TRANSFORMER EMF OF AN ELECTROMAGNETIC FLOWMETER AND METHODS FOR ITS SUPPRESSION

I. N. Kraev, V. N. Titov

(Presented by the scientific seminar of the Faculty of Physics and Engineering)

It is known that in electromagnetic flowmeters one of the sources of interference is the alternating magnetic field of the instrument, which induces in the measuring circuit a voltage not associated with the motion of the liquid. This voltage (transformer emf) arises because in the alternating magnetic field there is a loop formed by the two electrodes, the column of liquid between them, and the leads coming from the electrodes.

The magnitude of the transformer emf depends on the area of the loop penetrated by the magnetic flux, the intensity of the magnetic field, and the frequency of its variation.

The area of the loop must be reduced, while leaving a small value of the emf induced in it for compensation of stable interferences of the same frequency arising from inhomogeneity of the magnetic field. The leads from the electrodes should be carefully shielded. Usually the magnitude of the transformer emf is comparable with the magnitude of the useful signal and amounts to several mV.

Suppression of the transformer emf can be carried out in the sensor itself, at the input of the secondary instrument, and also in such a way that, when the secondary instrument possesses phase selectivity, the recording instrument does not respond to the transformer emf.

Reduction of the errors caused by the transformer emf can be accomplished by precise adjustment of the orientation of the leads from the collecting electrodes in the form of a noninductive loop, as shown in Fig. 1.

The loop formed by the liquid, the electrodes, and the leads from the electrodes lies in a plane parallel to the magnetic lines of force and therefore is surrounded by zero flux. Consequently, by the corresponding orientation of the loop it is possible to obtain a minimum signal from the sensor at zero flow rate [1, 2].

Figure 2 shows a circuit for suppressing the transformer emf when two leads are made from one of the electrodes. The movable leads embrace the pole of the electromagnet on two sides and are closed through a low-resistance potentiometer. The voltage obtained from the sensor is taken from the slider of the potentiometer and the second electrode. By changing the position of the potentiometer slider, with the liquid stationary, such a ...

its position such that the signal at the input to the amplifier is equal to zero [3, 4].

Another suppression method [5] is based on the use of two transducers connected as shown in Fig. 3. If the transformer e.m.f.’s arising in both transducers are equal in phase and magnitude, then they are compensated in the primary winding of the transformer. The useful e.m.f.’s, proportional to the magnitude of the flow rate in each receiver, are opposite in phase and are added in the pri—

![Fig. 1 diagram]

Fig. 1. Circuit for reducing errors caused by transformer e.m.f. by orienting the leads from the electrodes.

![Fig. 2 diagram]

Fig. 2. Circuit for suppressing transformer e.m.f. with a tapped lead to a potentiometer.

![Fig. 3 diagram]

Fig. 3. Circuit for suppressing transformer e.m.f. by means of two transducers.

secondary winding of the transformer. Since it is practically impossible to create two identical sensors and identical conditions for the leads from the electrodes, it is therefore necessary that the midpoint of the primary winding of the transformer be variable.

The complication of the design of electromagnetic flowmeters with such a method of compensating the transformer emf has led to their use only for laboratory investigations of flowmeters.

To reduce the influence of the transformer emf at the input of the secondary instrument, a symmetrical differential amplification circuit is used. In this case the parasitic emf will have the same phase and magnitude in each input lead and will be amplified much less than the useful signal, whose phases in the input leads of the amplifier are opposite.

The main difficulty in using this method is the need for a transformer stage for transition from the symmetrical amplification circuit to the ordinary one.

Compensation of the transformer emf can be accomplished by means of an additional coil placed in a special socket in the flowmeter branch pipe. For compensation, the voltage induced by the instrument’s magnetic field in this coil is used [6].

When the magnetic-field strength changes, the compensating voltage changes in the same way as the measured signal. Since the voltage induced in the coil is shifted in phase relative to the signal by 90°, the instrument circuit must use a phase-shifting cell. This method can be used by taking the compensating voltage not from a special coil but from the power winding of the transformer feeding the sensor electromagnet, Fig. 4.

Fig. 4. Circuit for suppressing the transformer emf by a voltage taken from the power winding of the transformer.

transformer [7], feeding the electromagnet of the sensor, Fig. 4. In addition, electromagnetic flowmeters have been developed in which, to avoid the appearance of transformer emf, the wires from the collecting electrodes are brought out through an opening made in the core of the electromagnet, which consists of two symmetrical parts having identical coils [8].

Another way to eliminate the influence of the transformer emf on the measurement result is to use a rectangular current waveform to supply the sensor magnet. Such a current waveform [9] induces

a rectangular signal in the measuring circuit, proportional to the velocity of the liquid.

The action of the transformer e.m.f. is manifested in the appearance of sharp pulses at the moment when the polarity of the current feeding the magnet is changed. This difference in waveform makes it possible to separate the flow signal from the transformer e.m.f. more effectively. The sharp peaks caused by the transformer e.m.f. can be suppressed by a gating circuit. To do this, the amplifier is gated at the moment when the polarity of the current feeding the sensor magnet is changed. The e.m.f. is measured only in those time intervals when the field is maximal and constant. Thus, the time derivative of the magnetic flux, and correspondingly the signal produced by the transformer e.m.f., are equal to zero.

A flowmeter with a rectangular waveform of the supply current makes use of the advantages and eliminates the shortcomings of instruments of this type with a permanent magnet and with a magnet supplied by alternating sinusoidal current.

Suppression of the transformer e.m.f. can be accomplished by using a special phase-selective element. As such an element one may use a reversible capacitor motor, installed in automatic electronic potentiometers and bridges for balancing the circuit [10]. For automatic compensation of the signal at the input of the secondary instrument, the voltage induced by the magnetic field of the sensor is used. The compensating voltage taken from a rheochord and the output voltage of the sensor are applied in antiphase to the input of the amplifier. When the magnitude of the input signal changes, a certain quantity \(\pm \Delta u\) appears at the input of the amplifier; it is amplified and applied to the excitation winding of the motor. The motor shaft is thereby displaced through an angle \(\pm \Delta \varphi\), and, through a system of levers, moves the slider of the rheochord, changing the magnitude of the compensating voltage.

Owing to the presence of the phase-selective element, the secondary instrument does not respond to disturbances induced by the magnetic field of the sensor. However, when such a compensation method is used, its dimensions increase and the cost of the instrument rises.

At the Tomsk Polytechnic Institute, an original method of compensating the transformer e.m.f. was investigated. The method is based on multiplying the signal taken from the sensor by a voltage whose phase coincides with that of the useful signal.

For a sinusoidal magnetic field, the signal obtained from the flowmeter at the sensor electrodes consists of two components:

\[ e_0 = (e_1 \sin \omega t + e_2 \cos \omega t), \tag{1} \]

where \(e_1 \sin \omega t\) is the voltage in phase with the magnetic field, proportional to the flow velocity;

\(e_2 \cos \omega t\) is the interference voltage, independent of the flow and shifted in phase by \(90^\circ\) relative to the magnetic field.

To obtain a voltage proportional to the liquid flow, it is necessary to isolate the first term in equation (1) and suppress the second term. In practice this can be done with a multiplication circuit.

The multiplier in Fig. 5 consists of a resistance bridge using semiconductors as resistance elements. An increase in sensitivity is achieved by connecting four photoresistors in the bridge circuit. The photoresistors were illuminated by means of MN-8 neon lamps. To compensate the transformer e.m.f., the voltage taken from the electrodes must be multiplied by a term of constant amplitude whose phase coincides with the mag-

...with a magnetic field \(e_3 \sin \omega t\). In our case, the neon lamps are supplied by the voltage \(e_3 \sin \omega t\). The voltage taken from the electrodes is applied to the bridge diagonal \(ef\).

The multiplier operates as follows: in the absence of illumination, all the photoresistors are under identical conditions, and the potential difference in the bridge diagonal \(mn\) is zero.

Fig. 5. Multiplier circuit. \(\mathrm{FC}_1, \mathrm{FC}_2, \mathrm{FC}_3, \mathrm{FC}_4\) are photoresistors; \(L_1, L_2, L_3, L_4\) are lamps.

During the first half-period, lamps \(L_1, L_4\) are lit, introducing an imbalance into the operation of the bridge, while lamps \(L_2, L_3\) are extinguished. The magnitude of the resistances \(\mathrm{FC}_1\) and \(\mathrm{FC}_4\) changes according to a harmonic law from \(R_T\) to \(R_c\) and back to \(R_T\), i.e., the frequency of the resistance variation is equal to the frequency of variation of \(e_3 \sin \omega t\). The maximum imbalance of the bridge occurs when the phase of the voltage supplying the lamps coincides with the phase of the voltage applied to the bridge diagonal \(ef\). At the end of the first half-period all the lamps are extinguished and the bridge is again balanced.

During the second half-period, \(e_3 \sin \omega t\) lights lamps \(L_2, L_3\), and the pattern of resistance variation is repeated.

As a result of multiplication we obtain

\[ e_0' = (e_1 \sin \omega t + e_2 \cos \omega t)e_3 \sin \omega t = \tag{2} \]

\[ = \frac{1}{2}(e_1 e_3 - e_1 e_3 \cos 2\omega t + e_2 e_3 \sin 2\omega t). \]

By placing a filter at the output of the multiplication circuit, one can isolate the pure signal proportional to the flow rate,

\[ e = \frac{1}{2} e_1 e_3. \tag{3} \]

Studies carried out on this multiplication circuit using \(\mathrm{FC}\)-K1 with a magnet supply frequency of 50 Hz and using \(\mathrm{FC}\)-A1 with a supply frequency of 400 Hz made it possible to simplify considerably the circuit of the secondary instrument and to eliminate the effect of transformer emf on the measurement result.

References

1. W. G. James, An induction flowmeter design suitable for radioactive liquids, The Review of Scientific Instruments, v. 22, 1955.

2. W. G. James, An ac. induction flowmeter, Instruments, v. 25, 1952.

3. H. D. Einhorn, Electromagnetic induction in water. Trans. Roy. Soc. S. Africa, v. 28, 1941.

4. A. W. Denison, M. P. Spencer and H. D. Green, A square-wave electromagnetic flowmeter for application to intact blood vessels, Circ. Res., v. 3, 1955.

5. A. W. Denison, M. P. Spencer, Factors involved in intact vessel electromagnetic flow recording, Federation Proc., v. 15, 1956.

6. L. M. Korsunsky. Measurement Technology, No. 6, 1957.

7. A. Kolin, J. L. Weissberg, and L. Gerberg, Electromagnetic measurement of blood flow and Sphygmomanometry in the intact animal, Proc. Soc. Exper. Biol. Med., v. 47, 1941.

8. U.S. Patent, No. 2734380, 1956.

9. H. W. Shirer et al., Proceedings of the J. R. E., v. 47, No. 11, 1959.

10. L. M. Korsunsky and E. M. Madikyan. Measurement Technology, No. 8, 1961.