JPH06137970A - Method for constituting load cell and the load cell - Google Patents

Abstract

PURPOSE:To make it possible to improve a load characteristic in a load cell for detecting load with a strain gage provided on a strain producing part on which tensile strain is caused and another strain producing part on which compression strain is caused. CONSTITUTION:A strain sensor is installed between a fixed side arm 22a and a movable side arm 23a projected from a fixed rigidity body part 22 and a movable rigidity body part 23 in a strain producing body 26 to constitute the load cell. In the case where each pair of notches 36a, 36a, 36b, 36b whose opposed side edge parts are oppositely cut are provided on a rectangular thin metal board 35 for constituting a strain sensor 27, the positions between these pairs of the notches are respectively regarded as strain producing parts 37a, 37b to form gage resistant parts 39a, 39b. And one width Wt between the notches 36a, 36a opposed to and positioned at the fixed side arm 22a is set smaller than the other width Wc between the notches 36b, 36b opposed to and positioned at the movable side arm 23a.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a load cell used for detecting various loads, specifically, a load applied to a strain-generating body by a change in electrical characteristics of a strain gauge attached to the strain-generating body of a predetermined shape. The present invention relates to a load cell adapted to electrically detect.

[0002]

2. Description of the Related Art Generally, in a load cell used for load detection in an electronic scale or the like, two beam parts each having a pair of strain-reducing parts with weakened rigidity are installed in parallel between rigid parts located at both ends. A strain gauge having a hollow quadrangular shape is attached to each strain gauge on the surface of the strain generating section, and these gauges are connected so as to form a Wheatstone bridge circuit, and are loaded on the strain body. The load circuit is configured to detect the load as the output of the bridge circuit. In this type of load cell, as shown in FIG. 11, for example, as shown in FIG. A case where a load cell 7 having a three-layer structure is constructed by constructing a Roberval mechanism with the upper and lower beam portions 4 and 5 and installing a strain sensor 6 between both rigid body portions 2 and 3 of the strain generating body 1. A.

In other words, in the load cell 7, the semi-circular notch is formed on the inner surface side at the connecting portions with the rigid body portions 2 and 3 at both ends of the upper and lower two beam portions 4 and 5 constituting the flexure element 1. 8d are provided respectively, and the strain-generating portions 9a ... 8d whose wall thickness is reduced by these notches 8a.
9d is formed, and the strain sensor 6 is installed between the left and right mounting arms 2a and 3a projecting from the respective rigid body portions 2 and 3 so as to face each other between the upper and lower beam portions 4 and 5. It is like this.

Now, the structure of the strain sensor 6 will be described with reference to FIG. 12. A pair of notches are formed on one end side of the thin rectangular metal substrate 8 so that opposite side edge portions are cut to face each other. 9a, 9a are provided so that the portion between the notches 9a, 9a is one strain-generating portion 10a, and the opposite side edge portion is also cut to face the other end of the metal substrate 8. A pair of notches 9b and 9b having the same size as the inserted notches 9a and 9a are provided, and a portion between the notches 9b and 9b is another strain generating portion 10b. Also,
A strain detecting circuit 12 is formed on the metal substrate 8 with an insulating layer 11 interposed therebetween.
Are formed. The distortion detecting circuit 12 is provided with the strain-generating portion 1 between the pair of notches 9a, 9a; 9b, 9b.
0a and 10b, and the gauge resistance portions 12a and 12b having a predetermined pattern formed on the gauge resistance portions 12a and 12b.
12c connected to a and 12b, and the flat cable 13 is connected to the conductive parts 12c ... 12c on one side of the metal substrate 8 by soldering or the like, Both gauge resistance part 12
The resistance change of a and 12b is taken out as an electric signal through the flat cable 13.

When the load cell 7 is used in an electronic scale, for example, as shown in FIG. 13, one rigid body portion (hereinafter referred to as a fixed rigid body portion) 2 is attached to a base 15 erected on a pedestal 14. Together with the other rigid body part (hereinafter,
A weighing pan 17 is attached to the movable rigid body portion 3 via a bracket 16. Then, as shown in FIG. 17, two dummy resistors 18, 18 are provided for the pair of gauge resistor portions 12a, 12b in the strain sensor 6.
By connecting so that the Wheatstone bridge circuit 19 is formed, the output terminal 19a connected between the gauge resistance portions 12a and 12b and the dummy resistors 18 and 18 are connected.
An output voltage e corresponding to a change in the resistance value of the gauge resistance portions 12a and 12b with respect to a predetermined input voltage E is output between the output terminal 19b and the output terminal 19b.

That is, when the object X to be weighed is loaded on the weighing pan 17, the load is applied to the movable rigid body portion 3 of the flexure element 1 via the bracket 16, so that the movable rigid body portion 3 is fixed. While being displaced downward with respect to the part 2, the movable rigid body side mounting arm (hereinafter, referred to as the movable side arm) 3a is also displaced with respect to the fixed rigid body side mounting arm (hereinafter, referred to as the fixed side arm) 2a. It will be displaced downward.

In this case, the metal substrate 8 constituting the strain sensor 6 is deformed into an S-shape with the strain-flexing portions 10a and 10b as bending points, and the strain on the fixed rigid body side is generated. Tensile stress acts on the portion 10a, and compressive stress acts on the strain generating portion 10b on the movable rigid body side. Therefore,
Gauge resistance portion 1 formed on the strain-flexing portion 10a on the fixed rigid body side
In the Wheatstone bridge circuit 19, 2a receives tensile stress and its resistance value increases, and the gauge resistance portion 12b formed in the strain generating portion 10b on the movable rigid body portion side receives compressive stress and its resistance value decreases. Output terminals 19a, 1
The change in the output voltage e between 9b becomes large, and good distortion sensitivity can be obtained.

[0008]

However, in the load cell having the three-layer structure as described above, the longitudinal dimension between the notches (straining portions) in the strain sensor is determined by the notch of the straining body constituting the Roberval mechanism. It was found that the load characteristics of the load cell did not show linearity because it was shorter than the longitudinal dimension between the strained portions).

That is, as shown in FIG. 15, the actual output voltage corresponds to the ideal characteristic obtained by connecting the maximum output voltage when a predetermined reference load Fs is applied and the output voltage in the unloaded state. It will be output to the positive side like.

The cause is considered to be as follows.

That is, in the load cell 7 schematically shown in FIG. 16, when a load F is applied to the movable rigid body portion 3 which constitutes the flexure element 1, the upper and lower beam portions 4, 5 of the flexure element 1 are applied.
Ideally, as shown in FIG. 17, the strained portions 9a and 9b on the fixed rigid body portion 2 side pivot downwardly, and the movable rigid body portion 3 translates downward by a distance ΔY accordingly. As a result, it is considered that the movable arm 3a, which is integral with the movable rigid body portion 3, also translates downward by the distance ΔY. Therefore, the displacement angle θ of the strain sensor 6 having the bending portion 10b on the movable rigid body side as the bending point should coincide with the displacement angle θ ′ having the bending portion 10a on the fixed rigid body side as the bending point.

However, since the longitudinal dimension L1 between the notches (strain portions) of the strain sensor 6 is shorter than the longitudinal dimension L2 between the notches (strain portions) of the strain body 1 which constitutes the Roberval mechanism, The movable arm 3a has the above distance ΔY.
In order to move only downward, the strain sensor 6 on the movable rigid body portion side of the strain sensor 6 is moved to the fixed rigid body portion 10a.
Therefore, the movable arm 3a and the strain sensor 6 have to be horizontally extended by ΔX more than before the deformation.

Therefore, in actuality, as shown in FIG. 18, a bending load in the direction of the arrow in the figure acts on the movable rigid body portion 3, and the movable rigid body portion 3 is indicated by the two-dot chain line in the figure. As a result, the movable state arm 3a is inclined with respect to the ideal state as shown by the solid line, and accordingly, the movable arm 3a is also inclined from the horizontal state by the angle γ. Therefore, if the displacement angle on the fixed rigid body side of the strain sensor 6 is α and the displacement angle on the movable rigid body side is β, the relational expression of α = β + γ holds, and therefore the displacement angle α is larger than the displacement angle β. Become. Therefore, the strain sensor 6
It is considered that the tensile stress generated in the strain generating section 10a on the fixed rigid body side in the above is larger than the compressive stress generated in the strain generating section 10b on the movable rigid body side, and causes an error in the output voltage as described above. To be

For the purpose of improving the abnormal load characteristics of the load cell 7, for example, a load load test is performed by replacing the gauge resistance portion 12b on the compression side in the Wheatstone bridge circuit 19 of FIG. 14 with an equivalent dummy resistance. As a result, experimental results were obtained in which the output voltage was convex toward the minus side and showed a slightly small nonlinear error as shown by the chain line in FIG. The details of the load application test will be described later. Further, when the same load load test was performed by replacing the gauge resistance part 12a on the tension side with an equivalent dummy resistance, this time the experimental result showing a large non-linear error in which the output voltage is convex on the plus side as shown by the broken line. was gotten. In this case, since the load characteristic of the load cell 7 is considered to be an average of the output of the tension side gauge resistance portion 12a and the output of the compression side gauge resistance portion 12b, the overall output characteristic is the solid line in FIG. It is presumed that it exhibits a non-linear error that is convex on the positive side as shown in.

It should be noted that such a non-linear error is detected in a load cell in which a strain gauge provided in a strain generating portion where a compressive strain is generated and a strain gauge provided in a strain generating portion where a tensile strain is generated detect a load. Widely and generally seen with varying degrees.

The present invention uses a hollow quadrilateral strain element in which a pair of upper and lower beam portions are installed in parallel between a fixed rigid portion and a movable rigid portion located at both ends, and a strain generating portion in which tensile strain is generated is used. And a strain gauge that causes compressive strain. The strain gauge provided in the strain gauge deals with the above problem in the load cell, and the strain gauge is provided to increase the strain degree of the strain gauge. It is an object of the present invention to improve the load characteristics of this type of load cell, focusing on the fact that the nonlinear error increases in proportion to the increase in the output change amount.

[0017]

That is, the method of constructing a load cell according to the invention of claim 1 of the present application (hereinafter referred to as the first invention) is such that a fixed rigid body portion and a movable rigid body portion located at both ends are vertically moved. Using a substantially hollow quadrangular strain element in which a pair of beam portions is installed, it is provided in a strain element portion that causes a tensile strain and a strain element that causes a compressive strain when a load is applied to the movable rigid body portion. In a load cell configured to detect a load by a strain gauge, outputs of two strain gauges provided in a strain generating portion in which a tensile strain is generated and a strain generating portion in which a compressive strain is generated are different from each other with respect to a change in load. When a non-linear error in the opposite direction is generated, the degree of strain of at least one of the portions provided with the two strain gauges is changed so that the non-linear error is canceled out by the combined output characteristics of both strain gauges. Characterized by setting to characteristics To.

In the load cell according to the invention of claim 2 of the present application (hereinafter referred to as the second invention), strain gauges are provided in the strain generating portion on the fixed rigid body side and the strain generating portion on the movable rigid body side. In the load cell, the strain degree of the strain generating portion on the fixed rigid body side is different from the strain degree of the strain generating portion on the movable rigid body side.

[0019]

That is, according to the first aspect of the present invention, the strain gauges provided in the strain generating portion that causes tensile strain and the strain generating portion that causes compressive strain when a load is applied to the movable rigid body portion that constitutes the strain generating body. In a load cell configured to detect a load by means of a load cell, the outputs of two strain gauges provided in a strain generating portion in which a tensile strain occurs and a strain generating portion in which a compressive strain occurs If a non-linear error of
By changing the strain degree of at least one of the parts where the individual strain gauges are provided, the output characteristics that combine both strain gauges are set to a characteristic that cancels the nonlinear error so that the load cell is configured. Therefore, the combined output obtained by combining the two changes linearly with respect to the change in load, and good load detection accuracy can be obtained.

According to the second aspect of the invention, in a load cell in which strain gauges are provided on the strain generating portion on the fixed rigid body side and the strain generating portion on the movable rigid body side respectively, if the strain degrees of both strain generating portions are different. Therefore, the non-linear error of the load characteristic in this type of load cell is significantly improved, and even in this case, good load detection accuracy can be obtained.

[0021]

EXAMPLES Examples of the present invention will be described below.

As shown in FIGS. 1 and 2, in a load cell 21 according to the embodiment, two beam parts 24 and 25 are arranged in parallel between a fixed rigid part 22 and a movable rigid part 23 located at both ends. A hollow quadrangular strain element 26 that is erected, and a strain sensor 27 that is erected between a fixed side arm 22a and a movable side arm 23a that project from the fixed rigid body portion 22 and the movable rigid body portion 23 so as to face each other. Have and. 28d of substantially semicircular shape are provided on the inner surface side at the connecting portions of the upper and lower two beam portions 24, 25 constituting the strain generating body 26 with the rigid body portions 22, 23, respectively. 28d, the strain generating portions 29a ... 29d whose thickness is reduced are formed by the notches 28a. Also,
An end portion of the fixed rigid body portion 22 constituting the flexure element 26 is fixed to a base 31 standing on a pedestal 30, and a weighing pan 33 is attached to an end portion of the movable rigid body portion 23 via a bracket 32. It is supposed to be supported.

In the embodiment, the flexure body 26a having the fixed arm 22a and the movable arm 23a, and the four connecting bolts 34 on the upper surface of the body 26a.
The upper member 26 that is attached so as to be separable and separable via 34
The strain sensor 26 is formed by b and the strain sensor 27 is provided between the fixed side arm 22a and the movable side arm 23a as shown in FIG.

The strain sensor 27 has substantially the same structure as the strain sensor 6 shown in FIG. 12, and the thin rectangular metal substrate 35 has one end 35a attached to the fixed arm 22a. A pair of notches 36a, 36a whose adjacent side edge portions are cut to face each other, and the movable side arm 2
A pair of notches 36b, 36 having side edge portions adjacent to the other end portion 35b attached to 3a, which are cut to face each other.
b is provided, and the portions between the pair of notches are strain generating portions 37a and 37b, respectively. Also,
The insulating layer 38 provided on the metal substrate 35 includes a pair of gauge resistance portions 39a and 39b provided on the strain-flexing portions 37a and 37b, and the gauge resistance portions 39a and 39a.
39c is formed with a conductive portion 39c ... 39c connected to 39b. In this case as well, the gauge resistance portion 3 is provided in the conductive portion 39c ... 39c.
A flat cable 40 for extracting the resistance change of 9a and 39b as an electric signal is connected.

In this embodiment, one notch 36a, 36a corresponding to the fixed arm 22a is provided.
By making the width dimension between them smaller than the width dimension between the other notches 36b, 36b positioned corresponding to the movable side arm 23a, the strain degree of the strain generating portion 37a in which one gauge resistance portion 39a is formed is the other. Is set to be larger than the strain degree of the strain generating portion 37b in which the gauge resistance portion 39b is formed.

Here, as shown in FIGS. 4 and 5, the fixed arm 22a and the movable arm 23a of the strain-flexing element 26 have a pair of recessed stepped portions 41, which are cut in a rectangular parallelepiped shape, respectively. 41 are provided so as to face each other, and a hooking portion 42, in which the tip portions of these recessed step portions 41, 41 are further recessed,
42 is an end portion 35 of the metal substrate 35 that constitutes the strain sensor 27.
a and 35b are respectively hooked, and a lid plate fitted from above the recessed stepped portions 41 and 41 so as to sandwich the upper surfaces of the end portions 35a and 35b of the metal substrate 35. By fixing the bolts 43, 43 with bolts 44, 44 penetrating the cover plates 43, 43 and screwed into the movable side arm 22a or the fixed side arm 23a, the strain sensor 2
7 is fixed to the flexure body 26a.

Then, as shown in FIG. 6, two dummy resistors 45, 45 are connected to the pair of gauge resistors 39a, 39b in the strain sensor 27 so that a Wheatstone bridge circuit 46 is formed. Then, a predetermined DC voltage E is applied to both ends of both gauge resistance sections 39a and 39b, and an output terminal 47a connected between both gauge resistance sections 39a and 39b and an output connected between both dummy resistances 45 and 45. If the output voltage e output to the terminal 47b is taken out as a load signal, the weight of the object to be weighed on the weighing pan 33 can be measured.

According to the configuration of the load cell 21 according to the embodiment, when a predetermined load / load test is performed, the load characteristic in which the output voltage increases linearly with increase in load as shown by the solid line in FIG. Will be obtained.

This load test is carried out, for example, as follows, that is, a weight Y of a predetermined weight (for example, 0.5 kgf) is placed on the center position of the weighing pan 33 shown in FIG. 1, and Wheatstone at that time is placed. The output voltage e of the bridge circuit 46 is measured. When the measurement is completed, a weight Y is further added, the output voltage e is measured in the same manner, and the work is repeated until a predetermined maximum load (for example, 2.5 kgf) is reached.

In this case, when the load resistance test is conducted by replacing the gauge resistance portion 39a on the tension side in the Wheatstone bridge circuit 46 with an equivalent dummy resistance, as shown in FIG.
On the other hand, while showing a large non-linear error in which the output voltage is convex on the plus side as indicated by the broken line in FIG. 3, when the load resistance test is performed by replacing the gauge resistance part 39b on the compression side with an equivalent dummy resistance, this time the broken line shows As shown, a result showing a large non-linear error in which the output voltage is convex on the negative side was obtained. Then, the load Fo when the error voltage Vc becomes maximum when the tension-side gauge resistance portion 39a is replaced with a dummy resistance
Was loaded with the compression-side gauge resistance portion 39b replaced with a dummy resistance, the error voltage Vt at that time was substantially equal to the error voltage Vc. This is because, in the strain sensor 27, the width dimension of the strain generating portion 37a on which the tensile stress acts is reduced to increase the degree of strain of the strain generating portion 37a, and as a result, the gauge resistance portion 3 formed in the strain generating portion 37a.
It is considered that this is because the amount of change in the resistance value at 9a becomes large. Therefore, the tension side gauge resistance portion 3
Non-linear error due to 9a and gauge resistance part 39 on the compression side
The non-linear error caused by b is canceled out, and the output characteristic of the combination of both gauge resistance portions 39a and 39b changes linearly with the load as described above.

Next, an experiment conducted for setting the optimum size of the strain sensor will be described.

As shown in FIG. 8, a pair of notches 50a, 50a are formed on one end side of the metal substrate 49 formed in a rectangular shape so that the opposite side edge portions are cut in opposite directions.
The notches 50a and 50a are also formed on the other end side of the metal substrate 49.
A pair of notches 50b, 50b are also provided at intervals in the longitudinal direction from which the opposite side edge portions are cut so as to face each other. In that case, when the width dimension between the pair of notches located on the left side in the drawing is Wt and the width dimension between the pair of notches located on the right side is Wc, the width dimension ratio (Wt / Wc) Were prepared as shown in Table 1 below, and a plurality of sheets were prepared.

[0033]

[Table 1] Then, each pair of notches 50 in each metal substrate 49.
a, 50a; 50b and 50b, two gauge pattern forming parts 51 ...
Each sample was formed by forming a strain detection circuit similar to that shown in FIG. Then, as shown in FIG. 1, each of the test materials was incorporated into a flexure element, and the above-described load-loading test was performed, and the non-linear error was obtained for each of the test materials according to the following process.

First, after subtracting the output voltage in the no-load state (hereinafter referred to as the no-load voltage) from the output voltage when the maximum load is applied (hereinafter referred to as the maximum voltage), the value is divided by the maximum load. Then, the rate of change of the output voltage with respect to the load is calculated. Therefore, if the no-load voltage is added to the value obtained by multiplying the rate of change by a predetermined load, an ideal output voltage for the load can be obtained.

Next, for each measured load, the reference voltage obtained as described above is subtracted from the actual output voltage, and the maximum value is set as the error voltage, and the error voltage is calculated from the above maximum voltage. Divide the value by subtracting the no-load voltage,
The value obtained by multiplying the result by 100 was finally taken as the nonlinear error. In this case, if the actual output voltage is higher than the reference voltage, the nonlinear error becomes positive, and if the actual output voltage is small, the nonlinear error becomes negative.

The experimental results are shown in FIG. That is, it has been found that the smaller the width dimension ratio (Wt / Wc), the more the non-linear error tends to be negative, and that the non-linear error is close to 0 when the width dimension ratio (Wt / Wc) is approximately 0.6.

Next, another embodiment of the load cell 61 will be described with reference to FIG.

The load cell 61 according to this embodiment is shown in FIG.
As shown in 0, it has a fixed rigid body portion 62 and a movable rigid body portion 63 located at both ends, and a pair of upper and lower beam portions 64, 64 laid in parallel between these rigid body portions 62, 63,
The entire load cell 61 is an upper member 72 made of sheet metal in which rigid body constituent portions 70 and 71 that constitute the upper half of both rigid body portions 62 and 63 and an upper beam portion 64 are integrally formed,
Similarly, a lower member 82 made of sheet metal integrally formed with rigid body forming portions 80 and 81 and lower beam portions 64 that form the lower half portions of the rigid body portions 62 and 63, and the upper member 72 and the lower member 82. Rigid body part 70, 71; 8
The intermediate member 90 made of sheet metal sandwiched between 0 and 81 is formed into a hollow quadrangular shape symmetrical in the front-rear, left-right and up-down directions, and both ends of the strain sensor 27 of FIG. Is fixed to the rectangular hole portion of the intermediate member 90 between the pair of beam portions 64, 64 with the pair of notches 36a, 36a having a small width dimension being arranged on the fixed rigid portion side. Therefore, the pair of notches 36b, 36b having a large width dimension in the strain sensor 27 are arranged on the movable rigid body side.

Therefore, also in the load cell 61 according to this embodiment, the same effect as in the above example can be obtained.

The present invention is not limited to the three-beam type load cell as in the embodiment, and the Roberval mechanism is constituted by the left and right rigid body portions and the pair of beam portions installed in parallel between both rigid body portions. The present invention can also be applied to a two-beam type load cell in which a strain gauge is provided in the strain generating portion of the strain generating body.

[0041]

According to the present invention, the outputs of the two strain gauges provided in the strain generating portion in which the tensile strain occurs in the load cell and the strain generating portion in which the compressive strain occurs in the load cell are mutually different with respect to the change of the load. When a non-linear error in the opposite direction is generated, the above 2
By changing the strain degree of at least one of the parts where individual strain gauges are provided, the output characteristics that combine both strain gauges are set to characteristics that cancel the above nonlinear error. The output changes linearly with a change in load, and good load detection accuracy can be obtained.

According to the second aspect of the invention, in the load cell in which the strain gauges on the fixed rigid body portion side and the movable rigid body portion side are provided with strain gauges respectively, if the strain degrees of both strain generating portions are different. Therefore, the non-linear error of the load characteristic in this type of load cell is significantly improved, and even in this case, good load detection accuracy can be obtained.

[Brief description of drawings]

FIG. 1 is a front view showing a usage state of a load cell according to an embodiment.

FIG. 2 is an overall perspective view of a load cell.

FIG. 3 is an overall perspective view of a strain sensor.

FIG. 4 is a view seen from the line AA of FIG.

5 is a cross-sectional view taken along the line BB of FIG.

FIG. 6 is a load detection circuit diagram of the embodiment.

FIG. 7 is a schematic diagram showing load characteristics with respect to a load in an example.

FIG. 8 is a plan view showing the structure of a test material used in a load test.

9 is a characteristic diagram showing a relationship between a width dimension ratio between notches and a non-linear error in the test material of FIG.

FIG. 10 is a partial cutaway overall perspective view showing another embodiment of the load cell.

FIG. 11 is an overall perspective view showing a conventional example of a load cell.

FIG. 12 is an overall perspective view of a strain sensor used in the load cell.

FIG. 13 is a front view showing a usage state of the load cell.

FIG. 14 is a load detection circuit diagram of a conventional example.

FIG. 15 is a schematic diagram showing load characteristics with respect to load in a conventional example.

FIG. 16 is a schematic diagram showing a state before deformation of a load cell in a conventional example.

FIG. 17 is a schematic diagram showing an ideal deformed state of the load cell in the conventional example.

FIG. 18 is a schematic diagram showing an actual deformed state of the load cell.

[Explanation of symbols]

 21 Load Cell 22 Fixed Rigid Body Part 23 Movable Rigid Body Part 24, 25 Beam Part 26 Strain Element 27 Strain Sensor 35 Metal Substrate 36a, 36b Notches 37a, 37b Strain Part 39a, 39b Gauge Resistance Part

Front Page Continuation (72) Inventor Shotaro Tamai 959 Shimoho, Ritto-cho, Kurita-gun, Shiga 1 Ishida Hoki Co., Ltd. Shiga Factory

Claims (2)

[Claims] 1. A load is applied to the movable rigid body portion by using a generally hollow quadrangular flexure body in which a pair of upper and lower beam portions are installed between a fixed rigid body portion and a movable rigid body portion located at both ends. When a load cell is configured to detect a load by a strain gauge provided in a strain generating portion in which tensile strain occurs and a strain generating portion in which compressive strain occurs, a strain generating portion in which tensile strain occurs and a strain generating portion in which compressive strain occurs The output of the two strain gauges installed in
When a non-linear error in the opposite direction to each other with respect to the change of the load is generated, the degree of strain of at least one of the portions where the two strain gauges are provided is changed, and the combined output characteristics of both strain gauges are changed to the above-mentioned non-linearity. A method of constructing a load cell, which is characterized by setting characteristics such that linear errors cancel each other out. 2. A load cell in which strain gauges are provided on the strain generating section on the fixed rigid body side and the strain generating section on the movable rigid body side respectively, and the strain degree of the strain generating section on the fixed rigid body side is A load cell characterized in that the strain degree of the flexure portion on the rigid body side is different.