NEW INSTRUMENTS AND METHODS OF MEASUREMENT

ELECTROMECHANICAL SENSORS

L. A. Goncharskii

Experience with the successful use of mechanically controlled electrovacuum devices (mechanotrons) in scientific laboratories, industry, and clinical practice has shown that these devices are now already emerging as an independent branch of applied electronics.

The principal field of application of mechanotrons is varied experimental and applied apparatus intended for the investigation and control of mechanical processes in various areas of science and technology. Here mechanotrons are used chiefly as highly sensitive electromechanical sensors.

At present, several basic types of mechanotrons manufactured by the electrovacuum industry of a number of countries have already become established. At the same time, electrovacuum laboratories continue the development of numerous types of new mechanotrons intended for use in new systems of measuring, control, automatic, and recording apparatus.

For a number of problems it proves expedient to construct specialized types of electromechanical sensors, rather than to adapt existing types of sensors by simultaneously complicating the subsequent stages of the apparatus intended for solving specific problems. It is precisely this circumstance that can explain the fact that, even recently, despite the existence of a number of designs of highly sensitive electromechanical sensors, intensive development continues of new sensors of this type, based on the use of new phenomena that make it possible to carry out effective mechanical control of electron and ion currents in vacuum.

In addition to the development of new designs of mechanotrons, there is a tendency to extend the fields of their use beyond the already established area of experimental techniques for investigating mechanical processes and measuring the values of individual mechanical quantities; recently there have also appeared tendencies to use electromechanical sensors for the purpose of measuring and investigating nonmechanical quantities. An analysis of the prospects for using mechanotrons in these areas of experimental technique as well shows the expediency of developing specialized designs of these tubes in order to implement electronic apparatus, completely new in its principles of operation, that can be used in scientific laboratories, industry, and other branches of the national economy.

Finally, recently work has been carried out on studying the possibilities of using mechanotrons as transducing elements of regulating and amplifying devices. In these areas of electronic technology as well, definite prospects have already been indicated that deserve preliminary

analysis and, if only for a general assessment of the possibilities and prospects for implementing such apparatus.

The development of new simplified methods for the mechanical control of electron and ion currents, as well as of mechano­tron designs accessible for manufacture even by nonspecialized laboratories, makes it possible for many scientific and factory laboratories to produce mechanotrons independently. This permits electron-mechanical sensors to be introduced much more rapidly into the practice of physical experimentation and to be used more widely in industrial control-and-measuring apparatus.

The development of new systems of electron-mechanical sensors, distinguished by high stability and reliability of operation at high sensitivity, has recently attracted attention to the study of the possibilities of using these sensors in apparatus employed for the automation of production processes. Electronic sensors have also found application in the monitoring equipment of automatic lines.

I. MECHANOTRONS

1. Longitudinal-control tubes

The longitudinal method of mechanical control of electron currents was the first to be successfully used for the implementation of reliably operating electron-mechanical sensors, which found wide application in experimental technique and applied apparatus. On the basis of this method, mechanotrons were produced by the electrovacuum industry of a number of countries for more than a decade3.

Simultaneously with the expansion of the fields of application of longitudinal-control mechanotrons, work was carried out to study the physical features of the longitudinal method of mechanical control of electron currents, with the aim of radically improving the parameters of such tubes, as required for their use in new areas of experimental technique where they had not previously been applied.

A new representative of longitudinal-control mechanotrons is the tube of bilateral conductivity. It has two flat indirectly heated cathodes 1 and 2 (Fig. 1, a), of which the first is fixed, while the second is mounted on the movable rod 3. The latter can oscillate in the elastic membrane 4 in the direction shown by the arrow[^2]. This tube has the same conductivity in both directions and can be used for regulating alternating-current circuits, as well as as a sensor operating in alternating-current circuits. Below we shall dwell on devices of this type, in which a rheostat of adjustable resistance is used, the values of which remain the same for currents of both directions.

Fig. 1.

Of practical interest are also mechanotrons of bilateral conductivity with different resistances for currents in opposite directions and with an adjustable ratio of these resistances. These are double diod—

tubes, the currents in each of the halves of which have opposite directions. A schematic diagram of one variant of such a tube is shown in Fig. 1, b. Here, two heated cathodes \(K\) are mounted on a common heater \(P\), parallel to which the anodes \(A\) are arranged. When the rod \(C\), on which the anodes are mounted, is swung in the direction shown by the arrow, the resistance of one half of the tube increases with a corresponding decrease in the resistance of its other half. Such tubes can be used as sensors operating on alternating current, but delivering to the load (a readout device or recorder) a pulsating current whose direction and mean value are determined by the displacement of the moving system of the electron tube. It is expedient to use such tubes also as relays delivering a current whose direction is determined by the direction of deflection of the moving rod. Similar relays may be used for switching the actuating circuits of automatically controlled systems, for example, in automatic potentiometers.

For parametric sensors, which also include electromechanical sensors, it is important for stable operation to choose a sufficiently low-stress temperature regime. This is due to the fact that, as the power dissipated inside the sensor increases, temperature errors rapidly increase, and with them the zero drift of the measuring device. It should be noted that certain types of continuous-control sensors, for example sensors manufactured by the firms Sylvania (USA), Mullard (England), and Philips (Holland), are distinguished by a rather severe thermal regime, consuming about 5 watts\(^3,4\). This is due not so much to the large power dissipated at the sensor anodes as to the increased power expended on heating the incandescent cathode.

The ratio of the powers dissipated at the anodes and in the cathodes of a mechanotron is characterized by the power-ratio coefficient

\[ K = \frac{P_a}{P_n}, \tag{1} \]

where \(P_a\) is the power dissipated at the anode of the tube, and \(P_n\) is the power consumed for heating the cathode. Substituting here the values

\[ P_a = I_a \cdot U_a \]

and

\[ P_n = I_a \cdot Y, \]

where \(U_a\) is the voltage at the anode of the tube, \(I_a\) is the anode current of the tube, and \(Y\) is the specific heating power \(\frac{P_n}{I_a}\) in watts per ampere, we obtain for \(K\) the expression

\[ K = \frac{U_a}{Y}. \tag{2} \]

In view of the circumstance that the limiting power that can be dissipated inside the sensor is restricted by its limiting thermal regime, determined by the conditions of stable operation, it proves essential to increase the value of the power-ratio coefficient. Increasing the latter makes it possible to raise the sensitivity of the sensor at an unchanged value of the power dissipated inside the sensor. Consequently, in order to increase the sensitivity of a sensor for which the value of the power dissipated in it is fixed, it is expedient to increase the anode voltage while simultaneously reducing the specific heating power.

The approximate values of the specific heating powers and the power-ratio coefficients for a number of types of modern cathodes of electron tubes, characterized by low work function, are given in Table I.\(^{20,21}\)

Table I

When the data of Table I are compared with the parameters of sensors, the reasons become obvious why the electronic longitudinal-control sensors manufactured earlier had values of the power coefficient considerably less than unity. These were predominantly diode sensors operating at anode voltages of about 10 V or somewhat higher.

From relation (2) it is evident that, under such conditions, the values of the power-ratio coefficient turn out to be considerably less than unity for all types of cathodes previously used in electromechanical sensors. Only the use of the metal-ceramic tubular cathode,\(^{21}\) developed comparatively recently, makes it possible to realize electronic diode-type sensors for which the value of the power-ratio coefficient exceeds unity.

In electromechanical sensors having oxide cathodes with a solid core, the only way to increase the value of the power-ratio coefficient without reducing the service life is to raise the anode voltage and to choose the cathode most suitable in its geometry. In particular, in the mechanotron manufactured by the firm “Radiocorporation” under the name vibrotron,\(^{15}\) the use of a heated cathode of an unfavorable shape led to a value of the power-ratio coefficient less than unity even at a relatively high anode voltage equal to 300 V.

From the data of Table I it is evident that replacing indirectly heated oxide cathodes with a solid core by directly heated oxide cathodes promotes a considerable increase in the values of the power-ratio coefficient. This gives grounds for recommending directly heated cathodes for triodes with a movable anode for longitudinal control. In particular, Fig. 1, в shows the circuit of an electronic sensor of this type satisfying the above condition. Here 1 is a fixed directly heated oxide cathode, and 2 is a fixed flat grid with very small apertures. At a very small distance from the grid there is a movable flat anode 3, oriented parallel to the plane of the grid. Anode 3 moves in the direction indicated by the arrow.

Figure 1, г shows the circuit of a dynamic-pressure sensor—the piezotron, whose electrode system corresponds to that shown in Fig. 1, в (the designations are analogous). The operation of this sensor is based on the deflection of membrane 3 as the external pressure increases.

In diode longitudinal-control sensors, attempts were also made to use a directly heated cathode.\(^{13,14}\) However, sensors of this type

of this type were distinguished by a very low voltage sensitivity. In the sensors of Hahn1, and also of Grekhova and Vasil’eva2, the main potential drop of the electric field occurred at the cathode surface, whereas the electric-field gradient at the surface of the movable anode had too small a value. Since the voltage sensitivity of a longitudinal-control sensor3 proves numerically equal to the electric-field strength at the surface of the plane anode, the reason for the low sensitivity of such sensors becomes obvious.

To increase the voltage sensitivity of a longitudinal-control electron sensor having a directly heated cathode, it proves necessary artificially to reduce the electric-field strength at the cathode surface and thereby increase the electric-field strength at the surface of the movable anode. One possible solution of this problem is the triode device described above, with a fixed grid located between the hot cathode and the movable anode. Another solution reduces to winding the directly heated cathode on a flat insulating plate.

The structural scheme of such a mechanotron corresponds to a diode with flat parallel electrodes, with the difference that, instead of a flat indirectly heated cathode, an insulating plate of analogous shape is placed, on which the directly heated cathode is wound. In this case the presence of an insulating surface near the directly heated cathode reduces the value of the electric-field gradient at its surface.

2. Probe-control tubes

It has already been shown above that, in longitudinal-control mechanotrons, replacing oxide cathodes with a continuous core and indirect heating by directly heated oxide cathodes contributes to a considerable increase in the values of the power-ratio coefficient. However, when directly heated cathodes are used in mechanotrons, it proves (in many cases) still more expedient to pass from longitudinal to probe control of the electron currents. Such a transition4 makes it possible to increase the voltage sensitivity of a number of electromechanical sensors by several orders of magnitude.

The voltage sensitivity of a probe-control mechanotron with a uniform field can be calculated from the formula4

\[ \varphi_{\mathrm{d}} = -\frac{U_a}{a}, \tag{3} \]

where \(U_a\) is the voltage at the anode of the tube, and \(a\) is the distance between the cold cathode and the hot cathode. Taking \(U_a = 500\ \mathrm{V}\) and \(a = 0.1\ \mathrm{mm}\), we obtain a voltage sensitivity on the order of \(50\,000\ \mathrm{V/cm}\). By using a mesh anode and placing behind it a collector4 of electron current, one can achieve a further considerable increase in the voltage sensitivity of the probe-control tube.

The relatively small power dissipated in a probe-control mechanotron with high voltage sensitivity makes it possible to reduce substantially the external surface of the metallic electrodes; in particular, it proves possible to make the electrodes from wire. This simplifies the technology of fabrication and degassing of the tube electrodes.

The arrangement scheme of the electrodes of a tube of this type is shown in Fig. 2, a. Here \(K\) is the directly heated cathode, \(A\) is the anode, made of

wire; \(X\) is a cold cathode, also made of wire, whose cross section is considerably smaller than the cross section of the anode. The moving electrode is the cold cathode, which can move in the direction indicated by the arrow.

A double version of this type of tube is shown in Fig. 2, b. The stationary hot cathode \(K\) with direct heating has a U-shaped form. On the outer side of the cathode there are two anodes \(A\), consisting of straight wires. The cone-shaped cold cathode \(X\) is placed between the two branches of the hot cathode.

Fig. 2.

A similar construction of a tube with probe control represents, to a certain extent, a development of the principles underlying the construction of the vibrotron,¹⁵ developed by Rose. According to Olson's proposal, in the vibrotron an anode having the form of a cone was used. From the mechanical point of view such a form of the movable electrode proved very favorable, since it made it possible to increase considerably the resonant frequency of the kinematic system of the tube (up to 12,000 cps). However, it had an adverse effect on the possibility of obtaining effective longitudinal mechanical control of the electron current of the tube.

On the contrary, when probe control is used, the concentration of the electric field at the end of the cone proves, in contrast to the longitudinal method of control, very favorable for obtaining high voltage sensitivity of the mechanotron.

In Fig. 2, c a version is presented of the construction of a tube with probe control, with a collector \(C\) for the anode current of the tube, located behind an anode \(A\) made of a straight wire of relatively small diameter. The movable cold cathode \(X\) controls the electron current emitted by the hot cathode \(K\). The greater part of the electrons bends around the thin rod of the anode and reaches the collector \(C\). The voltage sensitivity and internal resistance of such a mechanotron prove to be considerably greater than the corresponding parameters of an analogous triode mechanotron with probe control.

To obtain high current sensitivity in a mechanotron with probe control, the distance between the hot cathode and the anode should be reduced. However, in such a regime, at appreciable anode currents, the proportionality is disturbed between the gradient of the electric field near the probe surface and the potential difference of the probe and the equipotential surface at the level of which the probe is located.¹¹ In the case where the probe is in the immediate vicinity of the anode, the electron tube operates, to a certain degree, analogously to a diode with flat anodes and a directly heated cathode. This gives grounds for using Kuzunose’s empirical formula²² to obtain an approximate expression that makes it possible to estimate the current sensitivity of the tube to displacements of the movable electrode. When thin cathodes are used, the diameter of which is sufficiently small in comparison with the distance between the probe and the anode, the structure of the electric field near the cathode is obtained analogous to the structure of the electric field in a diode with a cathode of the same diameter, located at a distance from the anode equal to the distance of the probe from the anode of the tube (under the condition that the anode of the diode has a potential close to the potential of the equipotential surface at the level of which the probe is located).

The Kuzunose formula for the anode current of a kenotron of the indicated type has the form

\[ I_a = 2.33 \cdot 10^{-6}\,\frac{S_a \cdot U_a^{1.5}}{\beta^2 \cdot a^2}, \tag{4} \]

where \(S_a\) is the effective surface of the anode, \(U_a\) is the anode voltage, \(\beta\) is a coefficient whose value is determined by the ratio of the cathode diameter to the distance between the cathode and the anode (for small-diameter cathodes used in probe-control tubes, \(\beta\) may be taken equal to unity), and \(a\) is the distance between the hot cathode and the anode.

Substituting into this expression

\[ S = 2 \cdot L \cdot a, \tag{5} \]

where \(L\) is the total length of the active surface of the directly heated cathode, we obtain an expression for the anode current of a probe-control tube with a uniform field in the form

\[ I_a = 4.66 \cdot 10^{-6}\,\frac{L \cdot U_a^{1.5}\cdot (l-a)^{1.5}}{l^{1.5}\cdot a}. \tag{6} \]

Substituting (6) into the expression for the differential current sensitivity

\[ \psi_d = \left(\frac{\partial I_a}{\partial a}\right)_{dU_a = 0}, \tag{7} \]

we obtain

\[ \psi_d = -\,4.66 \cdot 10^{-6}\, \frac{L \cdot U_a^{1.5}\cdot (l-a)^{0.5}\cdot (l+0.5a)} {l^{1.5}a^2}. \tag{8} \]

Since \(a < l\), the approximate relation obtained may be transformed to the form

\[ \psi_d = -\,4.66 \cdot 10^{-6}\,\frac{L \cdot U_a^{1.5}}{a^2}. \tag{9} \]

Comparing (6) and (8), we obtain the approximate expression

\[ \psi_d = \frac{I_a}{a}. \tag{10} \]

From relation (10) it is evident that, at a fixed distance between the hot cathode and the anode, the sensitivity of a probe-control sensor proves to be proportional to its anode current. With a fixed cathode diameter, an increase in anode current is accompanied by an increase in cathode length. This, in turn, makes it possible to pass from tube designs with a straight cathode to designs of probe-control tubes with a denser arrangement of the cathode. In particular, it is reasonable here to use fastening of a directly heated cathode in a zigzag, as is done, for example, in the 5TsZS kenotron.

It is also expedient to form the directly heated cathode of a probe-control mechanotron in the form of a U-shaped loop, fastened by its ends to the movable rod of the tube. The plane of the loop is arranged parallel to the surfaces of the anode and the cold cathode.

In Fig. 2, a is shown the circuit of a dual probe-control sensor in which a directly heated cathode is used, wound in the form of

spiral of circular or rectangular form. On both sides of the directly heated spiral cathode \(K\) are located the anodes \(A\). Inside the spiral is a cold cathode \(X\), consisting of a straight thin rod whose axis coincides with the axis of the spiral. When the cold cathode is displaced in the direction indicated by the arrow, the ratio of the anode currents in the two halves of the double transducer changes. In electronic acceleration transducers it is possible to use a free spiral, sagging under the action of the inertial force, as the movable inertial mass of the transducer. In some cases it is advisable to make the cathode in the form of a spiral wound on a rigid frame.

The cold cathode may also be made of an insulator. The potential acquired by the surface of the insulator situated in the immediate vicinity of the hot cathode proves to be close to the potential of the corresponding sections of the cathode, but somewhat lower than it because of the temperature difference.

3. Ion glow-discharge tubes

The operation of ion glow-discharge transducers is based on two fundamental principles: on the geometrical restriction of the glow discharge and on the use of probes moving inside the discharge gap. With geometrical restriction of the discharge gap, the work required to carry an electric charge through it increases and, consequently, so does the voltage drop across the discharge gap at an unchanged discharge current. At present several methods of geometrical restriction of the discharge are known which make it possible to obtain a considerable sensitivity of the mechanotron.

The best known of these is the longitudinal method of mechanical control of a constricted discharge3. The schematic diagram of such a transducer is shown in Fig. 3, a. Inside the ion tube there are two electrodes—fixed electrode \(1\) and electrode \(2\), which can move in the direction indicated by the arrow. The surfaces of the two electrodes facing one another are flat and parallel. The gas pressure (usually of an inert gas) inside the tube is set so as to obtain a constricted glow discharge between the electrodes. To prevent the discharge from enveloping the side surface of the cathode, the latter is surrounded by a protective glass tube. Under these conditions it is possible to obtain a high sensitivity of the ion tube, with respect to voltage, to displacement of the movable electrode.

Fig. 3.

Figure 3, b shows the schematic diagram of a double transducer of this type. The movable flat anode \(3\), fastened to rod \(4\), sealed into an elastic membrane, can move in the direction indicated by the arrow, changing the voltage drop across both discharge gaps.

The magnitude of the internal differential resistance of ion transducers of this type, reaching \(100\text{–}200\ \text{k}\Omega\), is noteworthy.

The schematic diagram of an ion transducer with transverse control is shown in Fig. 3, c. Here the glow discharge between electrodes \(1\) and \(2\) passes through slot \(3\), one wall of which is fixed, while the other is formed by movable plate \(4\), fastened to rod \(5\), which passes through an elastic membrane.

...to the target. The narrowing of the target in which the discharge takes place promotes recombination of ions on the walls of the target and thereby increases the voltage drop across the device. An approximate form of the characteristic of the dependence of the voltage drop across the device on the position of the movable electrode is shown in Fig. 3, d. Sometimes electrode 1 is placed inside a hollow cylindrical electrode 2.

Here it should be noted that it is expedient to use transverse control of the heated cathode in ionic sensors. This makes it possible to reduce substantially the operating voltage on the sensor. In this case the basic circuit of the tube remains analogous to that shown in Fig. 3, c, with the difference that the cold cathode 1 is replaced by a hot cathode.

In Fig. 4, a is shown the circuit of a probe mechanotron of a glow discharge of direct current. It consists of a bulb filled with a rarefied inert gas, inside which there are two flat parallel electrodes 1 and 2.

Fig. 4.

Between them (and parallel to them) is placed electrode 3, which is a thin straight wire fastened to rod 4, which can move (owing to an elastic joint) in the direction indicated by the arrow. The potential of electrode 3, which plays the role of a probe, is in a definite correspondence with the potential of the discharge zone in which it is placed. During operation of the sensor a glow discharge is maintained between electrodes 1 and 2, and displacement of the probe is accompanied by a change in its potential relative to electrode 1. Consequently, the potential of the probe determines the position of the movable electrode.

In Fig. 4, b is shown the circuit of a double sensor of probe control, in which the discharge occurs between cathode 1 and anodes 2. On both sides of cathode 1 there are two thin straight probes 3. Cathode 1 is fastened to movable rod 4. Displacements of the latter in the direction indicated by the arrow are accompanied by corresponding changes of the current in the galvanometer connected to the probes[^12].

Recently the Decker Aviation Company in the USA has begun mass production of ionic sensors of probe control[^23], operating in a high-frequency electric field. The action of this sensor is based on the phenomenon of the occurrence of a space charge in a rarefied gas under a high-frequency discharge. The circuit of a measuring device with a high-frequency ionic sensor of probe control is shown in Fig. 4, c. The sensor itself is a tube 1 filled with rarefied gas, into which two straight electrodes 2 are sealed. The sensor is placed in the electric field of capacitor 3, supplied from high-frequency generator 4. When sensor 1 is displaced in the direction indicated by the arrow, the magnitude of the direct voltage taken from electrodes 2 changes. The maximum displacement of the sensor

is accompanied by a change in the voltage taken off in a range of 200 V. It is essential to note here the linear character of the dependence of the voltage taken off on the displacement of the sensor. The voltage sensitivity of the sensor proves to be about 20,000 V/cm. The diameter of the glass tube is about 6 mm, and the length about 18 mm. The long service life of the sensors—more than 5000 hours—is noteworthy; it is due, to a considerable extent, to the fact that the measuring device is very economical with respect to the power consumed: under normal operation the sensor consumes less than 0.05 W.

4. Pulsed-discharge ion lamps

The operation of pulsed-discharge ion lamps is based on regulating the discharge ignition voltage by displacing a movable electrode. Ion lamps of this type may have either a cold or a hot cathode.

The operation of lamps of the first of the above-mentioned types is based on regulating the discharge ignition voltage as a result of a change in the geometry of the discharge gap. Ion lamps may be either of longitudinal control (their structural design basically corresponds to the variants shown in Fig. 3, a, b) or of transverse control (basically corresponding to the circuit shown in Fig. 3, v).

It should also be noted here that in pulsed lamps with a cold cathode it is expedient to use a separate, third electrode, which is a movable electrode controlling the ignition of the discharge in the lamp. A lamp of this type is a mechanically controlled thyratron with a cold cathode. In this thyratron the control electrode is the movable electrode of a mechanotron. In single closing operations in such lamps, certain fluctuations of the voltage required for ignition of the discharge are permissible. However, when such sensors are used in circuits of relaxation oscillators and with pulsed direct-current supply, fluctuations of the ignition voltage are smoothed out, and the value of the mean discharge ignition voltage can already characterize fairly closely the position of the movable element.

Mechanically controlled thyratrons also belong to pulsed-discharge mechanotrons. A mechanically controlled thyratron has one movable electrode, whose displacement is used for mechanical regulation of the control coefficient of the thyratron. The circuit of a mechanically controlled thyratron essentially does not differ from that of any mechanically controlled electronic tube having a hot cathode and one movable electrode. By analogy with mechanically controlled electronic tubes, mechanically controlled thyratrons may be of longitudinal, probe, and differential control^5. Mechanically controlled thyratrons can operate both on direct and on alternating current. Later we shall touch upon certain features of both modes of operation of mechanically controlled thyratrons.

Mechanically controlled thyratrons may have external control by means of a kinematic coupling through an elastic element of the wall of the envelope of an ion lamp. It is also possible to realize mechanically controlled thyratrons with internal control. In particular, in thyratron acceleration sensors the movable electrode itself may be used as the inertial mass. At fixed values of the voltages on the thyratron electrodes, the discharge within it is ignited at a definite value of the acceleration with which the body on which the sensor is mounted moves.

In another possible variant of a thyratron with internal mechanical control, the movable electrode is fastened to a freely suspended magnet. Under the action of an external magnetic field, the latter rotates, carrying along ...

constituting an electrode. Such a tyratron with magnetic control may be used either to control a specified intensity of the magnetic field, or to control a specified orientation of the tyratron relative to the magnetic field.

II. MEASURERS OF MECHANICAL QUANTITIES

1. Micrometers

The high sensitivity of mechanotrons to the position of the movable electrode facilitated the use of these tubes primarily as highly sensitive transducers for electronic micrometers12. In the latter, bridge circuits operating on direct current, with a galvanometer (pointer type) as the indicating device, are used predominantly; moreover, the sensitivity of the micrometer reaches \(10^{-6}\) and even \(10^{-7}\) cm and is limited not by the electronic elements of the device, but by the stability of operation of its mechanical elements and by thermal deformations of its individual parts. It is appropriate to note here that such high sensitivity of the micrometer to small displacements proves especially important for detecting small periodic changes in linear dimensions or displacements, where continuous displacement of the zero of the measuring device is no longer dangerous, provided that the sensitivity of the device to small displacements remains unchanged.

In connection with the nonlinearity of the characteristics of most electronic transducers used in micrometers, in the latter one usually resorts to pairing transducers having identical characteristics. In this case, voltage fluctuations in the supply circuit cause relatively small shifts of the zero. It should be noted here that the dependence of the sensitivity of most electromechanical transducers on the applied voltage requires stabilization of the latter, primarily for feeding the anode circuits.

Fig. 5.

The circuit for connecting a high-frequency transducer to a galvanometer of high internal resistance is shown in Fig. 4, c. In Fig. 5, a is shown a bridge circuit for connecting an ionic transducer 1 of longitudinal control with a hindered glow discharge. In the same circuit is shown the connection of a gas-discharge voltage stabilizer 2, used for powering the micrometer.

Among the advantages of gas-discharge transducers is the possibility of performing measurements immediately after switching on the device. In this they compare favorably with electronic transducers, which require a certain interval of time for heating the filament and establishing a stationary thermal regime. The period required to establish a stationary thermal regime proves, for transducers with a heated cathode, to be considerably longer than the period required to establish a stationary thermal regime of a transducer with a directly heated cathode.

Thanks to the possibility of switching ion sensors on only for the duration of the measurement, it becomes possible, with their aid, to implement a measuring device for periodic readout of the monitored dimension over a long period of time while powering it from a dry battery. In the circuit shown in Fig. 5, a, there is a switch 3, which permits periodic, short-term connection of the battery 4 supplying the circuit.

The use of a logometer for reading the indications of an electronic micrometer makes it possible to use unstabilized sources for its power supply. Figure 5, b shows the circuit for connecting a logometer into the circuit of an electronic micrometer.

Electro-mechanical sensors with a hot cathode have unidirectional conductivity. This makes it possible to use an alternating-current mains supply for their power, while at the output of the sensor obtaining a rectified signal suitable for feeding to highly sensitive magnetoelectric indicating and recording devices.

Figure 5, v shows the circuit of an electronic micrometer operating on alternating current and equipped, for reading the indications, with a sensitive direct-current galvanometer. When a magnetic stabilizer of the supply voltage is used in such a device, one can obtain a simple and convenient micrometer operating from an alternating-current mains supply.

2. Vibrotrons

Initially, the name vibrotron was given to the RCA-5734 electron tube, so called because it was used chiefly for studying various oscillatory processes3. Later, electro-mechanical sensors used in vibration-measuring apparatus came to be called by this name.

For the study of vibrations, electro-mechanical sensors of displacement, velocity, and acceleration are used. Of these, the greatest practical interest at present is represented by the first and the last. Electronic displacement sensors for oscillatory processes—seismotrons—are of two types: with external and internal control.

Fig. 6.

The basic circuit of a seismotron with external control is shown in Fig. 6, a. Inside the frame 1, an inertial mass 3 is suspended on elastic springs 2; it is kinematically connected with the electronic sensor of small displacements 4. If the frequency of the oscillations is considerably higher than the natural frequency of the oscillatory system formed by the mass 3 and the spring suspension 2, then the mass 3 remains practically motionless, while the housing of the vibrotron 1 oscillates relative to this mass in the direction indicated by the arrow.

In such vibrotrons, standard electronic sensors of small displacements are used, distinguished by a comparatively small working travel of the movable electrode. The choice of the length of the lever extending the rod of the movable ...

...electrode is determined by the range of displacements for which the measuring device is designed, so as to obtain the maximum amplitude of the sensor signals and not risk overloading it.

Displacement sensors with internal control are considerably simpler in their construction. In Fig. 6, b a diagram of such a seismotron is given. Inside the electronic tube 1, on an elastic suspension 2, an inertial mass 3 is fixed, forming the movable electrode of the tube. The resonant frequency of the kinematic system of the seismotron, in this variant as well, must be lower than the lowest component of the range of frequencies under investigation. When the tube mounted on the object under study oscillates, the elastically suspended electrode remains at rest. Therefore the position of this electrode relative to the other electrodes of the tube changes periodically in accordance with the oscillations of the object. As a result, signals appear in the anode circuit of the tube which characterize the controlled oscillations. In Fig. 6, c a diagram is shown of an electron-mechanical level, which can be used as a seismotron for horizontal oscillations.

The signals generated by seismotrons of the types described above depend, in practice, little on the frequency of the oscillations (if the latter is considerably higher than the resonant frequency of the kinematic system of the sensor), and their amplitudes prove proportional to the displacements of the controlled object.

Electronic acceleration sensors—accelerotrons—have kinematic systems analogous in their construction to those shown in Fig. 6, a, b, and c. The difference between the kinematic systems of seismotrons and accelerotrons consists in the choice of resonant frequencies. The resonant frequency of the kinematic system of an accelerotron must be considerably higher than the highest-frequency component of the mechanical process under investigation. In this case the displacement of the inertial mass of the sensor proves proportional to the acceleration with which the sensor moves. The sensitivity of the accelerotron is inversely proportional to the resonant frequency of its kinematic system. Therefore the natural frequency of the kinematic system should not be set much higher than is necessary for satisfactory reproduction of the accelerations of the range under investigation.

Displacement and acceleration sensors with external control are ordinary electron-mechanical sensors of small displacements, coupled by means of a kinematic system with an inertial mass. The operating characteristics of such a device depend to a considerable degree on the structural design of the kinematic system, which does not present substantial difficulties and is, to a significant extent, analogous to seismic sensors and acceleration sensors with other systems of electrical transducers.

Therefore we shall consider only certain most important features of the construction of electronic displacement and acceleration sensors with internal control. The best-known sensors of this type are tubes manufactured by the firms Sylvania³, Mullard²⁴, and Philips²⁵. They are diodes of longitudinal control with movable anodes and are distinguished by high current sensitivity and a small value of the operating voltage on the anodes. Signals from such sensors are usually recorded directly by means of an electromagnetic oscillograph, without preliminary amplification. The essential defects of such tubes include the excessively large value of the power expended on heating the cathode of the tube.

We shall describe several designs of electronic acceleration sensors distinguished by a considerably smaller thermal load on the tube cathode.

Among these, first of all, should be included electronic sensors with probe control. In Fig. 6, d a diagram is given of the simplest, in its construction, electronic acceleration sensor with probe control and a sagging cathode. It consists of an anode \(A\) and a cold cathode \(K\), between which

there is a filamentary directly heated cathode \(K\). Under accelerated motion of the transducer in the direction indicated by the arrow, cathode \(K\) sags under the action of the inertial force, assuming the position shown by the dotted line. As a material for making the sagging cathode it proves expedient to use oxidized tungsten, which has high strength, quite sufficient for measuring considerable accelerations.

The sagging cathode may also have the form of a U-shaped loop, the plane of which is arranged parallel to the planes of the anode and the cold cathode. The inertial force applied to the loop, under accelerated motion of the transducer, bends the loop and changes its distance from the anode and the cold cathode.

When an electronic transducer of probe control is used for measuring relatively small accelerations, it proves expedient to employ the cold cathode as the movable electrode. The circuit of such a transducer is shown in Fig. 6, \(d\). Here the anode \(A\) and the hot cathode \(K\), stretched rather tightly, are the fixed electrodes of the transducer, while the cold cathode \(X\), fastened to the elastic rod \(C\), is the inertial mass of the acceleration transducer.

In addition to vacuum accelerometers, it also proves possible to use gas-discharge acceleration transducers of the glow-discharge type \(^{27}\). Among the essential advantages of transducers of this type one should note the absence of a heated cathode, which contributes to a considerable increase in the reliability of the transducer. The absence of a heated cathode also makes it possible to use the transducer for measurements immediately after switching on the installation, without waiting for the establishment of thermal equilibrium, which is obligatory for transducers having a hot cathode. The circuit of an ionic acceleration transducer in which longitudinal control by a constricted glow discharge is used is shown in Fig. 6, \(e\). Here, between two cathodes \(K\), the lateral surfaces of which are covered by glass tubes \(T\), there is a flat movable anode \(A\), fastened to an elastic plate \(P\). Accelerated motion of the transducer in the direction indicated by the arrow is accompanied by a corresponding change in the gaps between the electrodes, in which the voltage drop in one discharge gap increases, while in the second discharge gap it correspondingly decreases. A double transducer of this type is intended for operation in a bridge circuit. It is essential that the end of the protective tube \(T\) protrude beyond the edge of cathode \(K\). Such a geometry promotes better concentration of the constricted discharge inside the discharge gap. In addition, the protruding end of tube \(T\) proves to be a convenient limiter of the displacements of the movable anode, preventing short-circuiting of the electrodes under excessive overloads of the transducer.

One of the most important features of vibrotrons is the small magnitude of the mechanical forces required to displace the electrode, due to the special features of the mechanical control of electronic and ionic currents. This makes it possible to use in transducers of this type inertial masses of very small magnitude. And this, in turn, makes it possible to considerably facilitate the damping of the natural oscillations of the kinematic system of the transducer; as is known, the damping of the movable electrode of a transducer with an internal inertial mass presents particular difficulties, and it is easier to carry out damping of the internal electrode in gas-discharge transducers, in which it is possible to use the internal friction in the gas filling the transducer.

Among the methods for damping electromechanical acceleration transducers with an internal inertial mass, the most highly developed is the method \(^{18}\) based on the use of a kinematic system with two degrees of freedom. The essence of this method reduces to the elastic mounting of the acceleration transducer, on which, in turn, an elastic kinematic system with a resonant frequency equal to the resonant frequency of the movab-

of the sensor electrode, but having much greater damping. The presence of kinematic coupling facilitates the transfer of the energy of the oscillatory motion from the internal oscillating body to the external one, where it is expended in overcoming the force of friction. Experimental investigation of this method of damping oscillations\(^{18}\) confirmed its effectiveness.

We have already noted above that there is no fundamental difference between a seismotron and an accelerotron. Sensors of both types differ only in the ratio of the controlled range of oscillations and the resonant frequency of the kinematic system of the sensor. Thus, for example, a portable electronic sensor of low-frequency accelerations, when mounted on an oscillating rod, proves to be a good receiver of its longitudinal sonic and ultrasonic vibrations, i.e. it operates as a seismotron. It should be pointed out that the use of a seismotron as a sensor makes it possible to simplify substantially the technical and scientific apparatus used for recording oscillatory processes, by reducing or completely eliminating amplifiers.

3. Force Meters

Instruments of this type include pressure meters (electronic manometers or piezotrons), electronic dynamometers, and electronic strain gauges.

Electronic manometers may be of external and internal control. Instruments of the first type include chiefly membrane-type manometers (Fig. 7, a), in which the deformation of the membrane \(M\) under the action of the pressure of a gas or liquid is measured by means of an electromechanical sensor \(D\) of small displacements\(^{9}\). Instead of an elastic membrane, Bourdon tubes or Vidie capsules may also be used as pressure receivers. In view of the obviousness of such a replacement, these variants of manometers are not shown in the figure.

Fig. 7.

Electronic manometers of internal control include sensors having an elastic element of the wall of the bulb, deforming under the action of the controlled pressure. In Fig. 1, g a diagram is given of a piezotron, part of whose bulb wall is made in the form of an elastic membrane that bends under external pressure. Electromechanical pressure sensors of this type are suitable for investigations of dynamic pressures. In particular, the use\(^{28}\) of such sensors for recording and visual observation of the course of pressure inside the cylinders of internal-combustion engines proves expedient.

In Fig. 7, b a diagram is shown of another design variant of an electromechanical pressure sensor of internal control, in which the element sensing the measured pressures is a flattened curved glass tube \(T\), sealed into the glass wall of the bulb of an electronic tube. The end \(E\) of this glass tube is located in the immediate vicinity of the hot cathode \(K\) of a probe-control tube (\(A\) is the anode of this tube) and performs the functions of a movable cold cathode. When the pressure of the medium in which the sensor is located changes, deformation of the glass tube \(T\) occurs and, correspondingly, the distance changes between the electrodes \(K\) and \(A\), on the one hand, and the end \(E\) of the glass tube, on the other. As a result, the anode current of the sensor becomes a function of the controlled pressure.

The last two types of electromechanical manometers are suitable for measuring comparatively high pressures, whereas a membrane-type manometer with an electromechanical sensor of the membrane-deflection arrow can be used both for measuring high and comparatively low pressures. Manometers of this type can be applied to the measurement of pressures beginning at \(10^{-2}\) mm of mercury and up to very high pressures.

It is worth mentioning the possibility of using a membrane manometer also for measuring lower pressures, provided that a system analogous to the McLeod manometer is used, in which the moving mercury compresses the control volume of gas into a narrow gap formed on one side by a flat fixed wall and on the other side by the membrane of a sensitive electromechanical manometer.

The use of electromechanical sensors with external control as dynamometers is based on measuring deformations caused by applied forces. The high sensitivity of electromechanical sensors makes it possible to use dynamometer kinematic systems distinguished by high rigidity, very small elastic aftereffects, and residual deformations. This makes it possible to implement dynamometric devices having high natural frequencies and suitable for the investigation of rapidly varying dynamic processes.

Especially interesting is the use of electromechanical sensors for strain-gauge investigations. The signals generated by them prove to be so intense that in many cases they can be connected directly to standard electromagnetic and electronic oscillographs for the purpose of recording dynamic pressures.

Let us give several relations illustrating the immediate possibility of this. Defining the current sensitivity \(\iota_d\) of the strain gauge as the partial derivative of the current in it with respect to the acting mechanical stress \(\sigma\), at constant voltage on the sensor anode,

\[ \iota_d=\left(\frac{\partial I_a}{\partial \sigma}\right)_{\text{with } dU_a=0} \tag{11} \]

and substituting here the value

\[ \sigma=\frac{E\cdot \Delta L}{L}, \tag{12} \]

where \(L\) is the base of the strain gauge, \(\Delta L\) is its elongation, and \(E\) is the modulus of elasticity of the material being tested, we obtain

\[ \iota_d=\frac{L\cdot \psi_d}{E}, \tag{13} \]

where \(\psi_d\) is the current sensitivity of the sensor to small displacements.

Defining the voltage sensitivity \(v_d\) of the strain gauge as the partial derivative of the voltage on the anode of the mechanotron with respect to the acting mechanical stress,

\[ v_d=\left(\frac{\partial U_a}{\partial \sigma}\right)_{dI_a=0} \tag{14} \]

and substituting (12) here, we obtain

\[ \varphi_d=\frac{L\cdot \varphi_d}{E}, \tag{15} \]

where \(\varphi_d\) is the voltage sensitivity of the sensor to small displacements.

Taking, for example, the values \(L=2\) cm, \(\psi_d=1\) A/cm, \(\varphi_d=10\,000\) V/cm, \(E=20\,000\) kg/mm\(^2\), we obtain, respectively, \(\iota_d=100\) μA\(\cdot\)mm\(^2\)/kg, \(v_d=1\) V\(\cdot\)mm\(^2\)/kg.

It is not difficult to see from the data presented that the obtained values for electron-mechanical strain gauges exceed by hundreds of times the sensitivities of wire resistance strain gauges having the same bases.

By using small-sized electron-mechanical transducers, it is possible to construct highly sensitive small-sized strain gauges that can easily be placed in the small cavities of the machines and mechanisms being monitored; this makes it possible to simplify considerably the experimental apparatus used for studying the operation of machines and mechanisms under conditions of their normal operation and testing.

It should also be noted that electron-mechanical dynamometers can be used in experimental physics as well.

4. Acoustic Instruments

The high sensitivity of electron-mechanical transducers to small displacements and small mechanical forces makes their use especially attractive as highly sensitive transducers in acoustic instruments15. First of all, electronic microphones were implemented. The basic circuit of an electronic microphone consists of a membrane, which is the receiver of sound vibrations, and an electron transducer connected to it that senses the displacements of this membrane. Especially successful was the use of a vibrotron as the transducer in electronic microphones. In the circuit described, the electronic transducer operates as an externally controlled seismotron, recording the vibrations of the membrane. Such microphones prove especially convenient for reproducing the low frequencies of the audible range of sound vibrations29.

It is also of interest to point out a second, no less promising method for implementing electronic microphones with an internally controlled seismotron, whose resonant frequency lies below the level of the sound frequencies reproduced by the microphone. It consists of a conical membrane, at the apex of which an extremely light seismotron is mounted. The vibrations of the latter are accompanied by the generation of electrical oscillations corresponding to the sound waves incident on the microphone membrane. With a sufficiently small seismotron mass, it is possible to obtain a good frequency response for such a microphone.

If the kinematic system of a microphone with an externally controlled transducer has its natural frequency considerably below the range of sound vibrations, then it becomes an electron-mechanical sound-pressure meter.

One of the first applications of the electron-mechanical transducer in acoustics was its use as a pickup1 for reproducing mechanical sound recording15. The frequency characteristics of such a pickup and its sensitivity proved quite sufficient for this purpose, and the RCA firm, in particular with Olson’s participation, developed several designs of such pickups. Such instruments are, in essence, analyzers of microrelief. It was therefore quite natural to use similar transducers as sensors in instruments for monitoring surface roughness7, 8.

In analyzing surface roughness, one usually tries not to reproduce surface waves caused by periodic oscillations in the position of the cutter or another machining tool. These low-frequency components can be filtered out by means of electrical filters. Good results, as Bidwell showed, are obtained by using mechanical, and especially hydraulic, filters7.

The possibility of obtaining signals of considerable intensity from an electron-mechanical transducer opens up prospects for using a device analogous to an electron-mechanical pickup as a generator

electrical oscillations of audio frequencies. For this purpose the pickup needle must slide along a groove of sinusoidal shape; to this end, an ordinary gramophone-record player may also be used, on which records with sinusoidal tracks of different wavelengths are placed. Such a device makes it possible to obtain a series of fixed frequencies of electrical oscillations. To obtain an audio generator with a continuous range of obtainable frequencies, it is expedient to retain, as the motor, a synchronous motor3 supplied from an alternating-current mains or from a stabilized low-frequency generator, and to use an adjustable transmission from the motor to the record rotated by it.

Using electromechanical tubes of sufficiently high power sensitivity, it is not difficult to obtain a signal of sufficiently great intensity that in many cases it can be used without preliminary amplification. Thus, for example, with the aid of an electromechanical transducer with a sensitivity of about 10,000 V/cm, it is not difficult to obtain an amplitude of electrical oscillations of about 25 V with a sound-recording amplitude of 25 μm. Specialized transducers make it possible to take from such an audio generator sinusoidal signals with a power of up to 1 W and even more, without any amplification.

It is also possible to construct a microrelief analyzer, using for this purpose an extremely light seismotron attached directly to a needle moving over the surface. (It is desirable that the mass of the seismotron be comparable with the mass of the needle on which it is mounted.) Here, too, the use of a hydraulic filter proves expedient.

The kinematic systems of electromechanical transducers for acoustic apparatus proposed and developed earlier usually consisted of a rigid rod, the middle part of which is fastened to an elastic support (which, in particular, is a membrane). One end of the rod is subjected to the action of the mechanical force being monitored, while at the other end of the rod is mounted the electrode of a mechanically controlled electron tube.

Comparatively recently the possibility was demonstrated of implementing in mechanotrons another kinematic system, in which the rod, the end of which is acted upon in the transverse direction, is a hollow metal tube elastically fixed at the other end. Inside the tube there is a second (stationary) electrode of a long-control-gap glow-discharge mechanotron. Its internal electrode is covered along its entire length with a protective layer of insulator, except for the end exposed opposite the side wall of the tube, which performs the function of the movable electrode of the gas-discharge tube. Transverse oscillations of the tube are accompanied by corresponding oscillations of the voltage on the tube. With sufficient rigidity of the tube, such a system makes it possible to obtain considerably higher resonant frequencies and, consequently, considerably better frequency characteristics of electromechanical transducers.

In the case where a needle is attached to the free end of the tube, a pickup transducer or a surface-frequency analyzer is obtained; and when a kinematic connection with a membrane is implemented, a microphone or manometer transducer is obtained.

5. Hydromechanical Instruments

The use of electromechanical transducers in experimental techniques for studying hydrodynamic and aerodynamic processes makes it possible to improve substantially the most important parameters of measuring and recording apparatus, and also to implement a number of new instruments, the manufacture of which on the basis of other types of transducers is often very difficult.

First of all, mention should be made here of devices for measuring velocity and recording pulsations in the velocity of liquid and gas flows. The diagram of such a sensor, based on the principle of a hydraulic pendulum, is shown in Fig. 8, a. The flow-velocity sensor consists of an electromechanical sensor 1, on whose movable rod 2 a plate 3 is fastened, placed inside the flow perpendicular to the direction of motion. The pressure of the jet on the plate causes it to deflect through an angle proportional to the flow velocity. The entire sensor, except for plate 3, is enclosed in a streamlined housing, introducing minimal distortions into the controlled flow.

In order to be able to record pulsations of the flow velocity, it proves expedient to reduce the mass of body 3 to a minimum. Brecher and Pragin30 used, as body 3 receiving the pressure of the flow, a thin wire fixed on the movable rod of an RCA-5734 vibrotron. The diagram of the sensor developed by them for monitoring pulsations of the velocity of blood in an artery and in a vein is shown in Fig. 8, b. Here 1 is the tube in which the controlled flow flows, 2 is an RCA-5734 tube, 3 is the movable rod of the tube, on which wire 4, immersed in the controlled flow, is fastened. The natural frequency of the movable system of the sensor proved to be about 500 c/s. The use of an electromechanical sensor made it possible to implement a number of other designs of similar instruments33, 34.

Fig. 8

Of some interest is also the possibility of using, for measuring the flow velocity, the sensor described above, made in the form of a tube elastically fixed at one end with a stationary internal electrode. The angle of inclination of the tube, and consequently the sensor signal, depend unambiguously on the flow velocity. The dynamic characteristics of such a sensor may prove better than those of a sensor implemented on the basis of the use of a vibrotron30, 33.

For the purposes of recording dynamic pressures in a liquid flow and hydraulic shocks (for example, impact loads when waves strike the surfaces of engineering structures), it proves expedient to use sensors with a bending membrane, for example of the type shown in Fig. 1, g. The possibility of using membranes having a sufficiently high resonant frequency makes it possible to apply such sensors to the study of very rapid processes in liquid and gaseous media. For example, it proves possible to use such sensors for studying detonation in the cylinders of internal-combustion engines.

Among the problems in which the use of electromechanical sensors gives substantial results should also be included electromechanical differential-type manometers, in which the deflection of the membrane is recorded by an electromechanical sensor.

III. MEASURERS OF NONMECHANICAL QUANTITIES

1. Thermal devices

The use of electromechanical sensors in thermometric apparatus is based on the conversion of thermal deformations of the receiver of the controlled quantity into displacement of the movable electrode of the mechanotron. The very high sensitivity of electromechanical sensors to displacements of the movable electrode makes it possible to carry out highly sensitive

thermometric instruments, which can be used both as temperature sensors and as sensors for a number of other quantities whose values can be converted into the corresponding changes in temperature.

We shall first describe certain types of electromechanical thermometers, and then show some methods of using such devices as essential elements of other, non-thermometric apparatus.

In Fig. 9, a, a diagram is shown of an electromechanical thermometer with a wire sensing element formed by two wires made of different metals (for example, bronze \(1\) and tungsten \(2\)). The first of them connects the tension spring \(3\) with the movable rod \(4\) of the electromechanical sensor of small displacements \(5\), while the second connects the tension spring with the housing of the sensor. Such a kinematic scheme ensures that the instrument responds only

Fig. 9.

to the change in the difference of the lengths of the wires caused by a change in the air temperature, while excluding the influence on the readings of the device of thermal deformations of the instrument frame. Sections \(6\) of the wires, located in the immediate vicinity of the sensor, should preferably be made of the same metal in order to reduce the influence of the elevated temperature of the electron tube on the thermometer readings.

Assuming the lengths of wires \(1\) and \(2\) at \(0^\circ\mathrm{C}\) to be equal to \(L_0\), and denoting their thermal expansion coefficients by \(\alpha_1\) and \(\alpha_2\), we obtain expressions for the sensitivity of the electromechanical temperature sensor with respect to voltage:

\[ v_t=(\alpha_2-\alpha_1)\cdot L_0\cdot \varphi \tag{16} \]

and with respect to current:

\[ i_t=(\alpha_2-\alpha_1)\cdot L_0\cdot \psi. \tag{17} \]

Taking, for example, \(\alpha_2-\alpha_1=10^{-5}\ \mathrm{cm/degree}\), \(L_0=10\ \mathrm{cm}\), \(\varphi=10000\ \mathrm{V/cm}\), \(\psi=1\ \mathrm{A/cm}\), we obtain the values \(v_t=1\ \mathrm{V/degree}\), \(i_t=10^{-4}\ \mathrm{A/degree}\).

If, as the temperature-sensing element, a body is used that makes it possible to obtain substantially greater deformations per degree than a straight wire, then it becomes possible to implement thermometric-device sensors that are more sensitive both in current and in voltage. In particular, the use of bimetallic plates proves convenient for this purpose. In Fig. 9, b, a diagram of such a device is shown, in which one end of the bimetallic plate \(1\) is attached to the frame of the temperature sensor, while the free end of this plate is connected to the movable electrode of the mechanically controlled electron tube \(2\), which monitors the displacement of the free end of the bimetallic plate. A change in temperature is accompanied by a corresponding change in the anode current of the electromechanical sensor.

The high sensitivity of electromechanical temperature sensors makes it possible, on the one hand, to use pointer galvanometers for direct reading of small temperature changes and, on the other hand, easily to implement highly sensitive temperature regulators and temperature recorders without substantial amplification of the sensor signal, and sometimes even entirely without it.

without amplifying it. In particular, this proves convenient when using thermal integrating light sensors for spectroscopic studies. Later we shall touch on the principles of implementing such sensors on the basis of mechanotrons.

When choosing the most suitable mechanotron for an electro-mechanical thermometer, one should select tubes combining maximum sensitivity with minimum power dissipated during operation, so as to ensure minimal influence of the heat released by the tube on the thermometer readings. Therefore, gas-discharge sensors of small displacements with a cold cathode, which can be switched on only for the time required to read the monitored temperature, prove especially convenient for this purpose.

In addition to the externally controlled electro-mechanical thermometers described above, it also proves expedient to use internally controlled electro-mechanical thermometers. Let us show the principle of implementing one such sensor (see Fig. 9, v). Here the movable rod 1 of the mechanically controlled electron tube is fastened to a torsion bifilar suspension 2. The suspension loops are made of materials having substantially different thermal coefficients of linear expansion. A movable electrode 3, controlling the anode current of the tube, is fastened to rod 1. When the temperature of the tube changes, the two loops are deformed differently, and rod 1 rotates through an angle proportional to the change in the temperature being monitored. The suspension is fastened in a frame mounted inside the electron tube, and is tensioned by thread 4 with the aid of spring 5. It is not difficult to show that an electro-mechanical thermometer of this type has much higher sensitivity in comparison with the above-described types of externally controlled electro-mechanical thermometers. Use of the described circuit proves especially interesting in implementing certain types of thermal converters.

In many kinds of experimental and industrial apparatus, thermometric elements are used as auxiliary devices employed to convert the monitored quantity into electrical signals convenient for measurement or recording. In this connection, in a number of cases it proves expedient to use electro-mechanical converters. Let us show this with several typical examples.

One of the most promising applications of such devices is the implementation of thermal converters for measuring high-frequency currents. By directing the current to be measured into one of the two wires 1 or 2 of the circuit shown in Fig. 9, a, we obtain the possibility of taking from electro-mechanical sensor 5 a signal proportional to the power released by the current. Such a device may have a sensitivity many times greater than that of analogous devices based on the use of thermoresistances and thermocouples.

An analogous kinematic scheme can also be implemented inside the tube, i.e., with an internally controlled sensor. In this case, however, it is expedient to use the circuit shown in Fig. 9, v, and to pass the current through one of the suspension loops of the torsion balance or through the wire to which one of the loops is attached.

The sensor device whose circuit is shown in Fig. 9, a, may also be designed so that wires 1 and 2 are made of the same material and have the same diameter. In such a design the device proves suitable for measuring the temperature difference between two media. Thus, for example, if one of the wires is covered with moistened material, we obtain a device for direct reading of the psychrometric difference.

By passing one and the same current successively through identical wires 1 and 2, we obtain the possibility of using such a device

as a sensor of the differing thermal conductivity of the media in which the two wires are located. Consequently, such a device is suitable for use as a gas analyzer, in which one wire is in the reference medium and the other wire is in the controlled medium. Such a device can also be used for direct monitoring of moisture deficit. For this purpose one of the wires is placed in an atmosphere of air saturated with water vapor, while the other is placed in the controlled atmosphere having the same temperature. For direct measurement of absolute humidity by means of such a device, one wire should be immersed in dried air, and the second wire in the controlled atmosphere of the same temperature.

Finally, it is possible to use the device according to scheme 9,a with identical wires 1 and 2 in a regime in which the same constant current is passed through both wires, while through wire 1 there is also passed an even weaker alternating current of comparatively low frequency (of the order of a hertz), or only an alternating current. Then, as a result of the periodic cooling of the wire, its length will change periodically, and pulsations of the anode current will be observed in the anode circuit of the electromechanical sensor, the depth of which is determined by the thermal conductivity of the medium, depending on its chemical composition, velocity of motion, or gas pressure. Consequently, on such a principle it is possible to realize gas analyzers, meters of air humidity, the velocity of motion of gases and liquids, as well as certain other parameters of the medium surrounding the wire.

2. Magnetic Instruments

The operation of many magnetic measuring instruments is based on measuring forces due to the controlled magnetic phenomena. Usually the acting forces are measured by means of mechanical or opto-mechanical devices with direct reading, which for the most part are simple and reliable instruments. The use of electromechanical sensors of these forces proves in many cases to be more expedient from the standpoint of increasing sensitivity and extending the ranges of measurement.

Here one should first of all note the electromechanical magnetometer of external and internal control. The circuit of an electromechanical magnetometer of external control is extremely simple. A straight permanent magnet is fastened to the movable rod of the mechanotron. Under the action of an external magnetic field the rod then experiences a torque, and the angle of inclination of the rod is registered by the mechanotron. In an electromechanical magnetometer of internal control, the straight magnet is fastened to a suspension and is rigidly connected with the movable electrode of a mechanically controlled electron tube. In both types of electromechanical magnetometers described above, the permanent magnet may be replaced by a coil through which a direct current is passed.

The portability of such sensors makes it possible to carry out direct measurements of the magnetic-field strength and the values of its components at various points of the field, with the measured quantities read directly on a pointer galvanometer. The electromechanical magnetometers described can be used for purposes of magnetic flaw detection.

It may also be expedient to use them for determining the force necessary to detach a magnet from the surface of a specimen.

Such a device is suitable for use in meters of the thickness of strips made of ferromagnetic materials, for determining the thickness of protective coatings on steel and other ferromagnetic materials, and also for determining the thickness of a nickel layer covering non-ferromagnetic materials.

3. Electrono-mechanical amplifiers

The use of electromechanical transducers that transform an electrical signal into proportional mechanical forces for subsequent action on the kinematic system of a mechanotron opens up the prospect of producing electrono-mechanical amplifiers^27.

In order to obtain a substantial gain that justifies the implementation of such amplifying devices, it is necessary to use in them highly sensitive electromechanical transducers that make it possible effectively to convert the amplified signal into a displacement of the movable electrode of the mechanotron.

In Fig. 10, a a diagram is given of an electrono-mechanical amplifier with an externally controlled mechanotron. Here the signal to be amplified is fed to the coil of electromagnet 1, whose armature is mechanically connected with movable electrode 2 of mechanically controlled electron tube 3. The displacement of the armature under the action of the signal is accompanied by a corresponding change in the anode current of the tube, which may be greater than the primary current.

Fig. 10.

The use of such an amplifier may be meaningful when amplifying very small voltages. For this purpose it proves expedient to make the coil of electromagnet 1 of as low a resistance as possible. In such an amplifying device one can obtain a rather considerable voltage gain.

However, for purposes of physical experiment, electrono-mechanical amplifiers with internally controlled mechanotrons are of considerably greater interest. The basic circuit of such an amplifier is shown in Fig. 10, b. Here coil 1 acts on one of the permanent magnets 2 of an astatic magnetic system 3, analogous to an electromagnetic galvanometer with rotating magnets. To this same astatic system is attached movable electrode 4 of mechanically controlled electron tube 5.

For a rough estimate of the voltage gain of such a system, one may assume that the suspension system of a sensitive electromagnetic galvanometer, analogous to that used in such an amplifier, has a sensitivity of about \(10^{-6}\) volts per scale division located at a distance of 1 meter from the galvanometer. This corresponds to a displacement by 10 microns of the end of the movable electrode located at a distance of 2 cm from the axis of rotation of the suspension. With a mechanotron sensitivity of \(10\,000\) V/cm, a change in the voltage applied to the coil by \(10^{-6}\) V leads to a change in the voltage on the mechanotron by 10 V, which gives a voltage gain equal to \(10^7\).

An analogous amplifying device can also be realized in the case where the moving system used is the rotating frame of a magnetoelectric galvanometer.

4. Electrono-mechanical regulators

The possibility of easy regulation of the internal resistance of mechanotrons as a result of a relatively small displacement of the movable electrode, requiring insignificant amplification, makes it possible to implement

contactless resistances capable of dissipating relatively large powers within themselves and requiring relatively small efforts for regulation.

Mechanically controlled electron tubes can be used as rheostats in direct- and alternating-current circuits, which is of considerable interest for the technique of physical experimentation. In Fig. 11 the principal circuits of the main types of electron-mechanical rheostats are shown. In Fig. 11, a a circuit of a diode rheostat with longitudinal control and flat parallel electrodes is shown. The resistance of such a rheostat is

Fig. 11.

\[ R=\frac{a^3}{A\cdot S\cdot U_a^{0.5}}, \tag{18} \]

where \(a\) is the distance between the electrodes of the tube in cm, \(A=2.33\cdot 10^{-6}\), \(S\) is the active surface of the hot cathode in \(\text{cm}^2\), and \(U_a\) is the voltage across the rheostat. From this relation one sees the possibility of obtaining a thousandfold change in the resistance of the rheostat when the gap is varied over the range from \(0.1\) mm to \(3\) mm. Reducing the transverse dimensions of the electrodes of this diode to values smaller than the maximum separation of the electrodes makes it possible to increase still further the range of relative changes in the resistance of the rheostat. An example of such a solution may be a diode with a heated cathode about \(1\) mm wide, in which the electrodes can be separated by a distance of \(3\)—\(4\) mm and more.

A considerable range of resistance variation can be obtained with an electron rheostat of probe control, the circuit of which is shown in Fig. 11, b. Here \(A\) is the anode, \(X\) is the cold cathode, and \(K\) is the hot cathode. Displacement of the latter in the direction shown by the arrow is accompanied by very rapid changes in the resistance of the rheostat.

In Fig. 11, c the circuit of a rheostat with differential control is given. Here \(A\) is the anode, \(X\) is the cold cathode, and \(K\) is the hot cathode. Displacement of the anode \(A\) in the circuit of the cold cathode in the direction shown by the arrow is accompanied by rapid changes in the resistance of the rheostat. Rheostats with differential control make it possible to obtain a change in the resistance of the rheostat by up to \(5\)—\(6\) orders of magnitude when the movable electrode is displaced by several millimeters. The possibility of easily varying the resistance within extremely wide limits is of definite interest for certain problems of experimental technique.

The relatively small forces that are found to be necessary for moving the movable electrode of the regulator make it possible to use such rheostats for purposes of automatic regulation of individual circuits of electronic apparatus. In particular, such rheostats can also be used for purposes of automatic stabilization of the voltage or current of current sources used in individual experimental installations.

Electron rheostats of the type described above possess unilateral conductivity. This makes it possible to use them as mechanically regulated valves, which makes it very simple to carry out automatic stabilization of the voltage or of the rectifier output current.

For purposes of regulating alternating-current circuits, electron rheostats of bilateral conductivity can be used. A typical example of such a rheostat is a diode with two flat parallel heated cathodes, one of which is fixed and the other movable.

In conclusion, the information presented above on electromechanical transducers should be supplemented with information on electromechanical relays. Relays of this type are electromechanical transducers with a fixed displacement of the movable electrode and make it possible to obtain an electrical signal of fixed value if the action on the electromechanical tube exceeds a specified value. Thus, for example, by using displacement limiters for the movable element of an electronic micrometer, we obtain a relay of linear dimensions. By using displacement limiters for the movable element of an accelerotron, we obtain an acceleration relay. By using displacement limiters for the movable element of an electromechanical thermometer, we obtain a temperature relay. By limiting displacements in an electromechanical amplifier, we obtain an electromechanical relay that makes it possible to obtain very considerable amplification, in voltage and in current, of electrical pulses.

It may be supposed that individual types of electromechanical relays for various physical quantities have prospects for application in the technique of physical experiment and in a number of applied problems.

CITED LITERATURE

1. L. A. Goncharskii, Uspekhi fizicheskikh nauk 55, 81 (1955).
2. E. G. Hopkins, Journ. Inst. Electr. Eng. 3, March, 77 (1954).
3. V. Zeluff, Electronics 24, No. 2, 120 (1951).
4. S. Levy, Exp. Stress. Analysis 9, No. 2, 151 (1952).
5. L. A. Goncharskii, Trudy sektsii prikladnoi elektroniki VNITO, No. 1 (1954).
6. L. A. Vallet, Electronique, No. 88, 12 (1954).
7. J. B. Bidwell, Electronics 26, No. 11, 68 (1953).
8. W. A. Wildback, Rev. Scien. Instr. 25, No. 1, 102 (1954).
9. M. Ainsworth, Journ. Scien. Instr. 31, 12, 471 (1954).
10. L. A. Goncharskii, Electronic sensors of mechanical quantities, in the collection Applied Electronics, VSNITO Publishing House, Moscow, 1953.
11. L. A. Goncharskii, Zhurnal tekhnicheskoi fiziki 24, 1741 (1954).
12. L. A. Goncharskii, Izmeritel’naya tekhnika, No. 3, 64 (1955).
13. R. Gunn, Journ. of Appl. Mech. 7, No. 2, 49 (1940).
14. M. T. Grekhova and R. P. Vasil’ev, Zavodskaya laboratoriya 12, 882 (1946).
15. H. Olson, Journ. of Acoust. Soc. of Am. 19, No. 2, 307 (1947).
16. L. A. Vallet, Electronique, No. 89 (1954).
17. W. Ramberg, Journ. of Res. Nat. Bur. of Stand. 37, No. 6, 391 (1946).
18. S. Levy, E. V. Hobbs, W. O. Kroll, L. Mordfin, Journ. of Research. Nat. Bur. of Stand. 50, No. 2, 71 (1953).
19. J. R. Stovall, Electr. Manufact. 48, July, 102 (1951).
20. S. V. Ptitsyn, Physical phenomena in the oxide cathode, Gostekhizdat, M.—L., 1949.
21. G. A. Tyagunov and I. V. Yudinskaya, Thermionic cathodes of electron-vacuum devices, in the collection Advances in Electron-Vacuum Technology, Gosenergoizdat, M.—L., 1956.
22. N. Kuzunoza, Calculations of electron tubes and design of triodes, Gosenergoizdat, M.—L., 1934.
23. Tele-Tech. 13, No. 11, 67 (1954).
24. H. Bürnheim, VDI Zeitsch. 92, No. 23, 650 (1950).
25. K. Kretzmann, Funk-technik, No. 8, 217 (1951).
26. Hübner, Funk-technik 10, No. 7, 186 (1955).

1. L. A. Goncharskii, An Electron Tube with Mechanical Control, Gosenergoizdat, Moscow–Leningrad, 1956.

2. A. A. Baikov and L. A. Goncharskii, “A pressure meter,” Priborostroenie, No. 4 (1956).

3. Journal of Acoust. Soc. of Am. 26, No. 6, 1080 (1954).

4. G. A. Brecher and J. Praglin, Proc. Soc. Exp. Biol. a. Med. 83, No. 1, 155 (1953).

5. J. H. Green, Journ. Physiol. (London) 125, No. 1, 4 (1954).

6. P. Bättig and K. Bücher, Helv. Physiol. et Pharm. Acta 12, No. 12, 50 (1954).

7. A. Scher, T. Weigert, and A. Young, Science 118, No. 3055, 82 (1953).

8. A. Müller, Helv. Physiol. et Pharm. Acta 12, 300 (1954).

9. G. M. Zhdanov, Telemetering, Gosenergoizdat, Moscow–Leningrad, 1952.

1. Reference 13 in the source. ↩↩

2. Reference 14 in the source. ↩↩

3. Reference 1 in the source. ↩↩↩↩↩

4. Reference 11 in the source. ↩↩↩