Search

In this paper a method of drawing the signal flow graph of an active degenerate RLC network is given. From this signal flow graph the state equations of the network are obtained using Mason's graph gain formula.
The representation of networks in state equation form X(t) = AX(t) + But is advantageous in many respects. Recently Acar (1970) has provided a simple method for the formulation of state equations of active non-degenerate RLC networks using signal flow graph. The method is computationally more simple than earlier methods due to Dervisoglu (1964) and Kuh and Rohrer (1965), but unfortunately this cannot be applied for networks containing degeneracy. The purpose of the present paper is to provide a method for drawing the signal flow graph of an active RLC network containing degeneracy in the form of capacitance loops and inductance cutsets. By eliminating the dynamically superfluous variables with suitable linear equations of constraint due to degeneracies, the signal flow graph of the network may be properly drawn. By inspection of this signal flow graph the A and B matrices are directly obtainable using the graph gain formula. [ABSTRACT FROM AUTHOR]

In a series of papers and reviews Melehy (1964, 1965 a, 1965 b, 1966, 1968, 1970) has proposed an approach to thermodynamics via a somewhat unique definition of equilibrium, and in particular has illustrated his model by an examination of the equilibrium and steady-state conditions in semiconductor p-n junctions.
It is the purpose of the following paper to re-examine Melehy's hypotheses and to show that where his conclusions differ from those more generally accepted that it is due to misconception of semiconductor device principles. [ABSTRACT FROM AUTHOR]

It is well known that the network functions of an active network can be evaluated by means of complete trees or directed trees of its associated graphs. The paper presents and implements a matrix algorithm that generates all the directed trees or complete trees by a digital computer without redundancies due to duplications. The quantitative statistics of eight practical networks is given.
In the topological analysis of a linear active network, the problem of writing down network functions by inspection reduces ultimately to that of finding the set of directed trees or complete trees in the associated graphs. Since directed trees or complete trees of the associated graphs are in one-to-one correspondence with the corresponding major submatrices of certain associated matrices, the problem is equivalent to that of generating all the corresponding major submatrices efficiently. In this paper we shall present quantitative statistics reflecting upon the practicality of generating directed and complete trees based on an improved version of the matrix technique proposed by Mayeda et al. (1968) and Chen (1968). [ABSTRACT FROM AUTHOR]

In this paper an analysis of the first-order distortion to the current distribution in a short single-turn circular coil of high resistivity is presented. Both the in-phase and quadrature components arc given for different axial positions.
1. Introduction
There are several practical situations where short single-turn coils are used; these include induction heaters, plasma machines, magnetic forming of metals and radio-frequency circuits. In most of these applications the inductive impedance of the coil is greatly in excess of its resistive component and the current distribution may be analyzed for the greater part in terms of an infinite conductivity model (Barrault and McCartan 1966).
In other applications such as ohmic heaters and low-frequency air-cored magnetic field coils this situation is reversed and the current distribution is largely governed by the resistive component.
This paper presents an analysis of the first-order distortion to the current distribution resulting in this latter case from a weak inductive component. [ABSTRACT FROM AUTHOR]

The paper presents an active RC network which exhibits selective, all-pass and notch characteristics with simple adjustments of parameters. Two gain-parameters and passive-resistive elements are kept adjustable for this purpose, which also help to obtain variable selectivity and desired center frequency of the filter in the appropriate configurations. Discussions on the Q factor sensitivity of the filters have also been appended.
This paper describes an active RC second-order filter which can be adapted to operate in three different modes. As a band-pass filter it can be designed to have adjustable bandwidth and center frequency, which shows that the circuit may be used as a tuned selective filter. Secondly, depending on the choice of the system parameters, the circuit behaves as an all-pass network with adjustable delay. Finally, it is shown that the circuit operates as a very good rejection filter having continuously adjustable selectivity and notch frequency with no change in the shapes of its transmission characteristics. [ABSTRACT FROM AUTHOR]

The paper describes a circuit containing a junction transistor and a F.E.T. which behave as a one-port voltage controlled negative resistance of adjustable magnitude.
One-port negative resistance circuits, being very useful in synthesis of network functions, filter design and the design of regenerative circuits, have been developed by many authors (Waldmann and Beiri 1955, Indiresan 1959, Nagata 1965, Martinelli 1966). In this paper a new voltage controlled negative resistance circuit employing a transistor and a F.E.T. has been proposed which is different from the existing circuits regarding its network topology. The circuit requires no internal bias. As this circuit utilizes one transistor and a F.E.T., the circuit can be readily fabricated as an integrated module. [ABSTRACT FROM AUTHOR]

The realization of special types of the generalized-immittance converters will be given here using the second-generation current conveyor. Of particular interest in this paper is the new realization of the voltage generalized-immittance converter having conversion immittance function proportional to s. It will be seen that two realizations exist for each converter.
A special type of the current generalized-immittance converter was recently introduced, and used in one of the recent and most attractive RC active network synthesis procedures (Antoniou 1970). Antoniou (1970) used the operational amplifier as the active building block in the realization of the current generalized-immittance converter.
In a recent publication, Sedra and Smith (1970) introduced the second-generation current conveyor (will be abbreviated here as the current conveyor CC) and used it in the realization of some active network elements. The current conveyor is a versatile building block, as it can be used as a current summer, current differentiator, current integrator, current amplifier, etc.
In this paper the simulation of special types of both voltage and current generalized-immittance converters using the current conveyor will be given. [ABSTRACT FROM AUTHOR]

When harmonic oscillators utilizing elements with N-type and S-type voltage/current characteristics (elements with negative conductance or negative resistance) are calculated, usually only the non-linearity of the voltage/current characteristic is taken into account. However, the impedance of these elements is of a complex character, and the non-linearity of the reactive component influences the different modes of operation. This influence may be considerable and in some cases even dominating. If, however, it is taken into account, then the calculation of such systems becomes very difficult. The method used in this paper (Samoylo 1964, 1968) enables these problems to be treated correctly and at the same time considerably simplifies the mathematical calculations.
The results given in the paper may be used when harmonic oscillators utilizing tunnel diodes, Gunn diodes, avalanche diodes, avalanche transistors, unijunction transistors, p-n-p-n-type elements, etc., are studied.
The calculations are made considering the non-linearity of the active and reactive components of the impedance of these elements. The experimental results agree well with the theoretical results. [ABSTRACT FROM AUTHOR]

By inclusion of an external driving force, wave motion of any kind can be characterized by a dispersion function. This function is closely related to the energetic properties of wave motion, and then also to the averaged Lagrangian density. Linear and non-linear wave interaction can be analysed by inclusion of internal driving forces. Normalization procedures for the amplitudes can be avoided and time and space perturbations studied simultaneously. This analysis is further connected to slowly varying amplitudes and quasi-monochromatic waves. This paper presents the above-mentioned method and applies it to linear two-wave coupling, and non-linear three-wave coupling between positive and negative energy waves, and finally to amplitude modulation. The general equations obtained by this procedure are useful for general discussions. The simplicity of the method may prove useful in different applications.
The aim of this paper is to present a new and a simple method for analysis of wave interaction in dispersive media and to apply it on some illustrative problems. Further applications, which are found in plasma physics, microwave components and so on, will be presented later. [ABSTRACT FROM AUTHOR]

This paper describes gallium arsenide varactor diodes which develop effective values of f over 130 GHZ and will operate in double-diode parametric amplifiers at idler frequencies up to 43 GHZ. With such diodes it is possible to build wide band amplifiers which give good low noise performance without refrigeration. For example, an uncooled two-stage amplifier with a bandwidth of 500 MHZ at 26 dB gain gives an overall excess noise temperature at 7-25 GHZ of 85degreesK of which only 60degreesK is due to the varactor.
In recent years low noise parametric amplifiers which combine wide band-widths with high stability have been developed at S.E.R.L. for certain radar and satellite communication receivers. These amplifiers (Gittins et al. 1968) used the Pearson double-diode idler circuit (Pearson and Lunt 1964) as the cornerstone of the design, but the wide bandwidths were achieved by an innovation in which an extra resonant circuit of variable Q is coupled into the Pearson idler loop. Low noise temperature was obtained by cooling the diodes to liquid nitrogen temperature or below. However, refrigeration is unacceptable for many systems and, in particular, requirements at X-band now call for amplifiers with an excess noise temperature below 100degreesK when operating at ambient temperature. This sort of performance needs an idler frequency of at least 40 GHZ and an effective diode quality factor (f) greater than 100 GHZ. This paper describes the design and construction of new diodes which have such characteristics. [ABSTRACT FROM AUTHOR]

In this paper the gradual capture effect on the generation-recombination current in the space-charge region for an asymmetrical junction is studied. The theory is applied to the p-n InSb diodes. At forward bias, when the lifetime increases in the direction from p region to n region, the results we have obtained show that the current increases more rapidly with voltage than in the reverse case.
In this paper we shall take into account the gradual capture effect on the generation-recombination current in the space-charge region for an asymmetrical junction. [ABSTRACT FROM AUTHOR]

Plasma data given previously for the negative glow and Faraday dark space of near-normal cold cathode discharges in neon are supplemented by results, described briefly, for similar discharges without positive column or anode glow. The axial plasma electron concentration curves, which are sharply peaked near the cathode dark space when the cathode rays are strongly attenuated towards the anode, flatten markedly as the attenuation decreases.
The present paper contains a brief account and discussion of selected measurements of this kind made in neon (Grimley 1965). They supplement results for comparable discharges with an anode glow or positive column present, which have been described elsewhere (Woolsey et al. 1969). The critique of techniques given in the latter and in another paper (Emeleus 1969) will not be repeated here, beyond remarking that (a) current drain from the discharge by a probe near space-potential, although it will tend to produce some truncation of concentration maxima, seems unlikely to affect the main features of the results, and (b) it would nevertheless be desirable to repeat the measurements using ion-accelerating rather than space-potential data, although results obtained by the two methods in a similar discharge with rather larger concentrations agree well (Stern 1969). Unless otherwise stated, attention will be confined to the slowest (plasma) group of electrons. [ABSTRACT FROM AUTHOR]

For several years the so-called 'shorted emitter' principle has been used widely to improve the performance of silicon-controlled rectifiers.
In this paper the evolution of the emitter shunt is surveyed, and a review of simple theoretical principles is followed by an account of experiments which substantiate the general theory.
Since the publication of a paper by Aldrich and Holonyak (1959) which outlined a method of improving the temperature and rate of rise performance of the thyristor, many references to this so-called shorted emitter technique have appeared in the literature (Gentry et al. 1964, Gentry et al. 1965, Ruhl 1970, Chu 1970). However, in spite of the fact that the technique is now widely used in the production of commercial thyristors, little detail has been given in many of these references and no comprehensive survey is available.
Over the past few years, in our laboratories, a number of investigations have been carried out, and techniques have been established to make use of this shorted emitter principle.
It now seems an appropriate time to review this important field of semiconductor technology and give a brief account of experiments which have substantiated the theory. [ABSTRACT FROM AUTHOR]

This is the final part of a three-part series on the problem of efficient generation of trees of a graph. In this part of the paper we summarize most. of the existing techniques on tree generation that are based on decomposition and elementary tree and co-tree transformations, and provide a unified theory for their derivation.
In Parts I and II (Chen 1969d, e) we have shown that many of the existing techniques on generation of trees of a graph in the literature, seemingly so different in their appearance, can be formulated in one of the following two ways: the Wang algebra formulation and the matrix formulation. In the present Part III we shall show that the techniques on tree generation, based on decomposition and elementary tree and co-tree transformations, can be deduced from the Wang algebra formulation.
Throughout the paper the symbols n, b, and c are used to denote the numbers of nodes, edges, and components of a graph G, respectively. We denote the rank and the nullity of G by r and m, respectively. Unless specified to the contrary, we assume that G is connected. Definitions of terms used which are not explicitly defined may be found in Part I. [ABSTRACT FROM AUTHOR]

The well-known cyclotron resonance mechanism has been studied and energization of electrons in the magnetosphere has been explored. The relativistic variation of electron mass in the resonance process has been considered and the consequence of neglecting the relativistic electron mass variation has been discussed. At or near resonance the electron motion conforms with the motion of electrons in a potential well. Calculating the shape of the pseudo-potential for the magnetospheric plasma we have calculated the energy gained by the electrons. It is seen that 1 ev electron resonating with the electromagnetic wave (=10 rad sec and E=10 e.s.u.) acquires 1 Mev energy after 188 sec. The geophysical effect of the proposed energization is discussed.
It is well known that resonance of electrons with the electromagnetic waves propagating in the magnetoplasma energize the gyrating electrons. In our earlier paper (Nath and Singh 1970--hereafter referred as NS-I) we discussed the periodic energization of resonating electrons in the magnetosphere. In this paper we relax the requirement in Roberts and Buchsbaum's formulation (1964) and study the synchronous energization of the electrons in the magnetosphere. This is in accordance with the fact that high energy electrons have been detected in the magnetosphere and in the ionosphere. Davydovskii (1963) has suggested that under suitable conditions the charged particles could maintain synchronism with the electromagnetic waves and gain considerable amount of energy. [ABSTRACT FROM AUTHOR]

In this paper we consider how the electric potential varies in space and time during the formation of a plasma sheath. We give the variation of the wall potential in time, the potential profile at various stages of the sheath formation and the variation of the total current at the wall.
In this paper we extend the well-established static theory of plasma sheaths and discuss the growth of a sheath near a plane perfectly absorbing wall. Suppose that we introduce a plane perfectly absorbing sheet of material into a plasma which is electrically neutral and consists of electrons and a single type of singly ionized atoms. The mobile electrons will in general have a much higher velocity than the more massive ions and so initially more electrons than ions will strike the wall, and be absorbed, per unit time. Thus the wall will initially acquire a net negative charge. This charge will repel the electrons approaching the wall from the main body of the plasma and attract the ions. The net charge on the wall will continue to increase until the current density at the wall becomes zero. We wish to investigate the rate of growth of the potential of the wall and the profile of the electric potential at any time on the basis of a simple model. [ABSTRACT FROM AUTHOR]

In some applications it may be desirable to construct an aerial array to radiate minimum energy in a certain angular sector whilst providing high gain in the required direction. In this paper the far field expression for the radiation pattern of a broadside array is reduced to a function of several variables of the inter-element spacing. The element positions are then optimized to correspond to minimum radiation over a specified sector. Experimental results have been obtained in accordance with the theoretical predictions, but difficulty was experienced in achieving small signal measurement near the theoretical minima in the laboratory.
The purpose of the present paper is to show that a similar result is possible in the case of a non-uniformly spaced element array uniformly excited. [ABSTRACT FROM AUTHOR]

Collisions are considered as a drag force in the equation of motion of electrons, and its effect on the wave-particle resonance is shown. Expressions for energy gain and time period of energy oscillations are derived. The role of collisions on the synchronous and oscillatory energy gain and on the time period of energy oscillation for magnetospheric plasma is shown. Energy gain is considerably affected, due to the presence of collisions.
It is well known that electrons resonate with electromagnetic waves propagating through magnetoplasma. In our earlier paper (Nath and Singh 1971--hereafter referred to as NS--I) we discussed the periodic energization of resonating electrons in the magnetosphere. Certain magnetospheric and ionospheric observations have been successfully explained in terms of wave-particle resonance. In NS--I we assumed the collisionless ambient plasma. In this paper we relax the requirement and introduce the effect of collisions in Roberts and Buchsbamn's formulation (1964). This is done in accordance with the fact that most of the gaseous and solid-state plasma are characterized by significant collisional effects. The effect of collisions on wave-particle resonance has been studied in solid-state plasma (Vural and Bloom 1966, and Robinson 1967). In order to interpret some of the magnetospheric and ionospheric results on electron influx pulsations and the co-varying brems-strahlung x-ray flux, we have modified Roberts and Buchsbaum's resonance theory to account for the electron collision frequency. The collisions have been introduced as frictional forces in the equation of motion. With collisions each electron would behave as if it were subjected to a retarding force mvv. It is obvious that the collisions tend to decrease the velocity of the electrons and energy exponentially and eventually limit the energy gain of the resonating electrons. The limit of energy gain arising from the role of collisions is physically more feasible and provides... [ABSTRACT FROM AUTHOR]

This paper attempts to analyze doubly terminated non-symmetrical lattices. The study is restricted to (a) rendering the lattice resistive and (b) the driving-point impedances at the ports to be capacitive and inductive.
In network synthesis the symmetric-lattice structure possesses a main advantage in that any symmetrical realizable two-terminal pair network can be realized as a lattice and, as such, requires only two parameters to characterize such a network. Though some work is done on non-symmetrical lattices (Advani and Reddy 1969, Lee 1963, Miyata 1963), the onus has been on symmetrical lattices.
Hopefully, this paper may lead on to some interesting applications, one such potential application appears to be in the lumped approximation of divergent and convergent exponential lines (Ramachandran 1962). [ABSTRACT FROM AUTHOR]

The phase-space method of calculating electron velocity fluctuations generated by a thermionic cathode is extended to problems encountered in two-dimensional systems. Expressions are derived for volume and current density of the emitted electrons. Numerical results are obtained for the case of a semi-infinite, plane cathode immersed in a constant magnetic field. The two cases of interest, when the field is either parallel or perpendicular to the cathode, are considered separately. A method of reducing the cathode edge effect is discussed in the last section of the paper.
The main object of the paper is to present a method of calculating electron velocity fluctuations and the corresponding volume and current density distributions in systems which are basically two-dimensional in character. [ABSTRACT FROM AUTHOR]

This paper presents a new method of realization of second-degree filters having low-pass, band-pass and high-pass characteristics. Each of the characteristics requires two unity gain non-inverting type amplifiers in addition to passive resistances and capacitances. Details of sensitivity analysis have been reported. This method of realization gives appreciably low Q sensitivity due to the variation of passive elements. But the Q sensitivity due to the variation of the gain of the amplifier becomes a little high. This limits its application in high-Q filters.
This paper gives a new realization procedure for active second-order filters. Each of the band-pass, low-pass and high-pass characteristics requires two unity gain non-inverting amplifiers in conjunction with the passive resistances and capacitances. Details of sensitivity analysis show that the percentage variation of pole Q due to the percentage variation of passive R and C is sufficiently low. But the Q sensitivity due to the variation of amplifier gain is a little high. This dictates that the method will be practically useful in the case of low-Q filters. [ABSTRACT FROM AUTHOR]

This paper describes briefly the ideal characteristics required for tripping for faults within the first zone of a distance relay adapted to the protection of a transmission line. The quadrilateral characteristic on an R-X diagram, which is the ideal characteristic, is obtained using a new hybrid comparison technique. This technique possesses the flexibility of conforming to any desired characteristic by a simple change in the selection of current inputs. The relays are described in their block schematic form.
In this paper the first zone distance relay is considered and the desirable shapes of its characteristics are discussed. A new hybrid comparison technique is used to obtain these desirable characteristics. The comparator models are described in their block schematic form. [ABSTRACT FROM AUTHOR]

Recent studies in non-equilibrium thermodynamics have led to the development of a new general transport theory. Applied to charged systems, this theory makes it possible to separate completely the effects of the processes of diffusion and drift and to calculate the amount of transport work necessary to support a current through each one of such processes, even in the presence of charge-carrier generation or recombination. Thus, considering forward conduction in p-n junctions and heterojunctions to be simply carried by the injection of minority carriers, several non-ohmic terms of transport work have been calculated of which three terms are particularly significant.
This paper explains forward conduction in metal-n Schottky diodes to be caused, like in p-n junctions and heterojunctions, by the injection of minority carriers into the n-type semiconductor. This carrier injection originates from an inversion (p-type) layer which is assumed to be between the metal and n-type semiconductor. Such an inversion layer has been recognized earlier by other workers. Theoretical results, derived by this author in previous papers for p-n junctions and heterojunctions, are found to be accurately in agreement with extensive experimental measurements conducted by several authors on: Au-Si, Pt-Si and Au-GaAs diodes at temperatures ranging from 42 to 433 degrees K. [ABSTRACT FROM AUTHOR]

The electron energy distribution has been calculated as a function of both position and time for an electron swarm travelling across a uniform electric field discharge gap. From these distributions, integral equations have been developed for the discharge parameters representing ionization drift velocity and diffusion and these lead to an alternative form for the drift and diffusion equation.
The magnitude of the discharge parameters have been calculated for electron swarms in molecular hydrogen for 10<E/pо<120 v cm-1 torr-1. A comparison with the values calculated from the usual form of the integrals for these discharge parameters, shows that although the mean energy and rate of ionization are the same, the drift velocity is larger and the diffusion coefficient now behaves as a tensor.
In part I of this paper (Lucas 1967), the spatial/time variation of the mean energies of swarm electrons have been calculated. The results have shown that the mean energies depend upon the longitudinal position within the swarm, with the electrons at the head of the swarm having a higher mean energy than the electrons in the tail. The electrons had to pass through a voltage large compared with their mean energy before the variation of mean energies throughout the swarm was less than 1%. It was concluded that this non-equilibrium state may be represented by adding extra terms to the Boltzmann equation, and that the integral expressions representing the discharge parameters would have to be redefined. All these modifications have been given in this paper, and to illustrate their magnitude, calculations have been made for electron swarms in molecular hydrogen. [ABSTRACT FROM AUTHOR]

In a previous paper (Nandedkar and Bhagavat 1970) a quantum theory of finite d.c. resistivity of plasma was developed. In the present paper the effect of ion-molecule collisions on the noise emittance spectrum due to the electrons is considered. It is shown that the resultant noise power due to the electrons and the ions constituting the spectrum of electromagnetic radiations cancels out everywhere in the space.
In the present paper d.c. noise power generated by the plasma at thermal equilibrium, due to the electrons absorbing d.c. energy, in the presence of ion-molecule collisions is analysed. [ABSTRACT FROM AUTHOR]

In the present paper eigen-frequency of electronic damped oscillations in a plasma is derived. These oscillations are, however, not sustained. Further sustained electronic damped oscillations in the plasma in the presence of an external electric field are analysed.
Eigen-frequency of electronic damped oscillations in a plasma is derived in this paper. It is the intrinsic tendency of damped oscillations in the presence of local fields between charges of the plasma. These oscillations are, however, not sustained as such in practice. But if an electric field E be applied to the plasma model, sustained electronic damped oscillations result which are analysed here. [ABSTRACT FROM AUTHOR]

Theoretical investigation of random phase errors caused by surface irregularities in large reflector antennas has been the subject of many technical papers in recent years. This paper presents experimental results obtained by using a 4 ft diameter model whose reflector consists of an assembly of 54 triangular panels. Accurate control of artificial surface errors is achieved with the aid of micrometers attached to the corners of the panels. Thus the effect of random errors on the radiation characteristics of a large batch of nominally identical antennas can be successfully investigated. Good correlation between theoretical and experimental results has been obtained. The results also show that there is a significant advantage in using triangular panels to build antenna reflectors. [ABSTRACT FROM AUTHOR]

This paper describes the character of neuristor operation composed of a lumped parameter delay line with a negative resistance element parallel to the capacitance, as a function of the precise character of the negative resistance element. This model allows a neuristor pulse transmission in cases with very low characteristic impedance Z= (L/C). The experiments cover a Z value down to 11 . A lower characteristic impedance is possible. The model is analyzed applying Lienard's graphical method and linear approximations. As a result the criteria for characteristic parameters of negative resistance elements to operate as a neuristor are obtained. Based on the results, a neuristor with low inductance, small power, and high-speed operation can be designed. This circuit is convenient for neuristor logic, because, compared with the peculiar pulse width, it has a long delay time per section. The paper also describes neuristor logic without, an R-type junction, using only the threshold property of the neuristor. A conventional Crane's R-junction is also presented. For this neuristor model a destructive collision phenomenon exists independent of the refractory period phenomena. The mechanisms of these phenomena are clarified. [ABSTRACT FROM AUTHOR]

This paper deals with the fundamental properties of the two classes of physical components that exhibit a dual and here are consistently referred to as negative resistors and negative conductors. At the outset a definition of negative resistors and negative conductors is given in terms of essential linear and non-linear properties. Linear models are proposed. The conjoint occurrence of specific linear and non-linear properties is explained using a single-loop feedback circuit. Two postulates applying to active components enable one to determine whether a given network realization is either a negative resistor or a negative conductor. When a negative resistor or a negative conductor is to be used in a specified network, the correct choice between these two components is decisive when system stability is taken as a prerequisite. Some useful rules facilitating this choice are given in this paper.
A unified treatment of negative resuctors has been presented in this paper in an attempt to clear up obscurities and unsolved problems that still exist in connection with the essential nature and the physical limitations of negative resuctors. It has been shown that a consistent distinction between the two dual classes of negative resuctors enables one to draw unambiguous conclusions with regard to the stability problem.
Negative resuctors have been investigated in terms of their fundamental properties, linear as well as non-linear. The Linearity Postulate and the Non-linearity Postulate were each in turn found indispensable to explain various experimental results, of which the conjoint occurrence of specific linear and nonlinear properties is most significant.
The applications, which have been treated as an illustration, have shown that straightforward analysis using appropriate equivalent circuits leads to a better understanding of the potentialities and physical limitations of negative resuctors. [ABSTRACT FROM AUTHOR]

The expressions previously derived by the author for the resistance parameters of the composite transistors are used to investigate theoretically the open circuit characteristics of the compound configuration consisting of two-point contact type transistors. It is found that the system has, in general, a much smaller current gain, voltage gain, input resistance and higher output resistance. The characteristics are greatly modified when the second-point contact type transistor is replaced by a junction type.
Recently the author, in his paper (Singh 1969), has derived the general expressions for the resistance parameters of the composite transistors in terms of the open circuit parameters of the individual transistor. The characteristics of the compound systems in all the three configurations were, however, not theoretically investigated although these have already been discussed by Suen (1968) in terms of the hybrid parameters. Therefore in this paper the author presents the quantitative analysis of the characteristics of the Darlington configuration on the basis of its resistance parameters. [ABSTRACT FROM AUTHOR]

In this paper the recently developed theory of complementary variational principles is extended to include constraints. The theory is applied to electrostatic fields where the constraint corresponds to the specification of the total charge on a conductor.
In a recent paper (Arthurs and Robinson 1969 b), maximum-minimum principles for boundary value problems in electrostatics and magnetostatics were derived from the theory of complementary variational principles. For the simplest electrostatic case, extremum principles were found for the Dirichlet problem:
The purpose of this proper is to derive extremum principles for constraint problems in electrostatics of the form:. [ABSTRACT FROM AUTHOR]

Complementary (maximum and minimum) variational principles are used to obtain upper and lower bounds for capacities in microwave filter problems. The results are illustrated by calculations for a periodic system of cylindrical rods between two parallel ground planes.
In this paper we shall study the boundary value problems which arise when a periodic system of cylindrical rods lies between two parallel ground planes. Such systems are used in certain microwave filters.
The basic problem is to determine the capacitance of the system when either the odd or even mode is excited. The odd mode gives rise to a Dirichlet problem and the even mode gives rise to a mixed boundary value problem. Approximate solutions for the fields in these eases have been obtained by Richmond (1923) using a conformal transformation method, by Chisholm (1956) using a stationary (variational) principle, and by Cristal (1964) using a numerical method for the appropriate integral equation. An alternative approach to the problem is pointed out in this paper. It is based on the theory of maximum-minimum (complementary) variational principles, and leads to both upper and lower bounds for the capacitance. [ABSTRACT FROM AUTHOR]

In previous papers a theory of r.f. conductivity of plasma was discussed. The absorbed r.f. energy in plasma, in the region of backward waves, gives rise to noise in general (Nandodkar and Bhagavat 1969 a, b).
In the present paper a theory of finite d.c. conductivity of plasma is developed. During the process of electronic collisions with molecules in the plasma, when d.c. energy is absorbed, the distribution of electrons is altered. Rearrangement of the electrons with respect to different states of the distribution produces ' quantum jumps ' to give rise to photons which govern noise distribution in accordance with the temperature of the new distribution of electrons obtained. [ABSTRACT FROM AUTHOR]

For a wide, class of planar aperture antennas the aperture electric field is almost completely linearly polarized, the energy in the cross-polarization component being negligible. In the present paper it is shown that all the properties of such antennas can, in principle, be determined from any non-vanishing 'piece' of only one rectangular component of the electric field in the far-field region. And also it is possible to express the integral for the total power radiated in terms of only one rectangular component of the electric field in the far-field region. This has an immediate application in that only one rectangular component of the electric field in the far-field region need be measured in order to determine the total radiated power, knowledge of which is required for the computation of gain and efficiency.
In the present paper the problem of radiation from a planar aperture antenna, the aperture electric field of which is linearly polarized, will be formulated in terms of the pattern space factor. In many antennas the field over the aperture is almost completely linearly polarized, only a small fraction of the energy being in the cross-polarization component of the field (Silver 1949). A planar aperture in the yz plane will be considered. Let the aperture electric field be polarized along the y axis. [ABSTRACT FROM AUTHOR]

This paper is an extension of the recent work of Gupta (1968) in which the properties of a thin resonant antenna in a plasma are analysed using linearized hydrodynamic theory and assuming a general propagation constant for the current distribution, for an antenna having feed points displaced by /2.
In this paper the analysis is extended to include non-resonant antenna and also the dependence of the properties of both types of antennas on the size of the displacement and value of the propagation, constant. The field patterns in the plasma are calculated for two different values of the general propagation constant and for various lengths of the antennas. The results are compared with the free space radiation field strengths which are evaluated by setting the plasma frequency to zero. It is shown that the maximum field strength of both the resonant and the non-resonant types of antenna in free space increases considerably when the feed points are displaced to a certain optimum value. The variation of the magnitude of the field strength and of the radiation resistance on the ratio of the plasma frequency to the source frequency are also investigated. [ABSTRACT FROM AUTHOR]

When a radiating source is immersed in an isotropic compressive plasma, a longitudinal electroacoustic wave may be excited in addition to the usual electromagnetic wave. These waves produced by linear antennas have been studied by the author in a series of papers. The present paper, called part VI of the series, deals with some features of the electric field associated with electroacoustic waves excited by linear antennas. It is shown that the distribution of the field is oscillatory with broadside symmetry and that the maximum occurs almost in the broadside direction.
The investigations presented in the current paper indicate the electric field of electroacoustic waves to be oscillatory having a maximum almost in a broadside direction. About this direction, as expected for the RCFLA, the P mode electric field pattern is symmetrical. But even for the NREFLA there is some symmetry about the broadside direction which improves further with larger values of (c/v)> 10. Thus the P mode field pattern is symmetrical for both types of linear antennas and so differs from the EM mode field pattern which is know to be symmetrical for the RCFLA and non-symmetrical for the NREFLA.
In the end it may be concluded that for frequencies confined to the first Fraunhofer zone, the assumed sinusoidal current distribution with as the propagation constant forms a workable and reasonably simple model to describe the main salient features of linear antenna performance in a hot plasma. [ABSTRACT FROM AUTHOR]

The efficiency with which a complex electrical network may be analyzed by a digital computer using topological formulas depends largely upon the efficiency with which the trees and directed trees of the corresponding graph are generated. The problem of efficient generation of trees of a graph has been treated rather extensively in the literature. In this paper it is shown that many of the existing techniques in the literature, seemingly so different in their appearance, are actually the variants of the Wang algebra formulation. Thus, they can all be unified. The unified theory enables one to summarize many of these results systematically and to provide a simple and easy way for their derivations.
In the paper we have shown that many of the existing results described in the literature on the problem of generation of trees of a graph, seemingly so different in their appearance, are actually variants of the Wang algebra formulation. Thus, it enables one to provide a unified theory for their derivation and to summarize most of these results systematically.
The main advantage of the Wang algebra formulation lies in the fact that once a desired set of sets, which can be obtained easily by inspection, is derived from a given graph all the remaining manipulations are algebraic. It is not necessary for one to go back to the original graph to see whether a generated term forms a tree (co-tree) or not. However, the techniques are not very efficient for large graphs because duplicated terms are generated during the process. Apart from slowing down the process of generating trees, this limits the sizes of the graphs that can be analyzed since one must retain the set of all trees in computer memory and check each new tree against the list for possible duplication.
The application of the Wang algebra technique to the generation of directed trees and multi-trees has been considered by Chen (1966). A special case of which is recently given by Shinoda (1968). [ABSTRACT FROM AUTHOR]

In the previous papers (Bhagavat and Nandedkar 1988, Nandedkar and Bhagavat 1969) investigation of r.f. energy transport in the vicinity of damped oscillations in plasma, where absorption of r.f. energy is maximum, is carried out. Backward waves are shown to exist in the region of anomalous dispersion of the plasma. The absorbed r.f. energy in the region of backward waves appears as noise.
In the present paper, wavefront and signal velocities of a r.f. wave in plasma, in the region of backward waves are analyzed
Whatever may be the frequency of a r.f. wave, the velocity of wavefront in plasma, is always equal to the velocity of a T.E.M. wave in free space (eqn. (19)). Thus the wavefront velocity in the region of backward waves is equal to the wave velocity in free space. When the wavefront of the signal makes its way through the plasma, it finds the electrons which are capable of oscillating originally at rest. Originally, therefore, the plasma seems to be free space for the electromagnetic field propagation; only after the electrons are set into motion, can influence the phase and form of the electromagnetic wave. The propagation of wavefront thus proceeds undisturbed with the velocity of a T.E.M. wave in free space; independently of the character of oscillating electrons.
Figures 13 anti 14 show that the signal velocity for < increases with the increase of the frequency of the r.f. wave, in the region of backward waves (for backward wave region, refer to Bhagavat and Nandedkar 1968e). In this portion of the region v is less than c. At = the signal velocity becomes equal to the velocity of a T.E.M. wave in free space. For > the signal velocity decreases with increase of frequency; in this portion of the region, v is again smaller than c. The left half portion of the region of backward waves with respect to is a mirror image of the right half portion. One thing... [ABSTRACT FROM AUTHOR]

In previous papers (Bhagavat and Nandedkar 1968 b, c, d, e) the investigations of the damped oscillations, r.f. energy transport and backward waves in a plasma in the region of anomalous dispersion were carried out.
In this paper the absorption of energy and power of a r.f. wave in the same region is analysed. Permittivity of the plasma is found to decrease with increase of frequency in the absorption band. The absorption is maximum at the resonance. [ABSTRACT FROM AUTHOR]

The paper deals with the phenomenon of microwave tunnelling through a finite length waveguide filled by plasma in cut-off conditions. The transmission coefficient of the plasma region is calculated and on this basis a method for the determination of the free-electron density of the plasma is derived. This method is applied in the case of a cold cathode discharge with a double plane geometry inside an X-band waveguide for the experimental study of the plasma density as a function of the discharge current.
In this paper possibilities of extending the range of measurable electron densities above the cut-off value are analyzed by the study of microwave propagation along a waveguide of finite length filled by plasma. A fraction from the incident power can be found behind this region, due to a phenomenon of tunnelling. [ABSTRACT FROM AUTHOR]

In previous papers (Bhagavat and Nandedkar 1968 a, b, c, d, e, Nandedkar and Bhagavat 1969) the investigations of the damped oscillations, r.f. energy transport, backward waves and absorption of a r.f. wave in a plasma in the region of anomalous dispersion were made.
In the present paper an analysis of the absorbed r.f. wave in the plasma in the region of anomalous dispersion is carried out. The absorbed r.f. wave exhibits itself as noise. The peak of noise spectrum lies in the far infra-red region of the spectrum of electromagnetic waves. [ABSTRACT FROM AUTHOR]

This paper describes an investigation into the electrical behaviour of shallow implantations of boron and phosphorus in silicon. The clover-leaf Van der Pauw patterns were defined by planar masking techniques, and were ion implanted to produce a 'tailored' impurity profile of approximately uniform concentration with depth.
It was found that phosphorus implantations became fully electrically active after annealing at a temperature between 620degreesc and 750degreesc, whereas the boron implantations required an anneal in excess of 900degreesc. A subsidiary activity peak, whose magnitude increased with the ion dose, was observed for the boron implantations after annealing at a temperature around 520degreesc.
The principles of ion implantation have been discussed at length by other authors, for example Gibbons (1968), Kleinfelder (1967) and Nielsen (1967), and recently Large (1969) has described the techniques used at S.E.R.L. During the past two years, papers have been published showing the electrical properties of silicon implanted with various ion species, but very little information has appeared on the electrical behaviour of shallow implantations of boron and phosphorus. The present work was initiated to measure these properties in order that the electrical behaviour of ion-implanted silicon devices be more fully understood. [ABSTRACT FROM AUTHOR]

This paper treats the problem of finding the electric field generated within a rectangular cross section single turn conductor. Both the cases of infinite and finite conductivity are treated and an exact series solution valid in the limit of negligible displacement current is given.
The determination of the electric fields generated by single turn conductors encompassing various geometries is the kernel of an interesting set of electromagnetic problems.
The recent increased interest for azimuthally symmetric containment geometries in the field of controlled fusion research, more precisely marked by the development of Spherators, symmetrical Stellarators and toroidal theta pinch devices, calls for solutions to these problems.
In this paper denoted 'I' we derive and present analytical solutions pertaining to toroids with a rectangular cross section and with feed slits parallel to a major circle of the toroid.
The results are presented both in the form of an exact converging series and as digital computer derived plots and tables. [ABSTRACT FROM AUTHOR]

Synthesis of multi-port networks is attracting increasing attention in network theory. On the other hand comparatively little work has Been reported on active RC multi-port transfer function synthesis. This paper presents an elegant synthesis procedure to realize any arbitrary rational m n open-circuit voltage-transfer function matrix where m and n are, respectively, the numbers of the input and output ports. The realization results in a grounded network requiring (2m+1)n RC one-ports and (m+1)n grounded operational amplifiers and hence is attractive from the practical point of view.
The present paper discusses the realization of an arbitrary rational m n open-circuit voltage-transfer matrix where m is the number of inputs and n is the number of outputs. This type of network is of practical importance in systems where simultaneously more than one input and output are required. [ABSTRACT FROM AUTHOR]

This paper considers the sensitivity of the power gain of a tuned conjugate-matched amplifier to changes in active device parameters, for the general case wherein the active device may be embedded prior to termination. Prediction of the effect of simultaneous spreads in the eight parameters of the active device is based upon a simple expression--here also displayed as a three-dimensional model for ease of insight--relating conjugate-matched gain to two device parameters. These parameters are representative of device activity and non-reciprocity, and are separately invariant to common types of embedding. A new basis is proposed for the characterization of spread between two devices, each represented by a point in a three-dimensional space, and leads to a simple description of the effect of device spread upon power gain. The effect on gain spread of common types of device embedding is discussed and its simple description verified experimentally. A prerequisite to the treatment of batch spread is the realization of an 'average device', and is effected by standard statistical techniques. For the minimization of gain spread in amplifiers affected by batch spread in active device parameters, a heuristic design strategy for the best choice of device embedding is proposed.
This paper is concerned with the above problem as it arises with an active two-port network designed to exhibit useful power gain at a single frequency. It is, therefore, particularly pertinent to the design of narrow band amplifiers. Indeed, the illustrative numerical examples refer to a transistor tuned amplifier, although the underlying theory is valid for any three-terminal active device. It is the object of the paper to consider how an active device might best be connected between source and load terminations--but using an intervening embedding network if necessary--in order to realize a useful power gain while minimizing its spread. [ABSTRACT FROM AUTHOR]

This paper presents the graphical analysis and design considerations of multi-input sine comparators whose mathematical analysis has already been presented in an earlier paper by the authors (Anti Kumar and Parthasarathy 1967). Particular emphasis is placed on the design of a circuit, using semiconductor components, which provides a quadrilateral polar characteristic on the impedance plane. Novelties of design include a reactance reach, increased tolerance to fault resistance as the fault point moves towards the relay location, directional feature and independent control of each line segment constituting the quadrilateral, so that the fault area and tho relay characteristic could be closely matched. Test results show the negligible transient overreach and a constant operating time which is 1 1/2 cycles or less, and irrespective of fault location. The flexibility of design enables it to be used as a three-zone stepped time-distance relay or in conjunction with pilot schemes.
Test results show the versatility of the unit and the flexibility of the comparator theory. By suitable modifications of only the input signals, any desired characteristics can be easily obtained as dictated by system requirements. Special care has been taken to use similar switching blocks so that replacement is rendered simple. [ABSTRACT FROM AUTHOR]

In the previous papers (Bhagavat and Nandedkar 1968 b, c, d) the investigations of the damped oscillations and velocity of energy transport of a r.f. wave in a plasma in the region of anomalous dispersion were made.
This paper covers the analysis of phase and group velocities of a r.f. wave in this region. This analysis indicates the existence of backward waves which are experimentally verified. [ABSTRACT FROM AUTHOR]

This paper describes the development of a static comparator which responds to asynchronous conditions in a power system, by measuring the mean angular velocity between the e.m.f.'s of the system generators. The comparator initiates reestablishment of synchronism by disconnecting a selected number of machines or by effecting out of step tripping of a selected link in the power system, only when the measured angular velocity exceeds the critical angular velocity of the electro transmission.
This paper describes the development of a static comparator which responds to asynchronous conditions in a power system by measuring the mean angular velocity between the e.m.f.'s of the system generators. The comparator makes use of logic gates with transistor circuits which have proved to be very reliable for use in the design of control systems. [ABSTRACT FROM AUTHOR]

A rigorous solution has been given for the growth of pre-breakdown current in a uniform electric field discharge gap of finite dimensions. The electron current is generated by both ionizations in tho gas and the production of secondary electrons at the cathode due to both positive ions and non-resonance photons.
The previous approximate solution to this problem (Lucas 1965) is shown to be in very good agreement with the rigorous solution, and is of sufficient accuracy to be used for the analysis of pre-breakdown currents, experimentally measured by the Townsend method, to yield values for the ionization coefficients.
The approximate equation given in paper I to represent the growth of pre-breakdown current duc to electron ionization and the production of secondary ionization by both non-resonance radiation has been shown to represent the theoretical growth of pre-breakdown current to a high degree of accuracy. Modifications to this equation have been given in this paper to include both a distributed primary electron source released from the cathode and values of d/b 1. [ABSTRACT FROM AUTHOR]

In a previous paper (Bhagavat and Nandedkar 1968) an analysis of the frequency of damped oscillations in the presence of a r.f. field in a plasma model in equilibrium was made.
This paper makes a complete analysis of this subject in the presence of a d.c. field. An experimental investigation of the variation of with the electron density and a d.c. field is described. [ABSTRACT FROM AUTHOR]