing are sequentially stored in the RAM 1b. Incidentally, in order not to waste the good quality springs manufactured in this sampling manufacturing process, the sorter 7 is driven during this period in accordance with the detected spring free length.

Now, during one sampling manufacturing process or after it is completed, the non-defective product rate G based on the number n of springs within the allowable range, the average value 2 of the difference with respect to the target free length, and its standard deviation value σ are calculated. Note that the average free length τ may be used instead of the average value ΔL.

Here, the multivalue is calculated using the following formula.

G = n / Na = (1/N-Bo1 (ΔL1-2)2F'' from the second time onwards (
The control amount CI in the i-th sampling production is equal to the control amount in the immediately preceding sampling production by ΔC (for example,
0.01) (・C0+ΔCX (i-1)
), and calculate the three values mentioned above each time. After performing the sampling manufacturing process a preset m number of times in this way, it is determined which control amount the sampling manufacturing process will yield the best result.

As for the criteria for determining the optimal control amount, in the embodiment, each element is weighted as shown below.

Good product rate〉Average value〉Standard deviation In other words, when the maximum good product rate can be obtained at the first sampling manufacturing time during m sampling production, the value of C0◆ΔCX(i-1) is determined as the optimal control amount. do.

In addition, if there are two or more candidates for the maximum non-defective product rate, the candidate is determined based on the "average value" which is the second determining factor. Furthermore,
If it is not possible to narrow down the selection to one based on this average value, a judgment is made based on the third judgment factor "standard deviation".

In the example, the number of sampling production m and the number N of springs produced during each sampling production were not specified, but since these values have no statistical significance if they are small, a certain number of springs should be used. is necessary. That is, it is desirable to set m to several thousand times and N to about several hundred times. At this time, it is also important to set the value of the initial control amount C0 and the additional value ΔC in each sampling manufacturing process, but when manufacturing a spring with a relatively long free length, set m to a thick value and set the value of ΔC to a small value. It is desirable to do so. because,
This is because although the feedback amount is determined by the control amount, the variation becomes larger compared to when manufacturing a spring with a short free length, and detailed analysis is required.

Now, the reason why the priority order is set for each element in this way will be explained with reference to FIGS. 5 and 6.

The following explanation is a method of determining the optimal control mart based on the distribution of the difference with respect to the target free length, but the same applies to the distribution of the free length of manufactured springs.

Now, the free length is 50.00mm, the tolerance range is ±O,
Suppose that n samples of springs were manufactured for the purpose of manufacturing Oamm springs. At this time 50.00mm
Figures 5(A) and 5(B) are graphs in which the vertical axis represents the difference in frequency and the horizontal axis represents the frequency. It is assumed that the non-defective product rates in these figures are the same. Further, as a matter of course, the control amount in each graph is different.

Now, the average difference with respect to the desired spring natural length in Figure 5 (A) is about 0.0013 mm, while in Figure 5 (B
) is -0,0145 mm, which clearly shows that the control amount for sampling production in FIG. 5(A) is superior. Therefore, it is easy to understand the importance of anticipating cases in which the non-defective product rates will be the same and using that average value as the second criterion. In other words, it is also a guideline for determining whether it is possible to manufacture a spring with higher precision by making the tolerance range smaller (eg, ±0.04n+m, etc.).

Furthermore, assuming that the non-defective rate and the average value are the same, we use the third judgment criterion, the target! $! It was decided to make a judgment using a deviation of 0 (or an error of 02).

Figures 6 (A) and (B) show the same non-defective product rate and
This shows a case where the error with respect to the target free length is both 0.00 mm. In this case, it goes without saying that it is not good if the error is 0.00 mm frequently (the standard deviation is small), so sampling manufacturing using the controlled variable shown in Fig. 6 (B) (standard deviation 0) is not a good thing. It can be easily understood that the standard deviation 0 = approximately 0.026') is superior to that in FIG. 6(A) (standard deviation 0 = approximately 0.039). In particular, in the case of FIG. 6(B), since the standard deviation is small, it is conceivable to further reduce the allowable range for the spring free length.

In addition, by displaying the above-mentioned graph and the time-series changes of the three values serving as judgment criteria on the display unit 3, it becomes easy for the operator to understand the current situation.

The processing procedure in the embodiment based on the above configuration or principle can be summarized as shown in the flowcharts shown in FIGS. 7(A) and 7(B).

First, in step S1, the number m of sampling production is set using the keyboard 2. Thereafter, in step S2, the number N of springs to be manufactured in each sampling process is set, in step S3 the allowable range, in step S4 the initial control amount COs, and in step S5 the control amount increment value ΔC. In the next step S6, "1" is assigned to the variable i as an initial value.

Incidentally, it is determined whether sampling manufacturing has been completed or not based on the contents of this variable l.

In this way, the process moves to step S7, where a sampling manufacturing process is performed, the details of which will be described later.

When one sampling manufacturing process is completed, the variable i is incremented by one in step S8, and the variable i and the number of sampling manufacturing times m are compared in the next step S9. If it is determined to be "15 m", the process returns to step S7 to execute the next sampling manufacturing process, and steps 57 to S9 are repeated until it is determined that "i>m".

Now, if it is determined that "i>m", the process moves to step 510, and the optimum control amount is determined according to the determination criteria described above. Based on the optimal control amount thus obtained, spring manufacturing is executed in step Sll.

This process will be executed until a preset number of non-defective springs is reached or until the device is stopped.

Next, the contents of the sampling manufacturing process in step S7 are explained in the eighth step.
This will be explained according to FIGS. (A) and (B).

In step 3701, the control amount C in sampling manufacturing is determined based on the variable i representing the order of sampling manufacturing.
Calculate using the following formula.

Controlled amount C==CO◆(i-1)xΔC Therefore, the controlled amount at the time of the first sampling production becomes the initially set value C0.

In the next step 5702, "1" is assigned to variable j representing the number of springs manufactured during sampling manufacturing, and variable A representing the number of non-defective springs and variable B representing the sum of variations are each set to °°0. After completing the above initial settings, proceed to step 5703 to actually manufacture one spring.In the next step 5704,
Variation amount Δ with respect to the target free length detected by the length analyzer 5
Detect L and store its value in variable D(j). In step 5705, it is determined whether the variation amount D(j) is within the allowable range. If it is determined that it is within the allowable range, "1" is added to variable A in step 5706, and step S Moving on to F08. Also, dispersion: ftD(
j) is outside the allowable range, the process moves to step 5707, where one of the solenoids 71 and 72 of the reducer 7 is driven for a predetermined period of time based on the sign of the amount of variation, and the process moves to step 5708. Move.

In step 3708, the variation amount D(j) is added to the variable B, and in the next step 5709, the value of D(j) and the graph described above are displayed.

In step 5710, a feedback amount F is calculated. The calculation function for this amount of feedback has already been explained, but if I quote it again here, it becomes the following formula.

F=CXD (j) Then, in step 5711, the stepping motor 9 is driven based on the magnitude and sign of the obtained feedback amount F.

In the next step 5712, the variable j is incremented by one, and in step 5713, the variable j and the set value N are compared.

At this time, if it is determined that "j≦N", it means that N springs have not been manufactured yet, so step 57
Return to 03.

When N springs are manufactured in this way, step 571
If it is determined in step 3 that "' j >N", the processing from step 5714 onwards will be executed. Therefore, at this time, variable A stores the number of springs within the allowable range.

Further, the variable B stores the total value of the variations of N springs, and the variables D(1) to D(N) store the variations 2 of each spring.

Then, based on these values, steps 8714-5
716, the non-defective rate G (
i), average values X(i) and σ(i) are calculated, and these values obtained in the next step 5717 are stored in the RAM 1b. By performing the above processing for each sampling production, the non-defective product rate, average, and standard deviation specific to each sampling production can be obtained. Therefore, in step SIO explained earlier, the obtained variables G(i),
i), the optimal control amount is determined according to σ(i).

As described above, according to the present embodiment, by detecting the best condition for spring manufacturing before manufacturing the spring, it is possible to manufacture the spring in a condition with the best quality product rate.

Furthermore, all of this processing is done automatically, so
Even a person with little experience in spring manufacturing can reliably manufacture a spring with a good free length.

Note that it is desirable to carry out similar pretreatment even when installing a new wire after one winding of the wire has been used, or when manufacturing springs with different lengths in the middle. This is because there are subtle changes in the material, diameter, etc. of wire rods from different lots.

In addition, in the example, when the non-defective rate is within the set tolerance range, it is calculated by incrementing the counter, but the standard deviation value is calculated, and from that value, the non-defective rate is also calculated when the tolerance range is set. Since this is mathematically possible, it is not necessary to count each item one by one. That is, the optimum control amount can be determined only from the distribution of free lengths during the production of each sample.

Further, in the embodiment, the feedback amount is calculated each time using a function, but it may be stored as a table in the ROM 1a and derived.

In addition, when the number of sampling production is plotted on the horizontal axis and the non-defective product rate is plotted on the vertical axis, if a graph with the relationship shown in Figure 9 is obtained, for example, the subsequent sampling production is stopped. It is also possible to prevent unnecessary waste of wire rods. However, since it is impossible to judge whether the non-defective product rate has reached the maximum unless the non-defective product rate of the next sampling (point B) is measured, in reality, the number of samplings when the maximum non-defective product rate is detected + the number of samplings once. Make it necessary.

Furthermore, in Figure 9, if the maximum non-defective rate is obtained at 0 points, strictly speaking, it is determined that there is a setting (environment) that provides the maximum non-defective rate between points A and B that sandwich the 0 point. be done. Therefore,
Returning to the sampling state (point A in the figure) before the sampling where the maximum non-defective rate was detected, this time △C'
(However, △C.

By setting ΔC) as an increment value and executing the flowcharts shown in FIGS. 7 and 8 up to point B, it is possible to detect a spring manufacturing environment with an even better rate of non-defective products.

Furthermore, in this embodiment, only controlling the pitch tool 23 has been described, but the position of the point 22, for example, also has a considerable effect on the free length of the spring. Therefore, the position of this point 22 may also be finely adjusted and analyzed in the same manner.

Furthermore, in the embodiment, the spring 25 and the motor 32 are distinguished, but the spring feeding and spring pitch may be generated by driving one spring.

[Effects of the Invention] As described above, according to the spring manufacturing apparatus and method of the present invention, it is possible to manufacture springs with the best yield rate.

Further, in an apparatus for manufacturing a spring by forcibly bringing a wire into contact with a coiling point, it is possible to manufacture a large amount of springs with a good free length with simple operations.

[Brief explanation of drawings]

Figure 1 is a schematic diagram of the configuration of the spring manufacturing system in the example, Figures 2 (A) and (B) are diagrams for explaining the principle of spring manufacturing in the unexamined example, and Figure 3 is the inspection length in the example. Figure 4 is a diagram showing the external appearance of the sorter in the embodiment, Figures 5 (A), (B), and Figures 6 (A) and (B) are 7(A) and 7(B) are flowcharts showing an outline of the processing of the CPU in the embodiment; FIGS. 8(A) and (B) are flowcharts showing the processing contents of the CPU during sampling manufacturing in this embodiment, and FIG. 9 is a diagram showing an example of the relationship between the non-defective product rate and sampling manufacturing in this embodiment. In the figure, 1... CPU, 2... Keyboard, 3...
Display section, 4... Printer section, 5... Length analyzer, 5a.
...Detection section, 6...Spring manufacturing mechanism section, 7...Selector, 8...Driver, 9...Stepping, 20a,
20b...Feed roller, 21.28.31...
Guide, 22...Point, 23...Pitch tool, 25.32...Motor, 26a, 26b...Gear, 27...Cutter, 29...Rod, 30...
Driven member, 33... cam, 34... locking member,
36... Worm wheel, 37... Worm screw. :・4・12↓ Figure 3 Figure 4 Figure M5 (A) Figure 5 (8) Figure 6 (A) Figure 7 (A) Figure 7 (B)

Claims (3)

[Claims] (1) A setting means for setting a control amount related to the free length of the target spring; an adjustment means for adjusting the control amount set by the setting means; and a plurality of control means for each of the various control amounts adjusted by the adjustment means. a sampling means for manufacturing a spring; a detection means for detecting the free length of each spring manufactured by the sampling means; 1. A spring manufacturing device comprising: analysis means for analyzing a controlled amount. (2) A feeding means for feeding out the wire; a coiling point located in the direction of feeding and forcibly bending the wire in a predetermined direction by abutting the wire; and a coiling point that is continuously bent by the point. , an extrusion means for creating a coil pitch while reciprocating in a direction substantially perpendicular to the surface being bent; and a cutting means for cutting the wire in synchronization with one reciprocation of the extrusion movement by the extrusion means. means, a setting means for setting a desired free length, a detecting means for detecting the amount of difference between the free length of the manufactured spring and the desired free length, and converting the detected amount of the difference into a feedback amount. a function; an adjusting means for adjusting the extrusion amount of the extrusion means according to the feedback amount determined by the function; a plurality of sampling means for sampling and manufacturing a predetermined number of springs for each different function; and analysis means for analyzing an optimal function based on a distribution based on the free length of a spring manufactured based on the above, and the spring is thereafter manufactured based on the obtained optimal function. Spring manufacturing equipment. (3) A feedback process in which a feedback amount is determined by a predetermined function based on the amount of difference between the free length of the spring manufactured immediately before and the target free length, and the determined amount is fed back to the spring manufacturing; a sampling step of manufacturing a predetermined number of springs as samples for each function; a function changing step of changing the function for calculating the feedback amount each time a predetermined number of springs are manufactured in the sampling step; A spring manufacturing method comprising: an analysis step of analyzing an optimal function using a distribution based on the free length of a spring for each function; and a spring manufacturing step of manufacturing a spring based on the optimal function obtained in the analysis step.