JP2001358587A - Nonlinear/linear converter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide precision linearization by using an A/D converter and a calculator of low cost with less bits. SOLUTION: An 8-bit type is employed for an A/D converter 2A, a calculator 2B, and a D/A converter 2C. A nonlinear analogue input signal is supplied to both the A/D converter and positive side of a subtracter 2D. The output of A/D converter 2A is supplied to the calculator 2B, the output of calculator 2B is supplied to the D/A converter 2C, and the output of the D/A converter 2C is supplied to a negative side of the subtracter 2D. The calculator 2B is provided with a correction value conversion table 2B1. A correction value (digital value) for converting a nonlinear analogue input signal at the subtracter 2D into a linear one is stored corresponding to the digital value (input value) of 256 steps from the A/D converter 2A in the correction value conversion table 2B1.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a nonlinear / linear conversion for linearizing (linearizing) a non-linear electric signal such as a sine wave or an output signal from a resistance temperature detector, a thermocouple, a thermistor, a differential pressure type flow meter or the like. It is about a vessel.

[0002]

2. Description of the Related Art Conventionally, a linearizer as shown in FIG. 10 has been used as this kind of nonlinear / linear converter. The linearizer 1 includes an A / D converter 1A, a computing unit 1B, and a D / A converter 1C, and converts a non-linear analog input signal into a linear analog signal.

In FIG. 10, an A / D converter 1A includes:
For example, electromotive force (input voltage V) from a thermocouple (not shown)
Is provided as a non-linear analog input signal. A / D
The converter 1A converts an input non-linear analog input signal into a digital value. Here, it is assumed that a 12-bit A / D converter 1A, a computing unit 1B, and a D / A converter 1C are used, and a non-linear analog input signal is converted into a digital value with a resolution of 1/4096. .

FIG. 12A illustrates a non-linear analog input signal. In FIG. 12A, the horizontal axis is the temperature of the measurement target, and the vertical axis is the A / D converter 1A according to the temperature of the measurement target.
2 shows an input voltage V to the input. In this example, if the temperature span of the measurement target is 0 to 650 ° C., the A / D converter 1A
The value (maximum value) Vmax of the input voltage V corresponding to 650 ° C. from the value (minimum value: 0 V in this example) of the input voltage V corresponding to the temperature of 0 ° C. is set to the full span, and the input voltage V is set to 4096.
It is converted into digital values of the stages (see FIG. 12B).

[0005] The non-linear analog input signal converted into a digital value by the A / D converter 1A is supplied to a 12-bit arithmetic unit 1B. The arithmetic unit 1B is provided with a linear conversion table 1B1. That is, the A / D converter 1A
A digital value (linear conversion value) for linearly converting a non-linear analog input signal is stored in correspondence with the digital value (input value) of 4096 steps from (see FIG. 11). When a digital value is input from the A / D converter 1A, the arithmetic unit 1B obtains a linear conversion value corresponding to the digital value (input value) from the linear conversion table 1B1, and obtains the D / A
Output to converter 1C. FIG. 12C shows a state of conversion from nonlinear to linear in the arithmetic unit 1B. The D / A converter 1C converts the digital value from the arithmetic unit 1B into an analog value. In this way, a linear analog signal is obtained.

In this linearizer 1, the A / D converter 1
A, a computing unit 1B, and a D / A converter 1C use 12-bit ones, and use a linear conversion value having the maximum value Vmax of the input voltage V as a full span.
Assuming that ax is 100%, 100% ＝ 4096 = 0.
Linearization can be performed with an accuracy of about 024%, and can be sufficiently used as a linearizer that requires an accuracy of 0.1% or less.

In order to obtain sufficient accuracy in the above-described conventional nonlinear / linear converter (linearizer 1), an A / D converter
The converter 1A, the arithmetic unit 1B, and the D / A converter 1C generally have a high resolution of 12 bits or more.

[0008]

As described above, the conventional linearizer 1 is expensive because at least the A / D converter 1A and the arithmetic unit 1B have 12 bits or more. On the other hand, for example, an 8-bit A / D
If a converter or the like is used, the cost is low, but the accuracy of 0.1% or less cannot be secured. That is, when an 8-bit A / D converter is used, the full span Vmax is set to 1
If it is set to 00%, linearization can be performed only with an accuracy of about 100% ÷ 256 = 0.4%, and cannot be used as a linearizer that requires an accuracy of 0.1% or less.

The present invention has been made to solve such a problem, and an object of the present invention is to use a low-cost A / D converter and an arithmetic unit with a small number of bits to achieve high-precision linearization. It is an object of the present invention to provide a nonlinear / linear converter which can be converted.

[0010]

In order to achieve the above object, the present invention converts a non-linear analog input signal into a digital value, obtains a correction value according to the digital value,
This correction value is converted into an analog value to obtain an analog correction signal, and the non-linear analog input signal is combined with the analog correction signal to convert the non-linear analog input signal into a linear analog signal. According to the present invention, a non-linear analog input signal is converted to a digital value, a correction value (digital value) corresponding to the digital value is obtained, and the correction value is converted to an analog value to be an analog correction signal. . Then, the analog correction signal is combined with the non-linear analog input signal (addition or subtraction).
The non-linear analog input signal is converted to a linear analog signal. In this case, the correction value may be the difference between the non-linear analog input signal and the linear analog signal. If the non-linearity of the non-linear analog input signal is 25% of the full span, that is, the difference between the non-linear analog signal and the linear analog signal Assuming that the maximum value of the difference is 25% of the full span, for example, even if A / D conversion of 8 bits is performed, 25% ÷ 2
Linearization can be performed with an accuracy of about 56 = 0.1%.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail based on embodiments. FIG. 1 is a block diagram of a linearizer showing an embodiment of a nonlinear / linear converter according to the present invention. The linearizer 2 outputs an analog voltage signal having a linear temperature / voltage characteristic by using, for example, an electromotive force of a thermocouple having a non-linear temperature / voltage characteristic as a non-linear analog input signal. D converter 2A,
It comprises an arithmetic unit 2B, a D / A converter 2C and a subtractor 2D. These A / D converter 2A and arithmetic unit 2B
And the D / A converter 2C uses an 8-bit one.

In the linearizer 2, the non-linear analog input signal (input voltage V) is supplied to the A / D converter 2A and also to the + side of the subtractor 2D. A /
The D converter 2A converts the input voltage V to a digital value of 256 levels, with the maximum value Vmax being a full span from 0 [V] of the input voltage V corresponding to 0 ° C. to 650 ° C. (see FIG. 2). The non-linear analog input signal converted into a digital value by the A / D converter 2A is supplied to the calculator 2B.

The arithmetic unit 2B has a correction value conversion table 2B1.
Is provided. The correction value conversion table 2B1 includes A
25 where Vmax from / D converter 2A is full span
Subtractor 2D corresponding to the digital value (input value) of 6 stages
, A correction value (digital value) for linearly converting a non-linear analog input signal is stored (see FIG. 3). That is, in the present embodiment, the relationship between the nonlinear analog input signal (characteristic I shown in FIG. 4) input to the linearizer 2 and the linear analog signal to be output by the linearizer 2 (characteristic II shown in FIG. 4). Is known, the difference between the non-linear analog input signal and the linear analog signal is obtained as a characteristic III, and the maximum value VSm of the difference is obtained.
A digital value of 256 steps where ax is a full span is set as a correction value (see FIG. 5), and Vmax from the A / D converter 2A is
Is stored in association with a digital value of 256 steps with full span.

When a digital value is input from the A / D converter 2A, the arithmetic unit 2B obtains a correction value corresponding to the digital value (input value) from the correction value conversion table 2B1, and calculates this correction value as D / D Output to A converter 2C. The D / A converter 2C converts the correction value from the arithmetic unit 2B into an analog value, and supplies the analog value to the negative input of the subtractor 2D. The subtracter 2D converts the non-linear analog input signal into a linear analog signal by subtracting the negative analog correction signal from the non-linear analog input signal supplied to the + side.

In this case, if the nonlinearity of the nonlinear analog input signal is 25% of the full span Vmax, that is, the nonlinear analog input signal (characteristic I shown in FIG. 4) and the linear analog signal (characteristic II shown in FIG. 4) Is 25% of full span Vmax, the 8-bit A /
Even if the D converter 2A, the arithmetic unit 2B, and the D / A converter 2C are used, linearization can be performed with an accuracy of about 25% ÷ 256 = 0.1%. Thus, high-precision linearization can be performed using the inexpensive 8-bit A / D converter 2A, arithmetic unit 2B, and D / A converter 2C.

FIG. 6 is a circuit diagram showing a specific example of the linearizer according to the present invention. The linearizer 3 is a microcomputer having an 8-bit A / D converter, a CPU, and an I / O port [for example, a RISC (Reduced Instruction)
ion Set Computer) microcomputer] IC1 and operational amplifier I
C2-A, IC2-B, resistors R1 to R8, and capacitors C1 and C2.

A non-linear analog input signal (input voltage V) from an input terminal Iin is supplied to an operational amplifier I via a resistor R1.
The signal is supplied to the non-inverting input of C2-A and to the microcomputer IC1 via the connection point between the resistor R7 and the capacitor C1. FIG. 7 shows the microcomputer I
The outline of the internal configuration of C1 is shown. Microcomputer I
C1 is CPU3-1, RAM3-2, ROM3-3.
A / D converter 3-4, interface (I / O,
I / F) 3-5, 3-6. The ROM 3-3 stores a correction value conversion table. CPU 3-1
Performs a processing operation according to a program stored in the ROM 3-3.

In the microcomputer IC1, a non-linear analog input signal (input voltage V) is supplied to an A / D converter 3-4. A / D converter 3-4
Converts the maximum value Vmax of the input voltage V corresponding to 650 ° C. into a full span, and converts the input voltage V into a 256-stage digital value. The CPU 3-1 includes an A / D converter 3-4.
And a correction value corresponding to the digital value (input value) is obtained from a correction value conversion table in the ROM 3-3, and a PWM signal having a duty ratio corresponding to the correction value is obtained. Is created as software and output via the interface 3-6.

PWM from microcomputer IC1
The signal is applied to a smoothing circuit including a resistor R6 and a capacitor C2, and a voltage signal having a value corresponding to the duty ratio is applied to a non-inverting input of the operational amplifier IC2-B. That is, a voltage signal corresponding to the correction value obtained by the microcomputer IC1 is supplied to the operational amplifier IC2-
B is applied to the non-inverting input of B. Operational amplifier IC2-B
Distributes a voltage signal according to the correction value obtained by the microcomputer IC1 into positive and negative according to the setting of the bias voltage Vb, and supplies the analog signal as a analog correction signal to the non-inverting input of the operational amplifier IC2-A.

Accordingly, when the characteristic of the nonlinear analog input signal is such that it swells upward with respect to the linear analog signal, the analog correction signal is subtracted from the nonlinear analog input signal and is converted linearly from the operational amplifier IC2-A. Analog signal is output. In the case where the non-linear analog input signal has a characteristic that swells downward with respect to the linear analog signal, an analog correction signal is added to the non-linear analog input signal, and the analog signal converted linearly from the operational amplifier IC2-A. Is output.

In FIG. 6, it is assumed that a non-linear analog input signal expands vertically with respect to a linear analog signal.
In each case, a bipolar type circuit configuration in which correction is performed is adopted. If the non-linear analog input signal has a characteristic that expands only upward with respect to the linear analog signal, a unipolar circuit configuration as shown in FIG. 8 is used, and the non-linear analog input signal becomes a linear analog signal. On the other hand, in the case of the characteristic of expanding only downward, FIG.
By adopting a unipolar circuit configuration such as
Since the operational amplifier IC2-B is not required, the circuit configuration can be simplified.

In the above-described embodiment, the operation unit 2
In B, the correction value for converting non-linear to linear is determined from the correction value conversion table, but may be determined using a calculation formula. In the microcomputer IC1, the duty ratio PW corresponding to the correction value
Although the M signal is created by software, it may be created by a hardware circuit.

In the above-described embodiment, the resistance R6
The PWM signal from the microcomputer IC1 is converted (A / D-converted) into a voltage signal having a value corresponding to the duty ratio by a smoothing circuit including the capacitor C2 and the D / A converter. You may do so.

If the PWM signal is generated by a hardware circuit or a D / A converter is provided, the microcomputer IC1 becomes expensive. In the present embodiment, a PWM signal is created in a software manner, and D / A conversion is performed by a smoothing circuit including a resistor R6 and a capacitor C2, so that an A / D converter, a CPU, an I / O as a microcomputer IC1.
An inexpensive RICS microcomputer with an / O port is used to reduce costs.

In the above-described embodiment, the electromotive force (input voltage V) from the thermocouple is linearized as a non-linear analog input signal. However, a resistance temperature detector, a thermistor, a differential pressure type flow meter, etc. And the linearization of a non-linear electric signal such as a sine wave. In addition, it is also possible to nonlinearize a linear signal by the same procedure as that for linearizing a nonlinear signal. That is, by using the correction value conversion table, it is possible to convert linear data into any function such as sine wave, square root, and logarithm.

[0026]

As apparent from the above description, according to the present invention, a non-linear analog input signal is converted into a digital value, and a correction value corresponding to the digital value is obtained. Is converted into an analog correction signal, and this analog correction signal is combined (added or subtracted) with a non-linear analog input signal, and the non-linear analog input signal is converted into a linear analog signal. In this case, the correction value may be the difference between the non-linear analog input signal and the linear analog signal. If the non-linearity of the non-linear analog input signal is 25% of the full span, it is assumed that 8-bit A / D conversion is performed. Also, 25% ÷ 256 = 0.
Linearization can be performed with an accuracy of about 1%, and high-precision linearization can be performed using an inexpensive A / D converter and an arithmetic unit with a small number of bits.

[Brief description of the drawings]

FIG. 1 is a block diagram of a linearizer showing an embodiment of a nonlinear / linear converter according to the present invention.

FIG. 2 shows an example of an input voltage V having a full span of Vmax.
It is a figure showing the conversion situation to 56 steps of digital values.

FIG. 3 is a diagram showing a correction value conversion table for linearly converting a non-linear analog input signal.

FIG. 4 is a diagram showing a difference signal (characteristic III) between an input non-linear analog input signal (characteristic I) and an output linear analog signal (characteristic II).

FIG. 5 is a diagram showing a state of conversion of a correction value having a maximum difference VSmax as a full span into a digital value in 256 steps.

FIG. 6 is a circuit diagram showing a specific example (bipolar type) of the linearizer according to the present invention.

FIG. 7 is a diagram schematically showing an internal configuration of a microcomputer in the linearizer.

FIG. 8 is a diagram showing another circuit configuration example (unipolar type) of the linearizer.

FIG. 9 is a diagram showing another circuit configuration example (unipolar type) of the linearizer.

FIG. 10 is a block diagram showing a conventional linearizer.

FIG. 11 is a diagram showing a linear conversion table used for the linearizer.

FIG. 12 is a diagram illustrating a non-linear / linear conversion operation in the linearizer.

[Explanation of symbols]

2 Linearizer, 2A A / D converter, 2B arithmetic unit, 2B1 correction value conversion table, 2C D / A converter, 2D subtractor, 3 linearizer, IC1 microcomputer, IC2-A, IC2-B: operational amplifier, R1 to R8: resistor, C1, C2: capacitor, 3-
1 CPU, 3-2 RAM, 3-3 ROM, 3-4
... A / D converter, 3-5, 3-6 ... interface.

Claims (1)

[Claims] A / D conversion means for converting a non-linear analog input signal into a digital value, correction value calculation means for obtaining a correction value corresponding to a digital value from the A / D conversion means, and correction value calculation D / A conversion means for converting the correction value obtained by the means into an analog value to obtain an analog correction signal, and combining the analog correction signal with the non-linear analog input signal to convert the non-linear analog input signal into a linear signal. A non-linear / linear converter comprising: a correction unit for converting the signal into an analog signal.