JPH06249675A - Signal processor - Google Patents

Abstract

(57) [Summary] (Correction) [Purpose] To provide a signal processing device improved so as to obtain an output with little error even if noise is mixed in an electric signal. An exciting coil 14 is fixed to the outside of the conduit 10, and an exciting current I ₁ is supplied from a constant current source inside an exciting circuit 15. The detection electrodes 11 and 12 are connected to the respective input ends of the differential amplifier 16. The sampling circuit 17 includes a sampling switch SW ₁ , a hold capacitor C, a buffer amplifier Q _{1 and the} like, and a sampling signal SMP ₁
The output voltage of the differential amplifier 16 is sampled by the holding capacitor C and held in the hold capacitor C, and the held hold voltage VH is fetched by the microprocessor 21 through the analog / digital conversion circuit 18 and executes a predetermined calculation to execute the flow rate. The signal FS is output to the output terminal 20 and a predetermined timing signal TS ₁ is output to the analog / digital conversion circuit 18 to control the conversion operation.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal processing device for converting a physical quantity into an electric signal, subjecting the electric signal to predetermined signal processing, and outputting the electric signal. The present invention relates to a signal processing device improved so as to obtain a small output.

[0002]

2. Description of the Related Art An electromagnetic flow meter will be described as an example of a signal processing device. FIG. 4 shows a configuration of a conventional electromagnetic flow meter 1
It is a block diagram which shows an example. Reference numeral 10 is a ceramic pipe or a non-magnetic conduit made of stainless steel or the like with an inner lining for insulation.

Further, a pair of detection electrodes 11 and 12 are fixedly insulated from the conduit 10. The liquid contact electrode 13 is connected to the common potential point COM in order to ground the measurement fluid Q.

An exciting coil 14 is fixed to the outside of the conduit 10, and an exciting current I ₁ is supplied to the exciting coil 14 as a constant current from a constant current source inside the exciting circuit 15. On the other hand, the detection electrodes 11 and 12 are connected to respective input terminals of a differential amplifier 16 (the amplification degree is 1 for simplicity), and output to the sampling circuit 17 after common mode noise and the like are removed here. To be done.

The sampling circuit 17 includes a sampling switch SW ₁ , a hold capacitor C and a buffer amplifier Q _1.
The output voltage of the differential amplifier 16 is sampled by the sampling signal SMP ₁ and held in the hold capacitor C. The held hold voltage VH is taken into the microprocessor 19 via the analog / digital conversion circuit 18.

The microprocessor 19 executes a predetermined calculation using the fetched data and outputs it as a flow rate signal FS to the output end 20 and a predetermined timing signal TS.
₁ is output to the analog / digital conversion circuit 18 to control the conversion operation.

Further, the microprocessor 19 sends the sampling signal SMP ₁ for controlling the sampling switch SW ₁ of the sampling circuit 17 to the sampling switch S _1.
Output to W ₁ .

In addition to this, the microprocessor 19 causes the exciting circuit 15 to switch the exciting current I ₁ with the timing signal T.
The waveform of the exciting current I ₁ is controlled by outputting S ₂ and switching the constant current from the built-in constant current source.

Next, the operation of the electromagnetic flow meter constructed as above will be described with reference to the waveform diagram shown in FIG. Microprocessor 19 outputs a timing signal TS ₂ in the excitation circuit 15, the timing signal TS _2, a rectangular wave excitation circuit 15 to repeat the positive as shown in FIG. 5 (a), zero, negative, zero as 1 cycle An exciting current I ₁ having a waveform of is applied to the exciting coil 14.

Due to this exciting current I ₁ , a magnetic field B having substantially the same waveform as this is applied to the measurement fluid Q, and the detection electrode 1
An electromotive force e having a waveform as shown in FIG.

This electromotive force e has a form in which a rectangular wave signal voltage V ₀ resulting from the flow rate of the measurement fluid Q and a differential noise voltage N ₀ generated by switching the magnetic field B are superimposed. Of these, if only the noise voltage N ₀ is taken out,
It has a shape shown in FIG.

The differential amplifier 16 amplifies this electromotive force e and outputs it to the sampling circuit 17. The electromotive force e is sampled by the switch SW ₁ whose opening and closing is controlled by the sampling signal SMP ₁ . Timings t ₁ and t ₂ of the first and second periods for sampling are shown in FIG.
As shown in (3), it is a period of width T / 8 immediately before the exciting current I ₁ changes.

The hold voltages e _H1 and e _H2 held at these timings t ₁ and t ₂ are taken into the random access memory of the microprocessor 19 through the analog / digital conversion circuit 18. Hold voltage e _H1 , e _H2
Is represented by e _H1 = v ₁ + n ₁ (1) e _H2 = −n ₂ (2).

However, v ₁ is a signal voltage component, and n ₁ and n ₂ are differential noise components. If the sampling is performed in the third and fourth periods, the changing DC voltage component can be removed, but the description is omitted for simplicity.

Using the data shown in the equations (1) and (2) stored in the random access memory, the microprocessor 19 performs an operation of e _H1 + e _H2 = v ₁ + (n ₁ -n ₂ ) (3). Run. The signal processing is similarly performed in the third and fourth periods.

In these cases, if the differential noises n ₁ and n ₂ are small, then n ₁ ≈n ₂ and therefore the data (e _H1 + e _H2 )
Becomes equal to the signal voltage component v ₁ , and a correct signal can be obtained without being affected by noise.

However, the electromagnetic flowmeter to which the processing method (1) described above is applied shows the correct flow rate signal when the differential noises n ₁ and n ₂ are small.
_{When 1} and n ₂ become large, they cannot be ignored and cause an error.

Therefore, a processing method (2) is conceivable in which the waveform of the differential noise is identified, the differential noise is accurately calculated, and the differential noise is removed by using this. This processing method (2) will be described with reference to FIG.

FIG. 6 (a) is an excitation waveform having the same period T as the excitation waveform shown in FIG. 5 (a), and FIG. 6 (b) is a waveform of electromotive force including differential noise generated by this excitation, FIG.
(C) is a sample signal SMP ₂ for sampling the signal
Are shown respectively.

Electromotive force e generated on the detection electrodes 11 and 12
Is sampled by the switch SW ₁ whose opening and closing is controlled by the sampling signal SMP ₂ . The timings t _1A ′, t _1B ′, t _2A ′ and t _2B ′ of the 1A, 1B, 2A and 2B periods of sampling are shown in FIG.
T / 16 immediately before the exciting current I ₁ changes as shown in
Are two consecutive periods of width.

These timings t _1A ′, t _1B ′, t _2A
Hold voltage e _H1 ′, e _{H2 held} at ′, t _2B ′
′, E _H3 ′, e _H4 ′ are analog / digital conversion circuits 1
8 into the random access memory of the microprocessor 19.

Hold voltage e _H1 ′, e _H2 ′, e _H3 ′, e
_H4 ′ is e _H1 ′ = v ₁ + n ₁ ′ (4) e _H2 ′ = v ₁ + n ₁ ′ · r (5) e _H3 ′ = n ₂ ′ (6) e _H4 ′ = n ₂ ′ · r ( 7).

Here, n ₁ ′ and n ₂ ′ are timing t _1A
It is the differential noise at the time of ′, t _2A ′, and r is the attenuation constant represented by r = 1 / [exp (T / 16τ) (8). Note that τ is the decay time constant of the differential noise.

Therefore, from the equations (4) to (8), the signal voltage component v ₁ is as follows: v ₁ = (e _H3 ′ · e _H2 ′ _{−e H4} ′ · e _H1 ′) / (e _H3 ′ _{−e H4} ′) (9) The calculation can be performed by the microprocessor 19. That is, when the differential noise is large, the noise waveform can be identified by appropriate sampling to reduce the influence thereof, that is, the output error.

[0025]

However, in the signal processing method 2, when the amplitude of the differential noise is small or the time constant τ is small, the denominator of the expression (9) approaches zero and the output becomes indefinite. There is inconvenience.

Therefore, when the processing method 1 is suitable when the amplitude of the differential noise is small, and when the processing method 2 is suitable when the amplitude is large, it is conceivable to detect and switch an appropriate noise level. In this case, generally, since the outputs do not match, there is a problem in that they become discontinuous in terms of time and cause inconvenience.

[0027]

According to the present invention, a signal for converting a physical quantity into an electric signal by a sensor and subjecting the electric signal to a predetermined signal processing to output a sensor signal is provided as a structure for solving the above problems. In the processing device, the electric signal is input, the signal is processed by a predetermined first processing method, the first sensor signal X ₁ is output, and the noise index value N _x included in the electric signal is detected. a first signal processing means, second signal processing means for outputting a second sensor signal X ₂ and signal processing with different second processing scheme the first processing scheme previously with respect to the previous electric signal, the previous index When the magnitude of the value N _x is less than or equal to the first predetermined value N _a , the first sensor signal X ₁ is output, and when the value is greater than or equal to the first predetermined value N _a , the shift determination to the second processing method is performed. 1 determination means and the size of the above index value N _x The second sensor signal X ₂ is output when the second predetermined value N _{b is} larger than the first predetermined value N _a and the second sensor signal X ₂ is output when the second predetermined value N _b is _smaller than the second predetermined value N _b . Using the determination means and the index value N _x , the first predetermined value N _a , the second predetermined value N _b , the first sensor signal X ₁ , and the second sensor signal X ₂ , X ₃ = X ₁ [(N _{b -N x) / (N b -N} a)] + X 2 [(N x
-N _a ) / (N _b -N _a )] is executed to output the third sensor signal X ₃ from the third signal processing means having the third processing method. is there.

[0028]

[Operation] The first signal processing means receives an electric signal whose physical quantity is converted by a sensor, processes the signal by a predetermined first processing method, outputs a first sensor signal X ₁ , and includes the electric signal in the electric signal. The index value N _x of the generated noise is detected.

The second signal processing means performs signal processing on the previous electric signal by a second processing method different from the first processing method and outputs a second sensor signal X ₂ . First judging means and the second processing ahead of when the magnitude of the preceding index value N _x is when below a first predetermined value N _a and outputs a first sensor signal X ₁ previously described above the first predetermined value N _a Determine the transition to the method.

The second judging means outputs the second sensor signal X ₂ and outputs the second predetermined sensor signal X ₂ when the index value N _x is equal to or larger than the second predetermined value N _b which is larger than the first predetermined value N _a . When the value is less than or equal to N _b , the shift to the third processing method is determined.

The third signal processing means uses the above third processing method to obtain the index value N _x , the first predetermined value N _a , the second predetermined value N _b ,
The first sensor signal X _1, second by using the sensor signals _{X 2 X 3 = X 1 [} (N b -N x) / (N b -N a)] + X 2 [(N x
-N _a ) / (N _b -N _a )] is executed to output the third sensor signal X ₃ .

By the eclectic calculation by the third signal processing means, the index value N is set between the first sensor signal X ₁ by the first signal processing means and the second sensor signal X ₂ by the second signal processing means. It is possible to continuously output the sensor signal in the optimum state regardless of the variation of _x .

[0033]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 is a block diagram showing the configuration of one embodiment of the present invention. FIG. 1 is a calculation flowchart showing the calculation procedure of the embodiment shown in FIG. The configuration shown in FIG. 2 is basically the same as the configuration shown in FIG. 4 except for the contents of signal processing by the microprocessor.

Although not shown, the microprocessor 21 includes a random access memory (RAM) as a temporary data storage memory, and a read-only memory (RO) as a program storage memory for an arithmetic program.
M), microprocessor for controlling operation (μ
P), a bus connecting these components, and the like are stored.

The operation of the embodiment shown in FIG. 2 will be described below with reference to the operation flow chart shown in FIG. First, in step 1, timing t ₁ in the waveform diagram shown in FIG.
The signal voltage is sampled and held at a time point such as t ₂ and data e _H1 (= v ₁
+ N ₁₎ or stored as e _H2 (= -n _2).

Next, in step 2, the flow rate calculation is executed by the equation (3) based on the processing method (1), and the flow rate signal (X ₁ ′) is stored in a predetermined area of the RAM in the microprocessor 21 as data. [E _H1 + e _H2 (= v ₁ + (n ₁ −
n ₂ ))] is stored.

In step 3, the magnitude is taken as the noise index value, and the magnitude of the noise and the magnitude of the predetermined value are compared and judged. ROM in microprocessor 21
In the processing method (1) is stored _a predetermined noise value N _a that can be ignored by the influence of noise, and a judgment procedure for comparing the magnitude of the noise data N _x with the predetermined value N _a . The noise data N _x is the data e _H2 of the zero excitation period at the timing t ₂ stored in step 1, which represents the magnitude of noise.

The microprocessor 21 reads the determination procedure stored in the ROM and the predetermined value N _a , compares it with the noise data N _x stored in the RAM, and determines the magnitude relationship. If the relationship of N _x <N _a is satisfied, the flow rate signal v ₁ (=
X ₁ ) is output. If the relationship of N _x <N _a is not satisfied, the process proceeds to step 5.

In step 5, the signal voltage is sampled and held at timings t _1A ′, t _1B ′, t _2A ′, t _2B ′ in the waveform diagram shown in FIG. e _H1 ′, e _H2 ′, e _H3 ′,
It is stored as _eH4 '.

Next, the process proceeds to step 6, where the micro program is
Equation (9) stored in advance in the ROM
The data stored in RAM is calculated using the calculation procedure shown in.
E _H1´ 、 e_H2´ 、 e_H3´ 、 e_H4Read out ´
Issue v₂(= X₂) Is calculated and stored in the RAM.

In step 7, the magnitude of the noise and the magnitude of the predetermined value are compared and judged. In the ROM of the microprocessor 21, a predetermined value N _{b of} noise (N _a <N _b ) in which the influence of noise is negligible in the signal processing method (2), noise data N _x, and the predetermined value N _b are large or small. Stored is a judgment procedure for comparing the.

The microprocessor 21 reads the determination procedure stored in the ROM and the predetermined value N _b , compares it with the noise data N _x stored in the RAM, and determines the magnitude relationship.

If the relationship of N _x <N _b is satisfied, step 8
Flow rate signal v ₂ (= X ₂ ) stored in step 6
Is output. If the relationship of N _x <N _b is not satisfied, the process proceeds to step 9.

In step 9, signal processing is performed by the third processing method 3 (eclectic method). This calculation program is ROM
The microprocessor 21 uses the data stored in the RAM to execute the operation according to the following equation (10).

[0045] _{X 3 = X 1 [(N} b -N x) / (N b -N a)] + X 2 [(N x -N a) / (N b -N a)] (10) where X ₁ , X ₂ , X ₃ are flow rate signals v ₁ , v ₂ , respectively.
This is a sensor signal representative of v ₃ .

After the above calculation is executed, the process proceeds to step 10 to output the flow rate signal v ₃ as the sensor signal X ₃ .
For the above processing methods 1, 2, and 3, the noise level N
The relationship between the error ε and _x is as shown in FIG. The error in the processing method 3 has a characteristic close to the approximate straight line indicated by the one-dot chain line.
Can be minimized.

In the above description, an example in which the two processing methods are combined by the third processing method has been described, but the present invention is not limited to this, and similar effects can be obtained by sequentially switching some processing methods. it can.

Further, in the embodiment shown in FIGS. 1 and 2, the switching area is set by focusing on the size of the noise index value, but another index of the noise attribute is also considered.
It is also possible to set the switching area and the processing method in the dimension.

Further, in the embodiment, the flow rate and other unknowns are obtained by sampling four times during a half cycle. However, if the amplitude ratio of noise at positive and zero excitation is calculated, the unknown Since the number is reduced, it is possible to identify noise with three samplings.

[0050]

As described above in detail with the embodiments, according to the present invention, the processing method with the smallest error can be selected and used according to the noise index value, for example, its size. It is possible to obtain a sensor output with a small error regardless of fluctuations in noise, and it is possible to obtain a stable output without discontinuity in output that occurs due to switching of processing methods.

[Brief description of drawings]

FIG. 1 is a flowchart showing a calculation flow of an embodiment of the present invention.

FIG. 2 is a configuration diagram showing a configuration of an embodiment for executing the calculation shown in FIG.

FIG. 3 is a characteristic diagram illustrating the effect of the present embodiment.

FIG. 4 is a configuration diagram showing a configuration of a conventional electromagnetic flow meter.

5 is a first waveform diagram illustrating an operation of the electromagnetic flow meter shown in FIG.

FIG. 6 is a second waveform chart illustrating the operation of the electromagnetic flow meter shown in FIG.

[Explanation of symbols]

 10 Conduit 11, 12 Detection Electrode 14 Excitation Coil 15 Excitation Circuit 16 Differential Amplifier 17 Sampling Circuit 19, 21 Microprocessor

Claims (1)

[Claims] 1. A signal processing device for converting a physical quantity into an electric signal by a sensor, subjecting the electric signal to a predetermined signal processing to output a sensor signal, wherein the electric signal is input by a predetermined first processing method. First signal processing means for performing signal processing to output a first sensor signal X ₁ and detecting a noise index value N _x included in this electric signal, and the first processing method for the electric signal are different. Second signal processing is performed by the second processing method
A second signal processing means for outputting a sensor signal X _2, wherein when the index value N _x is greater than or equal the first predetermined value N _a to output the first sensor signal X ₁ when equal to or lower than the first predetermined value N _a A first judging means for judging the shift to the second processing method; and when the index value N _x is _a second predetermined value N _b which is larger than the first predetermined value N _a or more, the second sensor signal X ₂ is outputted. Second
When less than a predetermined value N _b and a second determining means for a shift determination to the third processing method, the index value N _x, the first predetermined value N _a,
Second predetermined value N _b , first sensor signal X ₁ , second sensor signal X
X ₃ = X ₁ using _{2 [(N b -N x)} / (N b -N a)] + X 2 [(N x
-N _a ) / (N _b -N _a )] and outputs a third sensor signal X ₃ by a third signal processing means having the third processing method. Processing equipment.