NEW INSTRUMENTS AND METHODS OF MEASUREMENT

ELECTROMETER TUBES

B. M. Tsarev

INTRODUCTION

The negligible amount of energy expended in electron tubes for controlling the electron flow from the cathode to the anode by means of a control grid has made it expedient to use electron tubes for measurement purposes, especially for measuring extremely weak currents. Not only special types of electrometer tubes, but also some ordinary mass-produced electron tubes used in radio engineering, successfully compete both with the most sensitive galvanometers and with various kinds of electrometers, simplifying, in particular, the technique of making measurements and the maintenance of the corresponding electrometric devices.

After the very first attempts to construct sensitive electrometric circuits using ordinary triodes and more complex multielectrode tubes, special electrometer tubes were developed. The further development of tube electrometry is proceeding precisely along these two paths: alongside the development of ever more advanced electrometer tubes, we see numerous examples of the application of ordinary electron tubes in special electrometric circuits and corresponding operating regimes. Therefore, in the present review we shall consider not only the designs and parameters of special types of electrometer tubes that have appeared during the last 10–15 years, but shall also briefly discuss the use, for electrometric purposes, of various types of mass-produced electron tubes.

The basic requirements imposed on electron tubes when they are used in electrometric circuits are: a minimum value of the grid current, which determines the smallest magnitude of currents measurable with circuits using such tubes, and stability of tube operation. The first requirement determines the choice of the tube design and operating regime. Fulfillment of the second requirement is also connected with the circuit in which the tube is connected. Therefore,

before proceeding to the characterization of the existing types of electrometer tubes, one should consider in greater detail the possibilities for the maximum reduction of grid currents in modern electron tubes and the question of the stability of tube operation in a circuit.

1. GRID CURRENTS, THEIR CLASSIFICATION, CAUSES AND METHODS OF REDUCING THEM

In what follows, when speaking of the control grid and grid currents, we shall have in mind the control electrode in the most general case and the currents in its circuit at negative potentials on this electrode. Very often, for example, in electrometer circuits triodes are used connected according to the so-called “inverted triode” circuit, when the role of an anode or electron collector is played by the grid located closer to the cathode, while the anode of the tube is the control electrode. Since at positive or small negative (down to \(-1.5\) V) voltages on the control electrode an appreciable fraction of the electron current emitted by the cathode reaches the latter, this mode of tube operation is of no interest from the standpoint of application to electrometry, and we shall confine ourselves to considering the case of sufficiently large negative biases on the control electrode.

Among the electrons emitted by the cathode there will always be some which possess initial emission velocities sufficient to overcome the existing negative potential of the control grid. Their number decreases with increasing negative grid potential, and the grid current then falls exponentially.

In practice, however, this electronic component of the grid current is partly compensated by other components of it, usually flowing in the opposite direction. At a certain bias on the control grid, called the “free-grid potential,” the grid-current curve \((I_{gz},\) Fig. 1) passes through zero, and at higher negative biases a grid current of the opposite direction is already observed; its nature proves to be much more complex than it seems at first glance.

In any electron tubes the grid current is composed of the following components:

1. The electronic grid current produced by the initial emission velocities of electrons from the surface of the cathode.

2. The ionic grid current, i.e., the current of positive ions arising in collisions of electrons with atoms and molecules of residual gases in the tube.

3. The leakage current through the insulation between the control grid and the other electrodes of the tube.

4. The current of electron emission from the grid, arising owing to heating of the grid to temperatures at which this current becomes

noticeable. As a rule, this current is observed only in tubes with heated oxide cathodes, in which the grid, along with considerable heating, is activated as a result of the deposition on it of a large amount of active substance from the cathode.

In addition, in special tubes with small grid currents, other causes of the appearance of grid current may also be observed:

Fig. 1. Characteristic of the grid current and its principal components.

1. Currents of photoelectron emission from the grid, caused by: a) light radiation from the cathode; b) illumination of the tube electrodes from outside; c) soft X-radiation arising inside the tube.

2. Currents of ionic emission from the cathode.

Let us consider separately the listed components of the grid current and the possibilities for reducing them.

The electron current of the grid \((I_{ge})\) in the case of a triode, as is known, is expressed by the equation:

\[ I_{ge}=m\cdot I_e \exp\left(\frac{eU_g}{kT_k}\right) \quad \text{for} \quad U_g^*\ll 0, \tag{1} \]

where \(I_e\) is the cathode emission, \(m\) is a dimensionless factor depending on the geometrical dimensions of the electrodes and, above all, on the degree to which the grid is filled by its turns, i.e., on the ratio of the diameter of the wire forming the grid to the pitch of its winding. In addition,

the quantity \(m\), as experiment shows, depends on the anode voltage, namely, it decreases as the latter is increased, and is completely independent of the grid potential. Here \(T_k\) denotes the cathode temperature, \(e\) the electron charge, \(k\) Boltzmann’s constant, and \(U_g^*\) the potential of the control grid. This equation is applicable, of course, only for negative values of \(U_g^*\), which represents the negative bias \(U_g\) applied to the grid with the corresponding correction for the contact potential difference \(U_k\) between the cathode and the control grid and for the magnitude of the potential minimum \(U_m\) created by the electron space charge between the cathode and the grid. Consequently, the relation must hold

\[ U_g^* = U_g + U_k + U_m < 0, \tag{2} \]

where \(U_k\) is the difference between the work functions of the cathode and grid surfaces.

As is evident from (1), the rate of decrease of the current with increasing negative bias on the grid is greater the lower the cathode temperature. From this point of view, in order to reduce the electron current to the grid it is more advantageous to use cathodes operating at lower temperatures, and therefore already in the first specimens of electrometer tubes we find cathodes of thoriated tungsten, used at present in most types of these tubes. The working temperature of oxide cathodes is still almost half as low \((800\text{—}900^\circ K\) instead of \(1700\text{—}1850^\circ K\) for thoriated cathodes), which have recently been used in a number of special electrometer tubes, despite the danger of the photoeffect caused by deposition of active material from the cathode onto the grid.

Fig. 2. a) Hausser tube; b) tube of type A154A.

Ion currents \((I_{gi})\), arising due to ionization of residual gases, can be considerably reduced by using in tubes

sufficiently active absorbers—getters, such as, for example, magnesium or barium. In most cases, however, special electrometer tubes use magnesium specifically, since barium can easily cause a photoelectric effect from the control grid, or leakage over the insulation of the latter, owing to its being deposited on the insulators. However, the principal method not only for reducing, but also for practically completely eliminating, ionic currents is to operate the tube at voltages on all electrodes that do not exceed the ionization potentials of most of the residual gases present in the tubes, i.e., first of all, the constituents of air ($\mathrm{O}_2$, $\mathrm{N}_2$ and Ar), and then the gases evolved by the parts of the tube (CO, $\mathrm{CO}_2$, $\mathrm{H}_2\mathrm{O}$, $\mathrm{H}_2$ and sometimes certain hydrocarbons). Therefore in most cases the anode voltages in electrometer tubes do not exceed 8–10 V. The provision of sufficiently good parameters is achieved by usually using, as electrometer tubes, so-called “two-grid tubes,” i.e., tubes with a first, retarding or cathode grid, which is at a small positive potential (4–6 V), and with a second control grid. In the case where triodes are used, they are most often used in the reverse-triode circuit, i.e., a positive potential is applied to the grid, and the anode serves as the control electrode. Examples of a two-grid tube and of a reverse triode may be, respectively, one of the first Hausser electrometer tubes1 (Fig. 2, a) and the RCA type A154 tube (Fig. 2, b).

Fig. 3. Type 4060 tube

Sometimes, as can be seen from the example of the Philips type 4060 tube (Fig. 3), the anode and the control electrode are arranged in the form of two plates on different sides of the cathode (the so-called “plation,” see also2), at different distances from it (the control electrode is closer to the cathode than the anode). In all the constructions listed, it is possible to obtain the required parameters at anode voltages of the order of 4–6 V and, in any case, not higher than 8–10 V.

As numerous investigations have shown, the ionic currents of the grid are proportional to the magnitude of the anode current and to the pressure of the residual gases in the tube (see, for example,3). However, in most cases of using tubes in electrometer circuits, currents caused by ionization of the gas are practically absent, and the small ionic currents that arise in some cases are due in their origin, as we shall see below, to emission of positive ions by the cathode.

Grid leakage currents \((I_{gc})\), i.e., the quality of its insulation relative to the other electrodes, can, by various means, be brought in practice to almost any desired values and, in essence, their minimum values are most often limited by the dimensions of the tube and by the voltages used for it.

In special electrometer tubes, as a rule, the control grid is mounted inside the tube on special insulators, glass or quartz, and the lead of the control grid is usually made on the dome of the bulb. In addition to leakage currents along the internal tube leads, there is consequently also leakage along the glass of the tube bulb, and chiefly along its outer surface. To reduce leakage along the bulb, in some tubes—for example, in the type D96475 tube of Western Electric (see⁴)—a long glass tube of small diameter is welded onto the dome of the bulb; the control-grid lead passes through this tube. To avoid the deposition of metallic conducting films on the inner walls of the tube, its lower opening is closed by a small metallic disk (Fig. 4). The insulators on which the control grid is mounted may be made either as straight rods or as curved ones (Figs. 2 and 3), in order to lengthen the leakage path. To protect the insulators from the deposition of metallic films on them, they are shielded by glass caps (Fig. 5), or special skirt-shaped insulators are used (Fig. 4). Sometimes the metallic holders on which the insulators are mounted are provided with a separate lead, to which, when the tube is connected into a circuit, a potential equal to the average operating potential of the control grid is applied; this makes it possible to reduce leakage along the insulator considerably, since both its ends are then practically at the same potential.

Fig. 4. Type D96475 tube.

Thanks to the measures listed above, the only principal leakage that remains is leakage along the outer surface of the bulb. To reduce this leakage, it is recommended that the bulb of the tube be thoroughly washed with pure alcohol and then dried by heating in a drying cabinet. In addition, the insulating properties of the bulb surface can be significantly improved by coating it with a thin layer of paraffin. Recently, for the same purpose, various organosilicon compounds have begun to be used, applied in a thin layer to the surface of the bulb⁵˒⁹. These protective coatings, which protect the glass surface from the absorption of moisture that severely worsens its insulating properties, are especially necessary when ordinary types of electron tubes are used as electrometer tubes.

In addition, in order to reduce leakage over the bulb it is recommended to put on it a protective metal ring in the form, for example, of a glued strip of tinfoil or a layer of aquadag (graphite). Such a ring may also be provided by the metal cup of the base found on some tubes. The protective ring or the base cup is then either grounded, or a potential equal to the average operating potential of the control grid is applied to it (Fig. 6).

Thermionic emission of the grid \((I_{gT})\) in special types of electrometer tubes is practically always absent, chiefly for two reasons: first, economical directly heated cathodes are used, mainly, i.e. of thoriated tungsten or oxide cathodes, and, second, the operating temperature of the cathode is always chosen as low as possible, i.e. of the order of \(650—700^\circ\mathrm{K}\) for oxide cathodes and \(1650—1750^\circ\mathrm{K}\) for thoriated tungsten. In addition, the grid situated closer to the cathode usually plays the role of the cathode counter-grid or, in the case of an inverted triode, the role of the anode, while the second grid or the anode is the controlling electrode; as a result its temperature is always sufficiently low and therefore even in the case of tubes with a heated oxide cathode the thermionic emission of the control grid is practically absent.

The four principal components considered by us usually give the picture of grid currents shown in Fig. 1. In the case of special electrometer tubes

Fig. 5. FP-54 type tube.

Fig. 6. Protective ring reducing leakage over the bulb.

or when ordinary tubes are used in electrometer regimes, ionic currents and thermionic emission of the grid, as indicated above,

are absent, and the grid currents usually consist only of the electron current and the leakage current (Fig. 7). But under these conditions other, weaker components indicated above also begin to play a significant role in the formation of the grid current.

Photoelectron emission from the grid under the influence of radiation from the cathode is practically entirely excluded, since in the case of an oxide cathode, when the activity of the grid surface may be sufficiently high, the wavelength of the cathode radiation lies beyond the red limit of the photoeffect; in the case of thoriated tungsten, the activity of the grid is always sufficiently low.

To eliminate the photoeffect from the grid under the influence of illumination from outside, it is recommended that the electrometer tube be placed in a light-tight cap. In addition, in the case of tubes with an oxide cathode, the photoeffect of the grid, as well as its thermionic emission, is considerably weakened by coating the grid with a layer of carbon (soot or graphite).

Fig. 7. Grid current of a type FP-54 tube.

Fig. 8. Dependence of the grid current on the anode voltage.

Finally, photoelectron emission from the grid may also be caused by soft X-radiation arising in the tube itself. It may be significant not only at anode voltages of the order of many tens of volts. Even at voltages of only about 10—12—15 V, supersoft X-radiation—so-called Bucky rays—is possible, quite sufficient for the appearance of photocurrents comparable with other components of the grid current. This explains the sharp increase in the grid current at anode voltages exceeding 8 V (Fig. 8), when phenomena of ionization of residual gases practically do not yet occur. Among other reasons, this circumstance also compels one to limit the anode voltages in electrometer circuits to values of the order of 6—8 V.

Electrometer Tubes

Ion emission from the cathode is possible for cathodes of any type. However, if in the case of pure and thoriated tungsten it is caused chiefly by volatile impurities in the metal and therefore decreases sharply with time, and, with sufficiently good treatment of the cathode, is entirely absent, then in the case of oxide cathodes it occurs continuously, owing to the continuous electrolysis of the oxide layer under the influence of the emission current, with the liberation of oxygen ions, and also owing to the evaporation of both metallic barium and its oxide from the cathode surface, accompanied in part by ionization.

Owing to the withdrawal from the cathode of an insignificant current, measured in a few tens of microamperes, and moreover at insignificant anode voltages, of which the voltage drop across the oxide layer of the cathode is insufficient for electrolytic dissociation of barium oxide, and also owing to the sufficiently low temperature of the cathode, which limits the process of evaporation of its oxide layer, the currents of ion emission from the cathode can be insignificant. In addition, as a result of applying a positive potential to the grid nearest the cathode, the positive ions leaving the cathode predominantly return back to the cathode, which prevents their reaching the control electrode.

Finally, in the case of the use of electrometer tubes as the first amplification stage for Geiger–Müller counters, one should also point out the possible influence on the operation of the counter of the radioactive radiation of thorium present in the tube in the case of a thoriated-tungsten cathode. In this respect the oxide cathode should have an advantage over the thoriated one. The absence of direct indications in the literature of such an influence is apparently explained by the extremely low intensity of this radiation and by the practical impossibility of distinguishing its influence on the operation of the counter from the influence of the many other factors causing the so-called background of the counter.

2. MAIN TYPES OF ELECTROMETER TUBES

In connection with the wide use of tube electrometer circuits, which make it possible to measure often extremely small currents, down to currents below \(10^{-15}\ a\) (down to \(10^{-17}\ a\)), and to detect still weaker currents, in recent years a considerable number of special electrometer tubes have been developed, satisfying the most varied conditions of their application. In construction all these tubes may be divided into three groups: tetrodes of the two-grid-tube type, i.e. with a positive cathode grid; triodes operating according to the inverse-triode circuit, i.e. with good insulation of the anode; and triodes of the “Plation” type, i.e. with two plates located on different sides of the cathode. Tubes of the first group have found the widest application, and, besides ordinary tetrodes,

Table 1

Basic design data of electrometer tubes

Electrometric Tubes

recently specially developed double tetrodes have come into use (see \({}^{6,7,8}\)), employed in balanced (bridge) circuits and making it possible to achieve, as we shall see in § 3, extremely high stability of operation of electrometric circuits at their high sensitivity.

The principal design data for the chief types of electrometric tubes are given in Table I. The parameters of tetrodes and triodes are given respectively in Tables II and III. In addition to the normal

Table II

Principal parameters of electrometric tubes (tetrodes)

operating conditions recommended for individual types of tubes, Table IV gives examples of special conditions with strongly reduced working voltages and cathode temperatures, which make it possible considerably to lower the limiting measurable values of the currents.

As can readily be seen from the tables, the greatest sensitivity of electric circuits can be attained with tetrodes. However, very often circuits of lower sensitivity are needed, namely for a current range from \(10^{-9}\) to \(10^{-12}\)—\(10^{-13}\) A, but cheaper, simpler in operation, and more mechanically stable when used in portable circuits. For this purpose both special triodes and, in certain cases, some types of ordinary electron tubes can be used successfully. However, in the latter case it is often necessary to

Table III

Basic parameters of electrometer triodes

Table IV

Examples of operating modes of electrometer tubes that make it possible to reduce the minimum measurable currents attainable

specially select individual specimens of tubes which, if possible, have smaller leakage between electrodes.

In connection with the wide use of light portable electrometer circuits, as well as portable dosimeters for various radiations, special types of small-size and even subminiature electrometer tubes have been developed which possess, in addition to small

compact dimensions, extremely low current consumption for heating the cathode, permitting the use of miniature batteries for their supply.

Fig. 9. Connection circuits of an electrometer tetrode.

Thus, for example, a subminiature tube of type VX41 (see⁹) requires only 13 milliwatts of power for heating the cathode. The cathode of this tube—an oxide-coated one with a core of nichrome wire 10 microns in diameter—has a filament current of only 10 ma. Despite its small dimensions and the bringing-out of the control grid at one end of the bulb together with the other electrodes, this tube has a grid current not exceeding \(10^{-14}\) a, which is achieved by coating the bulb of the tube near the place where the leads are sealed in with a thin film of an organosilicon compound. If the surface of the bulb of such a tube becomes contaminated, it is sufficient to wash it with pure alcohol, which does not dissolve the organosilicon film but removes the contamination; then, after washing with distilled water, the tube must be heated in a drying oven at a temperature of about \(100^\circ\text{C}\).

Fig. 10. Tube type SИ2.

It should finally be noted that, owing to identical insulation conditions for the leads of all electrodes (or, which is sufficient, of both grids), the VX41 tube permits its use in various connection circuits, namely (Fig. 9): as an electrometer triode, as a reverse triode, and as triodes with large and small amplification factors. The choice of one or another circuit depends on its application and on the desired range of currents to be measured.

3. STABILITY OF OPERATION OF ELECTROMETER TUBES AND METHODS FOR IMPROVING IT

The principal causes of insufficient stability in the operation of tube electrometer circuits are:

1. Instability of the circuit elements and chiefly of the resistances (especially high-ohmic ones), and of various kinds of contacts and connections.

1. Instability of the insulation of individual elements of the circuit and, in particular, of the tubes.
2. The presence of the influence of extraneous electric and magnetic fields.
3. Instability of the power sources.
4. Instability of the emission of the cathodes of electrometer tubes.
5. Insufficient mechanical strength and rigidity of the construction of individual types of tubes.

Consideration of the first four causes does not fall within the scope of the present review, and here we shall confine ourselves only to a few remarks. In particular, to ensure the stability of various kinds of contacts and connections in an electrometer circuit, they must be carefully soldered, in no case permitting either sliding contacts or the clamping of wires under terminals.

The second and third causes depend to a considerable extent on the thoroughness of the shielding both of the electrometer tube itself and of the entire circuit as a whole from the influence of external fields and of individual circuit elements from one another, and also on the cleanliness of the insulating parts of the circuit and the degree of humidity of the surrounding air. In particular, the chamber in which the electrometer tube is placed, serving simultaneously as an electromagnetic shield and as protection from the action of external light, is recommended to be dried by placing in it a vessel with calcium chloride or, preferably, phosphorus pentoxide. A number of remarks made earlier concerning the improvement of the insulation of the control grid of the tube may also be used in the construction of the circuit. Good materials for making insulators for fastening critical parts of the circuit requiring good insulation, and bushings for leads into the chamber with the tube, may be quartz and especially amber.

Insufficient stability of the power sources is to a considerable extent compensated by using, for example, storage batteries with as large as possible a reserve of capacity and discharge current, and by sufficiently long operation of the circuit in the switched-on state (up to several tens of hours) before making measurements, in order to attain stability of the battery voltages. In addition, fluctuations in the voltages of the power sources are well compensated by the use of two-tube balanced circuits. However, balanced circuits do not eliminate the influence of fluctuations of cathode emission, since these fluctuations are different in the two tubes operating in the circuit. As thorough investigations by Lafferty and Kingdon^6 have shown, fluctuations of the zero position of the galvanometer in the output circuit of an electrometer circuit may be of two kinds, namely: relatively rapid fluctuations of the “zero” and a comparatively slow displacement of it, most often to one side (drift). The main cause of this kind of instability in the operation of the circuit, and especially of drift, is the continuously continuing additional activation of the cold ends of the cathode during its operation. An increase in the duration-

of cathode activation (by a factor of 5) considerably reduced the drift. Rapid oscillations of the “zero” could be eliminated by shielding the cold ends of the filament, when the anode current of the tube was obtained only from its well-activated hot central sections. Replacement

Fig. 11. Schematic circuit diagram and electrode construction of a double electrometer tetrode.

of thoriated tungsten by an oxide filament almost completely eliminates the drift, but causes considerable rapid oscillations of the zero. The results of all these experiments are given in Table V. All these oscillations

Table V

Comparison of the operating stability of various tubes

are eliminated by the operation of two tubes in a bridge circuit. Therefore, for operation in such circuits a design of a doubled electrometric tube was proposed (Fig. 11), in which the cathode and the control-grid counter-winding were taken to be the same as in a tube of type FP54, while the two control grids and two anodes, which are halves of the grid and anode of this same tube, are arranged symmetrically on both sides of the cathode. Although the sensitivity of a bridge circuit operating with such a tube is reduced fourfold in comparison with a one-tube circuit using an ordinary FP54 tube (by a factor of two owing to the use of the bridge circuit, and by another factor of two because the anode current and the slope of the characteristic are reduced), nevertheless the stability of operation of the circuit with such a tube, as is seen from the same Table V, is considerably improved. This is explained by the fact that, owing to the influence of space charge, fluctuations of the cathode emission produce identical fluctuations of the anode currents of both halves of the tube, which mutually compensate one another when it operates in a bridge circuit. Such a tube is known under the name “FP54-split” (split FP54); analogous to it are the tubes: GL5674, which in external appearance is similar to the ordinary FP54 tube, since only one of its control grids is well insulated, and the DBM2 tube (Fig. 12). Since in the case of the V-shaped filament of the FP54-type tube it is difficult to achieve complete symmetry of both halves of the doubled tube, a straight filament was used in the DBM2 tube (see 7,8). Still greater symmetry was obtained in the doubled tube of type DBM6A with an oxide indirectly heated cathode (see 8). The principal parameters of doubled electrometric tubes are given in Table VI. The operating stabilities of the bridge circuit obtained with these tubes, expressed in millivolts per 1% change of heater current, are clearly characterized by the following results: whereas an ordinary FP54 tube gives, in a simple one-tube circuit, a stability of 50–100 mV, a doubled FP54-type tube gives 10–20 mV, and a DBM6A tube only 2–5 mV. In a special circuit compensating the influence of fluctuations of the electrode supply voltages, this value was brought, with the DBM6A tube, to 0.1 mV (see 8).

Fig. 12. Doubled electrometric tetrode.

Insufficient mechanical strength and rigidity in the construction of many electrometric tubes leads, in the case of portable installations or in the presence of vibrations in stationary installations caused, for example, by the presence of motors, to the appearance of such

Table VI

Basic parameters of dual electrometer tubes

Note. The parameters are given separately for each system.

called the microphone effect. To avoid the microphone effect, tube panels can usually be mounted on rubber, spongy gaskets, and their connection to the circuit made with flexible conductors. Owing to the small dimensions both of the tubes as a whole and of their parts, subminiature and small-size tubes have the most durable and rigid construction and are least subject to vibrations that cause the microphone effect; therefore they are applicable in portable instruments.

APPLICATION OF STANDARD MASS-PRODUCED TYPES OF ELECTRON TUBES IN ELECTROMETER CIRCUITS

Under normal operating conditions, most standard receiving-amplifying tubes have grid currents of the order of \(10^{-6}\)—\(10^{-7}\), more rarely \(10^{-8}\) A. By forcing them to operate at reduced cathode temperatures and at low potentials of all the other electrodes, these quantities can be reduced to \(10^{-9}\)—\(10^{-10}\) A, and only in specially selected specimens of certain tube types is it possible thereby to obtain still somewhat smaller values of grid currents. In most cases the reason for this is that, as a rule, the best insulation is possessed by the control grid nearest to the cathode, and types of tubes suitable for their good utili-

for use in special circuits of an electrometric tetrode or reverse triode. In this respect, many types of gridless tubes are more suitable for operation, for example, from the series of miniature and subminiature tubes or tubes of the “acorn” type. In tubes of these types all electrodes are in approximately equally good insulation conditions, and they can more easily be used in low-voltage modes of electrometric circuits, while obtaining much smaller values of grid currents, reaching \(10^{-12}\)—\(10^{-13}\), and sometimes even \(10^{-14}\) a. Thus, for example, an “acorn” tube of type 959, which is a directly heated pentode, when connected according to the circuit shown in Fig. 13, can give, in the mode of an electrometer of the reverse-triode type, a grid current of the order of \(10^{-12}\)—\(10^{-13}\) a, if, of course, the bulb of the tube is coated with a film of ceresin or a silicon-organic compound, and the filament is heated with a current of the order of 42—44 ma instead of the normal 50 ma. Table VII gives some examples of the use of mass-produced

Fig. 13. Circuit for using a type-959 “acorn” pentode as an electrometric tube.

Table VII

Operating conditions of mass-produced types of electron tubes in electrometric circuits

types of electron tubes in electrometer circuits. In this case, for a pentode of the UX222 type, the input resistance of the order of \(10^{14}\ \Omega\) is due almost exclusively to the current of positive ions from the cathode and, to a lesser extent, to leakage through the insulation. The reason for this is the use, as the control electrode, of the first grid, whose lead is made on the dome of the tube envelope.

References

1. K. W. Hausser, R. Jaeger, W. Wahle, Wiss. Veröffentlich. a. d. Siemenskonzern, 2, 325 (1922).
2. A. A. Shaposhnikov, Electron and Ion Devices, 1938, ch. VIII, § 13, pp. 293–294.
3. N. P. Morgulis, Zhurn. tekh. fiz., 1, 51 (1931); 3, 1276 (1933).
4. D. B. Penick, Bell Laboratory Record, 14, 71 (1935).
5. K. A. Andrianov and O. Gribarova, Organosilicon Compounds, Part II, TsBTIЭ, 1946; F. J. Norton, Gen. El. Rev., 47, 6 (1944).
6. J. M. Lafferty and K. H. Kingdon, J. Appl. Phys., 17, 894 (1946).
7. J. A. Darbyshire, Electronics Engineering, 18, 277 (1946).
8. G. C. Little, Electronics Engineering, 19, 365 (1947).
9. C. D. Gould, Electronics, 20, 106 (1947).
10. H. Daene and W. Husmann, A. E. G. Mitteilungen, 1937, No. 10.
11. G. P. Metcalf and B. J. Thompson, Phys. Rev., 36, 1489 (1930).
12. K. Henney, Electron Tubes in Industry, 2nd ed., 1937.
13. L. G. Fink, Electronics, 1941, p. 238.
14. J. Schintlmeister, Die Elektronenröhre als physikalisches Messgerät, pp. 24–31 (1942).
15. P. A. Macdonald, Physics, 7, 265 (1936).