JPH0575339B2 - - Google Patents

Description

[Detailed Description of the Invention] <Industrial Application Field> The present invention relates to an electromagnetic flowmeter that applies a magnetic field to a fluid to be measured and measures its flow rate, and particularly relates to an excitation method thereof and a signal processing method associated therewith. This invention relates to an improved electromagnetic flowmeter.

<Prior Art> Industrial electromagnetic flowmeters have conventionally adopted a commercial frequency excitation method using a commercial power supply. The commercial frequency excitation method has (a) fast response speed and low cost. (b) It has the advantage of being less susceptible to the effects of low-frequency random noise (hereinafter referred to as flow noise) that increases with flow velocity that occurs with slurry liquids and low conductivity fluids, but it However, it has the disadvantage that the zero point will fluctuate if left unattended for a long period of time, for example, about one day.

For this reason, a low frequency excitation method that excites at a low frequency of 1/2 of the commercial frequency or lower has been adopted. As is well known, the low frequency excitation method has the advantage of providing an electromagnetic flowmeter with a stable zero point. However, since the excitation frequency is low, it is close to the frequency of flow noise, and is therefore susceptible to the influence of flow noise, and this influence becomes particularly noticeable as the flow velocity increases. Furthermore, when damping is applied to reduce the influence of flow noise, the response becomes slow.

Therefore, as proposed in Japanese Patent Application No. 60-197168 (title of invention: electromagnetic flowmeter), a composite magnetic field is created by simultaneously flowing an excitation current component at a commercial frequency and an excitation current component at a lower frequency into an excitation coil. A composite excitation method has been proposed to form

<Problems to be Solved by the Invention> However, in this proposal, if differential noise generated due to switching of the excitation current, deformation of the excitation coil, or attachment of artifacts to the electrode is superimposed on the signal voltage, this problem may occur. There is a problem that zero error occurs due to this.

<Means for Solving the Problems> In order to solve the above problems, the present invention provides excitation means for supplying magnetic fields having two different frequencies, a first frequency and a second frequency lower than the first frequency;
a first demodulating means for discriminating and outputting a signal voltage excited by the exciting means and generated corresponding to the flow rate based on the first frequency; and a high-pass filtering means for high-pass filtering the output of the first demodulating means. , a second demodulating means for discriminating and demodulating the signal voltage based on a second frequency, a low-pass filtering means for low-pass filtering the output of the second demodulating means, and each of the high-pass filtering means and the low-pass filtering means. addition means for additively combining the outputs, the excitation means supplies an excitation current having a waveform obtained by multiplying the first frequency and the second frequency, and the first demodulation means supplies the excitation current having a waveform obtained by multiplying the first frequency and the second frequency Each time the level of the excitation current of the frequency component changes, three consecutive sampling values of the signal voltage sampled are multiplied by coefficients of -1, +2, and -1, respectively, and added. If a change in the level of the excitation current of the second frequency is included during these three consecutive samplings, instead of the three sampling values, each of the four consecutive sampling values is multiplied by 1, - Addition operations are performed by multiplying coefficients of 1, 1, and -1 by a sign determined by the excitation level of the second frequency.

<Example> Hereinafter, an example of the present invention will be described based on the drawings. FIG. 1 is a block diagram showing one embodiment of the present invention.

10 is a conduit for a detector of an electromagnetic flowmeter, and an insulating lining is provided on the inner surface of the conduit. 11
a and 11b are electrodes for detecting a signal voltage. Reference numeral 12 denotes an excitation coil, and a magnetic field generated thereby is applied to the fluid to be measured. An excitation current I _f is supplied to the excitation coil 12 from an excitation circuit 13 .

The excitation circuit 13 is configured as follows. The reference voltage E ₁ is applied to the non-inverting input (+) of the amplifier Q ₁ , the output of which is connected to the base of the transistor Q ₂ . The emitter of the transistor _Q2 is connected to the common COM via a resistor R _f and also to the inverting input terminal (-) of the amplifier _Q1 . An excitation voltage E _S is applied between the common COM and the collector of the transistor Q ₂ through the series circuit of switches SW ₂ and SW ₃ and the switch SW ₄ connected in parallel to this.
Applied via SW ₅ series circuit. The excitation coil 12 is connected to the connection point of switches SW ₂ and SW ₃ and the switch
Connected to the connection points of SW ₄ and SW ₅ , respectively. Timing signals S ₂ , S ₃ , S ₄ , and S ₅ are switches, respectively.
Controls the opening and closing of SW ₂ , SW ₃ , SW ₄ , and SW ₅ .

On the other hand, the signal voltage is detected by the electrodes 11a and 11b and output to the preamplifier 14. Preamplifier 1
4, the common mode voltage is removed and the impedance is converted, and the resultant signal is output to the output terminal 15.

The output of the preamplifier 14 at the output terminal 15 is converted into a digital signal by an analog/digital converter (A/D _L ) 16 and an analog/digital converter (A/D _H ) 17, respectively, and is randomly transmitted via a bus 18. The data is stored in access memory (RAM) 19. A read-only memory (ROM) 20 stores predetermined calculation programs and initial data, and is processed under the control of a processor (CPU) 21.
The calculation is performed according to the calculation procedure stored in the ROM 20, and the result is stored in the RAM 19. 2
Reference numeral 2 denotes a clock generator, and the clock generated here is divided into 1/n by a frequency divider 23 and supplied to the CPU 21 and the analog/digital converter 17 as a system clock Sh.

The CPU 21 outputs the timing for determining the waveform of the excitation current I _f to the timing signal output port (TO) 24 via the bus 18 in accordance with the arithmetic program stored in the ROM 20 . The timing signal output port 24 outputs timing signals S ₂ , S ₃ , S ₄ , and S ₅ for switching the excitation current according to this timing.

In addition, the timing signal output port 24 is
Timing signal according to timing specified by 21
S _l is output to analog/digital converter 16 and the output of preamplifier 14 is sampled.

On the other hand, using the data stored in RAM 19 by the calculation program stored in ROM 20,
A predetermined calculation is executed by the CPU 21, and the result of the calculation is stored in the RAM 19 and the bus 1
8, the digital/analog converter 25, and the output end 26 as a flow rate output.

Next, the operation of the embodiment shown in FIG. 1 will be explained using the timing diagram shown in FIG. 2, the flowchart shown in FIG. 3, and the operation diagram shown in FIG. 4.

The system clock _Sh obtained from the output of the frequency divider 23 shown in FIG. 1 has a waveform shown in FIG. 2a, and is supplied to the CPU 21.

In step 1 of FIG. 3, the CPU 21 sends a timing signal to the timing signal output port 24 via the bus 18 in synchronization with the interrupt timing of the system clock _Sh (see g in FIG. 2) in accordance with a predetermined calculation program stored in the ROM 20. A timing signal indicating the switching timing of the excitation waveform is output.

In step 2, the timing signal output port 24 receives this switching timing and outputs timing signals S ₅ (FIG. 2b), S ₄ (FIG. 2c), S ₃ (FIG. 2d), and S ₂ (FIG. 2d). Figure e) is output to the switches SW ₅ , SW ₄ , SW ₃ , and SW ₂ of the excitation circuit 13, respectively. Alternatively, output the timing signal S ₄ to the switches SW ₃ and SW ₄ at the same time, and output the timing signal S ₂ to the switch SW 3 and SW 4 simultaneously.
It is also possible to output to SW ₂ and SW ₅ at the same time. Excitation circuit 1
3 receives these timing signals and outputs an exciting current I _f having a waveform shown in FIG. 2 f to the exciting coil 12 . This excitation waveform is a waveform in which the timing number i is 0 to 15, forming one cycle and repeating it, as shown in Fig. 2 h and i, and Fig. 2 mainly shows the n cycle part. . This excitation waveform has a multiplicative waveform obtained by multiplying a low frequency waveform and a high frequency waveform.

Next, proceed to step 3. From step 3 to step 6, the analog/digital converter 16,
The procedure for reading data from 17 is shown.

In step 3, the system clock S _h (Figure 2a)
The data inputted from the analog/digital converter 17 every cycle in synchronization with
The data is stored in a predetermined data area H _i of the RAM 19.

Next, the process moves to step 4, where it is determined whether the read timing number i is O or not.

In step 6, it is determined whether or not the read timing number i is 8. If it is not 8, the process moves to step 8, and if it is 8, the process moves to step 7.

In step 7, the timing signal output port 2
The data input from the analog/digital converter 16 is converted into the data inputted from the analog/digital converter 16 by the sample timing according to the timing signal S _l (Fig. 2 k) outputted from the analog/digital converter 16.
As _shown in FIG.
-1), L ₀ (n), L ₀ (n+1), etc., and proceed to step 5.

Next, in step 6, the timing signal S _l (second
According to the sample timing according to FIG.
Under the control of step 1, the data is stored in predetermined data areas . . . , L ₁ (n-1), L ₁ (n), L ₁ (n+1), .

In step 8, it is determined whether the timing signal i is an odd number or not. If it is an odd number, the process moves to step 9, and if it is not an odd number, it is judged that the process moves to step 11.

Step 9 performs high frequency demodulation calculations. During demodulation calculation, the data H _i stored in the RAM 19
using the CPU 21 at the timing shown in Figure 2 m.
Under the control of , the high frequency demodulation calculation shown in _FIG . Through this demodulation calculation, the electrochemical DC voltage generated across the electrodes 11a and 11b is removed, and the differential noise is kept at a constant value and does not become an error factor. This calculation will be described in detail later. In addition, in Fig. 4, A
The constant is A= _Tc /(Tc
+ΔTc).

Next, move on to step 10. Here, a high-pass filtering operation F _Hi on the high frequency side is executed.

When performing filtering calculations, the CPU uses data e _Hi stored in the RAM 19 and the previous filtering calculation results.
Under the control of 21, the calculation is performed using the calculation formula shown in the column of high-pass filter calculation F _Hi shown in FIG.

Next, move on to step 11. In step 11, it is determined whether the timing number i is 0 or 8. If it is 0 or 8, the process moves to step 12, and if it is not 0 or 8, it is judged to move to step 14.

In step 12, low frequency demodulation calculations are performed. During the demodulation calculation, the data stored in the RAM 19..., L ₀ (n-1), L ₀ (n), L ₀ (n+1), L ₁ (n
-1), L ₁ (n), L ₁ (n+1), ..., and the ROM is controlled by the timing CPU 21 shown in FIG.
Low frequency demodulation calculation shown in Fig. 4 stored in 20
e Perform the calculation using the formula shown in _{the Li} column and write the result.
Store in RAM19. In addition, in Fig. 4,
Constant B is expressed as B=ΔT/(ΔT+T).

In step 13, low-pass filter calculation on the low frequency side is performed.
Run F _Li .

When performing the filtering calculation, the data e _L0 and e _L8 stored in the RAM 19 and the previous filtering calculation result are used.
Under the control of the CPU 21, the calculation is performed using the calculation formula shown in the column of low-pass filter calculation F _Li shown in FIG. 4, which is stored in the ROM 20, and the result is stored in the RAM 19.

In step 14, it is determined whether the timing number i is an odd number or not. If it is an odd number, the process proceeds to step 15, and if it is not an odd number, the process proceeds to step 16.

Step 15 performs an addition operation. Based on the control of the CPU 21, the result F _Hi of the high-pass filter calculation and the result F _Li of the low-pass filter calculation stored in the RAM 19 are used.
Addition operation e _A shown in Fig. 4 stored in ROM20
Perform calculations using the formula shown in the column and store the results in RAM.
19 and move on to step 17.

In step 17, the process waits until the next interrupt timing, and when the next interrupt timing arrives, the flow from step 1 to step 17 is executed again.

As described above, the signal voltage containing two frequencies, a low frequency and a high frequency, detected by the electrodes 11a and 11b is
The low frequency side and the high frequency side are read separately using a microcomputer, and the low frequency side is demodulated at low frequency and its output is passed through a low-pass filter, and the high frequency side is demodulated at high frequency and output. By outputting each through a high-pass filter, and adding and combining the outputs of the low-pass filter and high-pass filter, the cello point is stable, resistant to flow noise, and the response is stable. Good flow output can be obtained.

Furthermore, using FIGS. 5 and 6, it is shown that the high-frequency demodulation calculation e _Hi in the present invention can also eliminate the effects of differential noise caused by switching the excitation current or electrochemical DC voltage generated across the electrodes. I will explain.

The excitation current I _f shown in Fig. 5c is a transcription of the excitation current I _f shown in Fig. 2f, +I ₀ , 0, -I ₀
The waveform has three values, and the repetition of high frequency 0/+I ₀ and 0/-I ₀ is switched at low frequency. This means that the high frequency component is 0/1 and the low frequency component is -1/+
By considering that state 1 is repeated in each cycle, the exciting current I _f can be seen as a multiplicative combination of the waveform shown in FIG. 5c and the waveform shown in FIG. 5b.

At each timing point…, t _-1 , t ₀ , t ₁ , ~t ₁₅ ,…
The excitation current value in (Fig. 5a) is in the I _f column of Fig. 6,
The signal voltages associated with this are listed in the _eS column, assuming that a constant flow rate is flowing. Also,
The waveform of the differential noise e _o1 accompanying the switching of the excitation current I _f shown in FIG. 5 d is shown in FIG. 5 e,
The size at each timing point is listed in the column e _o1 in FIG. 6. The polarity of differential noise is determined by the direction of change in the excitation current, but since the direction of change is reversed every half cycle of the low frequency, differential noise also changes in the same cycle. In addition, the waveform of the electrochemical DC voltage e _o2 generated at the electrode deforms at a constant rate with time as shown in Figure 5, but the magnitude at each timing is as shown in Figure 6. It is written in the _o2 column. However, the DC voltage e _o2 is shown as having e ₀ at the timing point t ₀ and changing by ε as the timing point shifts by one.

Combining the above, the voltage e obtained as the output of the preamplifier 14 is as follows.

e=e _S +e _o1 +e _o2 ...(1) Here, if we simply calculate the voltage e _Hi from the difference between the voltage sample values e _i and e _i-1 at the two excitation levels, we get
If P _i is 1 when the low frequency waveform (Figure 5b) is at high level and -1 when it is low level, then e _Hi = P _i (e _i-1 − e _i )...(2) . However, i=1, 3, 5, . . . , which corresponds to executing one flow rate calculation for each fundamental cycle of the high frequency. The result of this calculation is shown in calculation result A in FIG. Looking at this result,
Since the change ε in the DC voltage is superimposed on the signal voltage and the differential noise and these are changing, ε remains as an error factor.

Next, by performing the calculation shown in the following formula, e _Hi = - (P _i-2 ×e _i-2 −P _i-1 ×e _i-1 ) + (P _i-1 ×e _i-1
−P _i ×e _i )=−P _i-2 ×e _i-2 +2P _i-1 ×e _i-1 −P _i ×e _i …(3)
becomes. However, i=1, 3, 5, . . . The result of this calculation is shown in calculation result B in FIG. 6, but the variation ε of the DC voltage and the like differ at each sampling time and become a cause of error. Here, according to calculation result B in FIG. 6, if the coefficients P _i-2 to P _i of the low frequency waveform change during sampling of the signal at timings (i-2) to i, Including errors.

Therefore, only when the state of the low frequency waveform changes and the coefficient P _i changes during the sampling of the signals from (i-2) to i, the calculation of the following equation is performed, and the calculation result C in Figure 6 is shown. Get results.

e _Hi =P _i-3 ×e _i-3 −P _i-2 ×e _i-2 +P _i-1 ×e _i-1 −P _i ×e _i
...(4) However, i=1, 9, 17, 25...

Therefore, as shown in the column of high frequency demodulation calculation e _Hi in FIG. 4, at timing points t ₁ , t ₉ , t ₁₇ . . .
At other timing points, by calculating based on equation (3), a high-frequency demodulated signal e _Hi is obtained in which the included differential noise e _o1 is constant and the fluctuation ε of the DC voltage is removed by the calculation. This high frequency demodulated signal e _Hi
is subjected to high-pass filtering in step 10 of Figure 3, so a certain differential noise component is also removed from the output F _Hi of step 10, and as a result, a high-frequency component flow rate signal free of errors can be obtained. .

In addition, in equations (1) to (4), the polarity of the flow rate signal is determined as shown in the column of signal voltage e _S in Figure 6, so the sign of each term in the calculation shown in Figure 4 is reversed. However, this is related to operational polarity and is not essential.

<Effects of the Invention> As specifically explained above in conjunction with the embodiments, according to the present invention, the magnet is excited by a two-frequency wave that is a multiplicative combination of two waveforms, a low frequency wave and a high frequency wave, and generated in response to the flow rate. By performing calculations on the signal voltage using two calculation formulas, the effects of differential noise contained in the signal voltage and DC voltage fluctuations caused by electrochemical causes can be removed, resulting in an error-free output. It can be done.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention;
Fig. 2 is a waveform diagram explaining the operation of the embodiment shown in Fig. 1, Fig. 3 is a flowchart showing the signal processing procedure of the embodiment shown in Fig. 1, and Fig. 4 is the flowchart of Fig. 3. Figure 5 is a waveform diagram illustrating how the calculation in Figure 4 can remove a certain amount of noise, and Figure 6 is a waveform diagram showing how the calculation in Figure 4 can remove a certain amount of noise at the time of pulling. FIG. 10... Conduit, 12... Excitation coil, 13...
Excitation circuit, 16, 17... Analog/digital converter, 18... Bus, 19... Random access memory, 20... Read only memory, 21...
...Microprocessor, 22...Clock generator, timing signal output port.

Claims (1)

[Claims] 1 Excitation means for supplying magnetic fields having two different frequencies, a first frequency and a second frequency lower than the first frequency, and a signal voltage excited by the excitation means and generated corresponding to the flow rate based on the first frequency. a first demodulating means for discriminating and outputting; a high-pass filtering means for high-pass filtering the output of the first demodulating means; a second demodulating means for discriminating and demodulating the signal voltage based on the second frequency; The excitation means includes low-pass filtering means for low-pass filtering the output of the second demodulation means, and addition means for additively combining the respective outputs of the high-pass filtering means and the low-pass filtering means, and the excitation means includes the supplying an excitation current having a waveform obtained by multiplying the first frequency and the second frequency;
The demodulation means increases the sampling value by -1 times, +2 times, and -1 times, respectively, for three consecutive sampling values of the signal voltage sampled every time the level of the excitation current of the first frequency component changes. If a change in the level of the excitation current of the second frequency is included during the three consecutive samplings, the four consecutive sampling values are used instead of the three sampling values. An electromagnetic flowmeter characterized in that an addition operation is performed by multiplying the coefficients of 1, -1, 1, and -1 by a sign determined by the excitation level of the second frequency.