JPS58115324A - Electromagnetic flowmeter - Google Patents

Abstract

PURPOSE:To improve a noise components and an offset voltage components with small power consumption, by practically performing a prescribed calculation when signal voltages of respective periods in which the stationary values of exciting current are negative, zero, positive, zero, negative, zero, positive, and zero are defined as prescribed values. CONSTITUTION:Induced voltage ea corresponding to exciting current Iw is generated between electrodes 23, 24 of an oscillator 2 in an electromagnetic flowmeter. An offset voltage components and a noise component are superposed to the induced voltage ea in addition to flow component Vs. Consequently signal voltages ea1-ea9 obtained by sampling the induced voltage ea at a fixed interval during the periods when the stationary values of the exciting current Iw are negative, zero, positive, zero, negative, zero, positive, and zero are found by the fomula 1, respectively. A signal processing circuit 3 practically operates (ea6-3ea5+2ea4+2ea3-3ea2+3a1) or (ea8-3ea7+2ea6+2ea5 -3ea4+ea3) to remove the offset voltae component Vn and the noise component Ve approximate to a cubic formula.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an improvement in a low frequency excitation type electromagnetic flow needle.

In general, an electromagnetic flow needle is configured to apply a magnetic field perpendicularly to the fluid flow direction, simultaneously detect changes in electrical signals in the fluid flow path, and measure the fluid flow rate based on this. ing. Most recent electromagnetic flow needles use low-frequency excitation methods, such as trapezoidal wave excitation and square wave excitation, which have superior zero point stability compared to AC excitation and DC excitation methods. There is. In this type of low-frequency excitation type electromagnetic flowmeter, in particular, an excitation current is supplied to the excitation coil of the electromagnetic flowmeter transmitter with a steady value that repeats in the order of negative, zero, positive, and zero. The signal voltage eal"a2" during each period when the steady value of the excitation current is negative, zero, positive, and zero.
Using a3"a4, the signal processing circuit can effectively convert K Cea
1+ea□) -(ea3+ea4'7) The device that calculates By the way, even with this type of electromagnetic flow needle, when the electrochemical DC voltage fluctuates and the offset voltage component changes, it is affected by this, and output fluctuations cannot be avoided. reaches several hundred mV, which is extremely large compared to the span of the flow rate component (1 to 10 mV), and its transient change cannot be regarded as a straight line.Therefore, the signal voltages ea2 and ea3 are multiplied by coefficients, respectively. After (
By performing the calculation ea4-3ea2 '' 3ea2- eal), the offset voltage component can be removed by quadratic approximation. However, according to experiments, the offset voltage component swings in the positive and negative directions, but the swing is between -1v and +1v. It was confirmed that the potential change always has an inflection point.Therefore, when removing the offset voltage component by quadratic approximation, it is necessary to remove the inflection point of the potential change. In addition, the minimum value of the excitation current in an electromagnetic flowmeter is determined by the ratio to the input conversion noise of the signal processing circuit, and since the input conversion noise is almost vector added, (ea + ea2' - ea3-ea4)
There is a book that performs the calculation (ea4-3ea3 ”3ea2
-ea□) is the ratio of 5yJ〒1”
: D-[Size=-284,4□, and the one that performs the latter calculation has 2.24 times the magnitude of the rounded excitation current that obtains the same magnitude of flow signal, and the power consumption is 5 It will be twice as big.

In the present invention, the steady value of the excitation current given from the electromagnetic flowmeter transmitter is negative, zero, positive, zero, negative, zero, positive, zero, and the signal voltages in each period are eal"a2"a3'ea4'ea
5"a6"a7' ea8, the signal processing circuit effectively calculates K(ea6-3ea5+2ea4+2ea3-3
ea2+ea1) or (ea8-3efL7+2ea
6+2ea5-3ea4+ea3) to realize a low frequency excitation type electromagnetic flow needle that can remove offset voltage components by cubic approximation with low power consumption.

FIG. 1 is a connection diagram showing an embodiment of the electromagnetic flowmeter of the present invention. In the figure, 1 is an excitation circuit that includes a DC constant current source 11 and a switch 12.1 that switches the current from the constant current source 11.
3. 2 is an electromagnetic flowmeter transmitter, which includes an exciting coil 21, a pipe 22 through which a current flows, and an electrode 23.2.
) 3 is a signal processing circuit, which is connected in series with an AC amplifier 31 that amplifies the voltage e induced between the electrodes 23 and 24 of the electromagnetic flowmeter transmitter 2, and to which the output voltage eb of the AC amplifier 31 is applied. sample hold circuit 3 on the 6 side
2 to 32. and these sample and hold circuits 32 ~
The outputs ec~af ec6 of 32 are sent to the coefficient multipliers 331~33.33, respectively. an arithmetic circuit 34 added through the arithmetic circuit 34, an inverter 35 that inverts the output ed of the arithmetic circuit 34, a switch 36 that switches between the output ed of the arithmetic circuit 34 and the output ex of the inverter 35, and a switch 36 selected by the switch 36. A sample hold circuit 37 that samples and holds e or e, a pulse Pla'lb that controls the switches 12 and 13 of the excitation circuit 1, and a pulse P2a = p2f that controls sampling of the sample hold circuits 32 to 32f, Changeover switch 36
pulse p3 that controls the sample and hold circuit 3
r? It has a timing pulse generating circuit 8 which generates a pulse p4 for controlling the stop/bring at the timing shown in FIG. The coefficients 1 to 6 of the coefficient units 33 to 33f are respectively 1-K6, so-called 1. 5-3 degrees are selected for K2- and 4-2 are selected for K3-, and the arithmetic unit 34 calculates (ecl-3e02" 2e03+ 2e04-3e
c5+"c6)" addition/subtraction operation is performed.

The operation of the present invention constructed in this way will be explained below with reference to the waveform diagram in FIG. 2. First, the switches 12 and 13 are controlled by drive pulses p1... During the period T when pla is on, the current from the constant current source 11 is sent to the excitation coil 21 in the positive direction, and plb
During the period T3 when Pl is on, the current Iw from the constant current source 11 is switched in the opposite direction to flow through the exciting coil 21 (7, Pl
During the period T2 and T4 in which both a''lb are shown as OFF, no current is passed through the excitation coil 21. Therefore, the excitation coil 21
As shown in FIG. 2, C has an excitation period T with a positive steady-state value, a negative excitation period T3, and a rest period T2IT4 with a zero value.
An excitation current Iw whose steady-state value repeats in the order of positive, zero, negative, and zero is supplied. Note that the excitation current IW is controlled by switches 12 and 13.
When switched, the constant current source 11 can only supply a finite output voltage, so it actually rises and falls with a delay due to the time constant due to the inductance and resistance of the exciting coil 21, and then reaches a steady value. is omitted in the figure. There is an excitation current wire wK between the electrodes 23 and 24 of the electromagnetic flowmeter transmitter 2.
A corresponding induced voltage e is generated, and the induced voltage eK is a as shown in Fig. 2.
a, the flow rate FK of the fluid flowing through the pipe 22
In addition to the proportional flow rate component v8, there is an offset voltage component vn due to an electrochemical DC potential or circuit, and a noise component V accompanying switching of the excitation current. are superimposed. As a result, the steady-state values of the excitation current shown in Figure 2 are negative, zero, positive, zero, negative,
Signal voltage eat"a2"a3 obtained by sampling the induced voltage e generated in each period of zero, positive, zero, and negative at a defined interval Δt
``a4''``a5''a6''a7'%8''a9 are given by the following equations.The offset voltage component V is expressed by Taylor expansion 1 and approximated by a cubic equation.

In the signal processing circuit 3, first, the induced voltage ea of the electromagnetic flow pointer oscillator 2 is amplified by the amplifier 31, and the signal voltage e, L1 to 'a9 K phase”4t;
The output of the b amplifier 31 is sequentially transferred to the sample hold circuit 321.
is held. Sample hold time @ 32. The hold value "cl" is held in the sample hold circuit 32b by the sampling pulse P2bK generated at the timing shown in FIG. 2. Similarly, the hold values ec2 to e05 of the sample hold circuits 32b to 32e are shown in FIG. 't y ) IJ that occurs at a timing like
y/pulses P2° to P2fK, sample and hold circuits 32 to 32. is held. Then, the sample and hold circuit 32a holds the voltage related to the signal voltage "a6," and at the time of death, the sample and hold circuits 32b to 32f hold the voltages related to ea5 to elL1, respectively, and the sample and hold circuits 32. At the point when the voltage associated with K'a7 is held, the sample and hold circuits 32b to 32K are eI
Voltages associated with L6 to 01□ are held. These sample and hold circuits 321-32. The hold voltages ``cl~e, s'' are the coefficient multipliers 331~33., respectively.
"cl-3e02+ 2ec3" in the arithmetic circuit 34 via
+2e, -3e, 5+·136) are performed.

As a result, at the output terminal of the arithmetic circuit 34, the output of the amplifier 31 corresponding to the sample and hold circuit 32sLK signal voltage @6K is held, the own arithmetic result ed1, and the output of the amplifier 31 corresponding to the signal voltage 01□ are held. The calculation result "d2" when the signal voltage e,8 is held, the calculation result e when the output of the amplifier 31 corresponding to the signal voltage e,8 is held, and the calculation result "d2" when the output of the amplifier 313 corresponding to the signal voltage -8 is held. A waveform voltage e is obtained by sequentially repeating the calculation result ed4. Then, e6, ~ed
4 are given by the following equations, where k is the gain of the amplifier 31.

edlxk(ea6-3elL5+2ea,+2e,3
-3e, 2+elL1)w4kV, (2)e
d211Ik(e, 7-36.6+2e, 5+2@,,
-3ea3+8,2)-4k(V,+2V,) (3
) ed3-k (eIL8-3e, 7+2e, 6+2e,
5-3ea, +esL3)--4kV,
(4) ed4-IJ-, -3e, 8+2e, 7+2
e, 6-3aIL5+e', 4)-4k(V, +2V,
) (5) Therefore, the output of the amplifier 31 corresponding to the signal voltage when the steady-state value of the sample-and-hold circuit 5zKIIlk current is positive or negative is held at the output of the arithmetic circuit 34, and the excitation current is not switched at the time of death. Although the accompanying noise component V is superimposed, the output of the amplifier 31 corresponding to the color signal voltage whose steady value is zero is held, and at the time of death, the offset voltage components vnO to vn3 approximated by the quintic equation are removed. The noise component ve accompanying the switching of the excitation current is also removed, and only the flow rate component VS can be obtained.

Therefore, the selector switch 36 is driven by the pulse p3 generated at the timing shown in FIG. 2, and the output of the arithmetic circuit 34 is
When □, the output e6 of the arithmetic circuit 34 and the output ed3 of the arithmetic circuit 34 are taken out via the inverter 35 and the output e6 of the arithmetic circuit 34 is taken out via the inverter 35, and the output ed3 is sent to the pull-hold circuit 37 alternately with the sampling pulse p generated at the timing shown in FIG. give. As a result, an output voltage e connected to the flow rate component VKfi is obtained at the output end of the sample hold circuit 37.

In this way, in the present invention, the noise component V and the offset voltage component accompanying switching of the excitation current can be removed by fifth-order approximation, and the offset voltage component can be effectively removed even when the offset voltage component has an inflection point. can. Moreover, the signal processing circuit 5
The input equivalent noise is 12+32+ 22+ 22+ 3
Since the flow rate component V is multiplied by 4, which is expressed as + 1 dew 5.68, it takes more time to obtain the same flow rate signal of large heat than when the offset voltage component is removed by quadratic approximation. The required excitation current may be 59χ, and the power consumption d is 35%.
K can also be reduced. Furthermore, this method has the characteristic that no integration error occurs due to fluctuations in flow rate, and no offset occurs due to ramp-like flow rate changes.

In addition, in the above, the steady value of the excitation current is negative, zero, positive zero, negative,
Using the calculation result of equation (2) using the signal voltage for each period of zero and the signal voltage for each period of positive, zero, negative, zero, positive, and zero (4
9 is an example of a case in which the calculation results of the formula ( ) are output alternately, but either one may be output. However, the case where the signals are output alternately as in the embodiment has the advantage that the response is twice as good.

Furthermore, in the above description, the case where the output eb of the amplifier 31 is directly transferred to the sample and hold circuit 32K is illustrated, but as shown in FIG. You can also sum it up to give. In this case, if the integration time T8 is selected to be an integral multiple of the commercial power supply cycle, the influence of power supply frequency noise can be removed. In FIG. 5, the integrator 39 includes a resistor RI, an operational amplifier Op, an integrating capacitor CI connected to the feedback circuit of op, a timing switch TS that controls the input integration time T8, and resets the integral value. has a reset switch R3, T8 and R8 are pulses P5. The one driven by P6 is illustrated. Further, the signal processing circuit 5 includes a microprocessor that performs digital calculations, as shown in FIG.
It may also be configured using 0. In FIG. 4, the output e of the integrator 39 is converted into a digital signal by the A/D converter 41 and given to the microprocessor 40. 4) A digital calculation corresponding to the formula is performed and a digital value corresponding to the flow rate component VIC is output, which is converted into an analog signal by the D/A converter 42 and held in the sample hold circuit 37, and the flow rate component VK is output to the output terminal. An example is shown that yields an associated output voltage eo. In FIG. 4, the integrator 39 applies the reference voltage 1 via the switch TSI driven by the pulse p7, and the A/D converter 4 integrates e over a certain period of time.
After converting the signal into a digital signal in step 1, the signal may be sent to a microprocessor 40 to perform span adjustment or the like. Furthermore, although the above example uses a rectangular wave as the excitation current, various waveforms may be used as necessary, such as a waveform obtained by converting the frequency of a trapezoidal liquid or a commercial AC power source and rectifying it after death. I can do it.

In this way, in the actual IN, the signal voltage in each period when the steady value of the excitation current changes in the order of negative, zero, positive, zero, negative, zero, positive, zero is eee ・al' a2 la a3
'a4'``a5'``a6'%7' ``A8, 9 and L signal processing circuit effectively K (''---3e, s + 2e, 4
+ 2e, 3-3e, L2 +”al ) t Nine is (
e, 8-3e, 7+ 2e,, + 2e, 5-3*,
4+ @IL3) Since the computation is performed, it is possible to obtain a low-frequency excitation type electromagnetic flow needle that can remove noise components and offset voltage components associated with excitation current switching using quintic approximation with low power consumption. It will be done.

[Brief explanation of drawings]

Figure 1 is a connection diagram showing one embodiment of the electromagnetic flow needle of the present invention, Figure 2 is a rounded waveform diagram explaining its operation, Figures 5 and 111.
FIG. 4 is a connection diagram showing another embodiment of the electromagnetic flow needle of the present invention. 1... Excitation circuit, 2... Electromagnetic flow pointer transmitter, 21.
... Excitation coil, 23.24 ... Electrode, 3 ... Signal processing circuit, 31 ... AC amplifier, 32-32 ...
・Sample hold circuit a f circuit, 34... Arithmetic unit, 37... Sample hold circuit, 39... Integrator, 40... Microphone equipment processor 7,
41...A/D converter, 42...D/A converter,
38...timing bar k x generator.

Claims (1)

[Claims] If the excitation coil of the electromagnetic flowmeter transmitter has a steady value of zero, negative,
An excitation circuit that repeatedly supplies an excitation current with a positive request and a steady value of an excitation current given from an electromagnetic flowmeter transmitter are used to detect signal voltages during each period of negative, zero, positive, zero, negative, zero, positive, and zero. 'al
"a2""a3"a4'ea5'ea6' ea7
``When a8 is set, the actual K (ea6-3ea5+2e1
4+2ea3-3ea2 ” “at” or “a8-3”
eEL7” 2elLe + 2e, 5-3elL4+
an electromagnetic flow needle having a signal processing circuit that performs calculations such as ea3) and obtains a flow rate signal by removing an offset voltage component by cubic approximation.