Showing papers in "Journal of the Institution of Electrical Engineers - Part III: Radio and Communication Engineering in 1948"

The Theoretical Precision with which an Arbitrary Radiation-Pattern may be Obtained from a Source of Finite Size

Abstract: It appears that it is possible to approximate as closely as desired to a specified radiation-pattern by a suitable distribution of field over an aperture of given size, though the necessary currents in the conducting elements of the source would in general be prohibitively large in comparison with the power radiated. The difficulty of obtaining a high degree of approximation, and in particular a power gain very much greater than that of a uniformly illuminated aperture, is thus a practical rather than a theoretical one. The same is true for the linear array of given length as for the continuous aperture if no limit is set to the number of elements. Even when this number is limited by the adoption of half-wavelength spacing, the broadside power gain is not a maximum when the amplitudes and phases of the elements are equal, unless the elements are ideal isotropic point-sources.

Forced oscillations in nearly sinusoidal systems

Abstract: A large class of radio circuits which are analytically equivalent to an oscillatory network in parallel with a non-linear negative resistance, are represented fairly accurately by the differential equation v? ? (?+sv??v2)v + ?2v = E?21 sin?1t where, ?/?, s/?, ?/? are small. The behaviour of the solutions of this equation near resonance has been discussed by Appleton, van der Pol and others. The paper contains a more complete discussion of the synchronized and quasi-periodic solutions near resonance, their phases, amplitudes and energy, and also the way in which one type of stable solution gives way to another as the parameters of the system vary, for instance as the electromotive force or detuning vary. It is shown that the phase and amplitude favourable to synchronization are prolonged just before synchronization. This agrees with Appleton's experimental results. It is also found that hysteresis occurs. The decrease in energy with the decrease in detuning is explained by the fact that the phase favourable to synchronization is that which opposes the motion and is prolonged.

A table of intermodulation products

Abstract: An expression is derived from which, by substitution of appropriate numerical values, an equation can be obtained for any of the intermodulation products which can be generated by the simultaneous transmission of any number of sinusoidal waves through a device with an output/input characteristic like V=a1v+a2v2+a3v3…. This expression is used as a basis for classification of products, and a table of representative products is drawn up. Information is given about the numbers and relative importance of products of different kinds.

Three-dimensional cathode-ray tube displays

Abstract: The display of three-dimensional information is considered with particular reference to a radar system in which a narrow pulsed beam of radio-frequency energy is used to explore automatically a volume of space. The three-dimensional displays are used to display the positions of the reflected ?signals? with respect to the three co-ordinates of the volume, in order that the radar can be used as an object-detecting and -locating system in three dimensions. In the past, ?configurational? methods have been used, wherein the shape of the intensity-modulated echo-trace on the cathode-ray tube is altered to add a third variable to the two already provided by the c.r.t. deflections. These types are not discussed, attention being concentrated on other methods. The ?truly three-dimensional displays? utilize a mechanical motion to add a third dimension to the screen of the cathode-ray tube, the picture appearing as an actual volume. The ?perspective displays? simulate the same picture as in the previous display without any moving mechanisms, rotation of the volume and/or stereoscopy being used to give the impression of three dimensions. The ?oblique displays? obtain the required three co-ordinates by combinations of two or more two-co-ordinate pictures on the same cathode-ray tube. The ?polychromatic displays? use colour to represent a spatial co-ordinate. The physiological and psychological problems of the human operator peculiar to three-dimensional displays are discussed, and a Section is included on the relative advantages of ?true? and ?relative? displays. Another Section on ?display sensitivity? leads to the general theoretical considerations involved in three-dimensional displays. The application of three-dimensional displays to other fields, for example X-ray photography, is briefly discussed. The revised form of the radar equation for use with three-dimensional displays is given in an Appendix.

Sketch for an algebra of relay and contactor circuits

Abstract: A system of symbols is described, by which the various elements (contacts and operating coils) of a relay circuit can be represented. The symbols are then linked together by signs, the resultant algebraic expression corresponding to the circuit, and the operation of the circuit can be studied by considering the behaviour of the algebraic expression according to rules which are developed in the paper. Methods are described whereby complex circuits may be simplified by algebraic manipulation of the symbolic expressions; and design of relay circuits to perform given operations is discussed.

Matrix Methods in the Solution of Ladder Networks

Abstract: The matrix algebra has a natural application to electric circuit theory, especially to the method of solution by means of substituted variables, such as symmetrical components. The solution of circuit equations by the method of diagonalizing the impedance matrix is developed in the paper, and applied to the solution of the difference equations of ladder networks. An extension to systems of difference equations follows. Diagonalizing is equivalent to the use of substituted currents, voltages and impedances such that there is no mutual coupling between the substituted networks. The substituted currents are then calculated without difficulty and the original circuit currents obtained by a simple transformation. The method has the additional advantage that, since the transformation is independent of the circuit constants, it may be applied to all circuits possessing the same degree of symmetry. Ladder networks may also be solved by regarding them as a series of four-terminal passive networks. The elements of matrix algebra are included for completeness.

Input impedances of centre-fed slot aerials near half-wave resonance

Abstract: Slot radiators of various widths were fed via balanced parallel-rod transmission lines connected near the centre points of opposite edges of the slots. Series components of impedance at the feed points were measured for each slot at a number of frequencies, including that at which the length of the slot was ½λ. In the first set of experiments the slot was free to radiate on both sides of the metal sheet in which it was cut. Further measurements are described in which the dimensions of the slot were kept constant whilst rectangular cavities of various dimensions were placed symmetrically behind the slot so as to prevent radiation from one face of the metal sheet.

Some effects of obstacles on the propagation of very short radio waves

Abstract: The paper deals with effects of localized obstacles, including reflection as well as shadowing, which are of practical importance in radio communication and radar, particularly on decimetre and centimetre wavelengths. In a digest of available theoretical results, later compared with the experimental findings, attention is drawn to some unresolved difficulties. In most of the practical problems discussed, prediction to a useful approximation is, however, shown to be possible.The experimental work described has been carried out by or for the Admiralty, and much of it had its origin in the special problems of naval communications and radar. It has, however, wider applications at almost every point; applications to radar navigation are specifically discussed.

Some aspects of the design of balanced rectifier modulators for precision applications

Abstract: The design of balanced rectifier modulators of ring and Cowan types is discussed, and it is shown how the performance of the modulator depends on rectifier characteristics, the circuit impedance in which the modulator operates, the resistance of the carrier generator and the carrier voltage.The main performance features discussed are efficiency, stability, production of unwanted modulation products, impedance, and carrier leak.It is shown that there are advantages in designing a modulator for maximum efficiency by choosing a circuit impedance of optimum value and a carrier supply of high resistance. Variations in efficiency due to variations in carrier voltage, circuit impedance and temperature are then a minimum although it may be necessary to have each rectifier shunted with a constant resistance to achieve this. The production of modulation products of higher order is greatest in a modulator of maximum efficiency in which the circuit providing the carrier is of high resistance, and is reduced considerably if inefficient working and a low-resistance circuit are used. Further reduction, if desired, can be obtained by the use of a large bias on the rectifiers, and it has been found possible to reduce the 3fc±f sidebands to over 26 db below the fc + f sidebands (where fc = carrier frequency and f = signal frequency). This method is of value only in ring modulators, as with the Cowan circuit it introduces products of the type 2fc + f.It is shown that the ordinary potentiometer adjustment for carrier leak compensates only the unbalances between the low forward resistances of the rectifiers, and there remains an out-of-balance current at small values of the instantaneous carrier voltage. This effect can be reduced by using a high-resistance carrier supply and by appropriate selection of the rectifiers; a further improvement, in the case of ring modulators, can be obtained by the use of large biases on the rectifiers. Carrier leaks as low as − 60 db relative to 1 volt in 600 ohms, or 40 db below sideband-level, when high-level signals are applied, can readily be obtained and maintained in modulators with at least 1 volt carrier peak voltage across the rectifiers.It is assumed throughout that the rectifiers are purely resistive, as with the diodes and crystal valves now available, the effect of the capacitance is negligible up to frequencies of the order of 1 Mc/s.A method of design of a ring modulator is described in which the input impedance remains relatively constant over the cycle of carrier voltage; this may have important applications in precision circuits.

Space-charge effects in beam tetrodes and other valves

Abstract: The space charge in the screen-anode space of a beam tetrode is examined, taking into account the effect of electrons returned from the virtual cathode into the region between the grid and cathode. As with previous investigations which do not take into account the effect of the returned electrons in the cathode-grid space, it is found that over a range of anode voltages just below the knee, three solutions of the problem satisfy Newton's laws of motion for the electrons and also Poisson's equation for the charge distributions. Two of these solutions give rise to discontinuities in the characteristic. The third, however, is continuous for all anode voltages from zero up to the knee, and corresponds closely to experiments on actual valves. In it, a virtual cathode which remains at a fixed point between the screen and the anode is set up. The question arises, What property of this solution determines its correspondence with experiment? It is found that Maupertuis's principle of least action, as an integral of momentum with respect to distance, can be applied. The principle of least action as it stands cannot, however, be applied to a single arbitrarily chosen electron, but by calculating the action per unit of time, it can be applied to all three solutions, whether electrons are reflected or not. This leads to the conclusion that the action per unit of time is least for the third solution. Since this corresponds to the conditions observed experimentally, it is asserted that, in problems concerned with the motion of charged particles in which more than one consistent solution can be found, the actual state set up will be that for which the action per second is least.A simple theory of the effect of perturbations of the space-charge density, due, for example, to the random fluctuations of the cathode emission, is developed. The rate of growth of a perturbation depends on its size, and if it is small, it is unlikely to grow to such an extent, before being swept to the anode, as to cause instability of a particular charge distribution. It is concluded that such perturbations do not form an alternative explanation of the fact that the state containing the virtual cathode is always that set up.The growth of the perturbations is shown to depend on whether the anode potential remains constant or can vary during the passage of a perturbation. It is concluded that the theory of the fluctuations of a space-charge-limited current cannot be dealt with accurately on the assumption that the fluctuation current is independent of the anode load.Finally, the fluctuations of the space-charge-limited current in diodes and triodes is described on the basis of the given theory of perturbations. For identical space-charge conditions near the cathode, the diode has a planar electric field which exercises no electron-optical action, while the triode has a strong focusing action in the space between grid and cathode. The conclusion, that the perturbations can grow to a greater extent in diodes than in triodes, agrees with the experimental results on the noise of diode and triode valves.

An anti-clutter radar receiver

Abstract: One of the major limitations to the efficacy of radar is the presence of responses from land, sea or rain-storms in the vicinity of a target, such responses being generally known as "clutter." If the clutter is of amplitude greater than the target-echo, detection of the echo is very difficult; in many cases, however, the clutter is of amplitude smaller than the target-echo, but the latter is lost owing to saturation of the receiver or display equipment by the clutter. It is shown, both theoretically and practically, that such loss can be prevented in the cases of sea and rain clutter by the use of a receiver, the amplification characteristic of which is such that the amplitude of an output signal is proportional to the logarithm of the amplitude of the corresponding input signal; in the case of land clutter, loss is not necessarily prevented, but the probability of such loss is reduced. A logarithmic receiver compresses input signals of any strength less than 100 db above the mean noise power, into output signals of voltage less than four times the output-noise deviation voltage, with negligible change in the detectability of small signals compared with the use of a linear receiver. In addition, with such a receiver, the difference in voltage of two output signals depends only on the ratio of the corresponding input signals, and not on their absolute amplitudes. These properties make a logarithmic receiver superior to a linear receiver for a number of subsidiary applications, some examples being given in the paper.

Field distributions near a centre-fed half-wave radiating slot

Abstract: The paper describes the methods and results of experiments made to map out the electromagnetic fields associated with a slot radiator. The radiator was a narrow slot about half a wavelength long cut in a large circular metal sheet of diameter about 9 wavelengths. The results include the relative magnitudes and phases of the electric vector over a quadrant of the metal sheet close to the slot, the relative magnitudes of the magnetic vector along the axes of the slot, and also some polar diagrams of radiation at relatively large distances from the slot, both on and off the sheet. The results obtained are correlated as well as possible with the theoretical expressions for the fields associated with a half-wave slot of infinitesimal width in a sheet of infinite extent. These theoretical expressions are derived from the usual formulae for the fields of a thin dipole by the use of an extension of Babinet's principle recently described by Booker.

The measurement of errors in radiogoniometers at high and very high frequencies

Abstract: A review of existing methods of measurement of the instrumental errors in inductive radiogoniometers is followed by detailed descriptions of three methods that have been used in the high- and very-high-frequency bands (3?100 Mc/s). The first, a field-coil interconnection method which requires no apparatus other than a signal generator and detector, has a limited application, but this disadvantage is offset by its simplicity. The second method makes use of a resistance potentiometer capable of a high order of accuracy at frequencies up to 30 Mc/s. In the third method two inductive piston attenuators are used. This apparatus has a very high order of accuracy, and may be used at the highest frequencies now employed for direction-finding purposes. The importance of the measurement of errors due to the electrical asymmetry of the field coils and search coil is emphasized, and methods that have been used at high and very high frequencies are described. The accuracy and limitations of the methods are discussed and their merits compared. Details of the construction of the measuring apparatus and of its calibration are given, and typical error measurements made on various goniometers are included.

Some methods for determining the power gain of microwave aerials

Abstract: Four methods of determining the absolute power gain of microwave aerials are described, and some indication of their limitations and advantages is given. Two of them require no knowledge of the aerial geometry or polar diagram, and can therefore be used for finding the gain of any type of aerial. The other two, however, require a knowledge of the field across the aerial aperture, and can only readily be applied to horns of rectangular cross-section; even then several approximations have to be made. A method for comparing other aerials with the standard is also described.

Circle diagrams of impedance or admittance for four-terminal networks

Abstract: The paper offers a theory of circle diagrams of impedance or admittance representing the relations between terminal impedances and input impedances for the general passive four-terminal network, including the cases of asymmetry and dissipation, so formulated that common transmission-line diagrams become special cases.After a review of earlier work on the subject, the relations between terminal and input impedances are expressed in terms of the iterative quantities of the network in a circle diagram. It is shown that this diagram is actually a graphical representation of the complex hyperbolic tangent function. Special circle diagrams are discussed corresponding to a non-dissipative network in the stop-band or pass-band, or at the cut-off frequency. These diagrams are studied by the application of projective geometry.An inversion of the ordinary circle diagram leads to the Smith diagram, which is described for both real and complex iterative impedances. It is pointed out that for a transmission line of less than ¼λ the error resulting from considering Z a pure resistance is of the same order of magnitude as that which would result from neglecting the damping.

Some sources of error in microwave milliwattmeters

Abstract: An approximate analysis is made of the factors that influence the intrinsic accuracy of a heated filament used as a transfer device for measuring power at microwavelengths in terms of a low-frequency calibration. It is shown that instruments incorporating such filaments are usually subject to errors of significant magnitude if the filaments are longer than about one-tenth of a wavelength, and that the errors are usually smaller in resistance milliwattmeters than in thermocouple instruments used in similar conditions. Experiments are described which show that serious errors, due to the inefficiency of the transformer used to match the power source to the filament of the instrument, can occur in addition to those described in the analysis. It is concluded that filament instruments cannot, in the present state of development, be relied on as standard transfer instruments for measuring microwave power in terms of a low-frequency calibration.

The design and use of resonant cavity wavemeters for spectrum measurements of pulsed transmitters at wavelengths near 10 cm

Abstract: Resonant cavities have been used to measure the frequency and check the performance of radar transmitters. The object of this paper is to describe the factors influencing the design of such cavities to cover frequency bands near 3000 Mc/s, and to estimate their accuracy in the measurement of both the mean frequency of the transmitter output and its spectral distribution about the mean. The principles described are of general application to the design and use of cavity wavemeters of high discrimination.

Radiogoniometers for high- and very-high-frequency direction-finding

Abstract: The paper is concerned with the design and performance of inductive radiogoniometers for use in d.f. systems operating in the frequency band 3?100 Mc/s. The general principles of design are discussed, and it is shown that the main considerations are the coupling law, the electrical symmetry of the field coils and the coupling factor. The principal component of the instrumental error is the coupling error, and two methods of reducing it are described. One involves the use of a distributed search-coil winding, and the other the use of a compound-wound type of search coil. It is shown that, if the turns of the search-coil winding are wound on a cylindrical former so that they cover the surface completely, the coupling error is negligible, irrespective of the form of the field coils. Four h.f. instruments designed on this principle are described, and details of their performance are given. In the compound-wound search coil the winding is in two sections, the planes of which are set at an angle to each other. It is shown that the coupling error is dependent upon this angular separation and upon the configuration of the magnetic field. An investigation into the design of the field coils indicates how the most suitable field configurations may be obtained. The design and construction of a v.h.f. goniometer embodying these principles is described, and its measured performance provides confirmation of the theory of operation. The importance of electrical symmetry of the field coils is pointed out, and the ways in which it is affected by the design of the goniometer are discussed. Comparison between goniometers of the two types shows that, in general, for medium and high frequencies the distributed search-coil winding is the more suitable, and for v.h.f. operation that with the compound-wound search coil. In an appendix the operation of the inductive goniometer as a phase shifter is considered theoretically, and it is shown that it has the same instrumental error when used in this manner as when used as an amplitude comparator in a direction-finder.

Impulsive interference in amplitude-modulation receivers

Abstract: An analysis is made of the response of an a.m. receiver to impulsive interference (of the type generated by car ignition systems) using the Fourier integral theorem. Conclusions concerning the general behaviour of receivers are deduced from the formulae developed. The method is then applied to the calculation of the performance of noise suppressor circuits which are classified as amplitude, differential and delay limiters. Simple expressions are derived for estimating the suppression attainable in any particular case.Some experiments are described in which the performance of the first two types of noise suppressor was measured. There is shown to be fair agreement with the calculated values.A comparison is made between the behaviour of frequency modulated, pulse-length-modulated and suppressed amplitude-modulated systems with impulsive interference.

Some Principles underlying the Design of Aerial Systems for High-Frequency Radio Direction-Finders in H.M. Ships

Abstract: The paper deals with the design of direction-finding aerial systems to cover the whole of the hf band, to work on the ground wave and to meet the conditions peculiar to HM ships The special problems encountered are intimately connected with the siting of the df aerial The aerial must be fitted to the top of a mast, and often at a considerable distance from the receiving compartmentThe most satisfactory design of df aerial to meet these conditions consists of two fixed crossed loops with an incorporated sense aerial, and a network for injecting test signals into the df and sense-aerial circuits The aerial system must be connected to the receiving apparatus by means of transmission lines which may be of considerable lengthThe paper describes the special problems experienced in achieving satisfactory performance in sensitivity, bearing accuracy and resolution of sense ambiguity, the compromises in design that have to be adopted and the performance figures obtained

Electron Optics and Space Charge in Simple Emission Systems with Circular Symmetry

Abstract: Properties of a simple type of emission system, such as is used in electron guns of cathode-ray tubes, have been investigated. It was found that the properties depend largely on the spacings between the electrodes of the system. For any grid-anode spacing, there exists a certain grid-cathode spacing referred to as the preference spacing, for which the current density in the focused beam is a maximum. The divergence of the emitted electron beam and the spherical aberration in it both have minimum values at this spacing, near which the crossover of the rays in the system occurs at a maximum distance from the cathode and in general the grid base is least. Some physical explanations of these effects are put forward. The presence of ions produced by traces of residual gas modifies the values of the preference spacings and tends to enhance the effects. From the results given it is seen that a direct application of first-order optical laws is not adequate to describe the crossover in an electron gun.

The design and operation of high-power broadcast transmitter units with their outputs combined in parallel

Abstract: The paper describes a method, developed by the British Broadcasting Corporation, of obtaining very large r.f. output power for broadcasting purposes by combining the outputs of two or more similar transmitter units. The principle is analogous to paralleling the outputs of a number of a.c. generators, with the difference that, in the case of transmitters, the frequency of the separate units is identical, the variables being phase and loading. The method of ensuring equality of the r.f. output voltage and modulation amplitude of the separate transmitter units is described. The circuits for combining and matching a variable number of transmitters to a common load are detailed, and a number of special aspects relating to the design of transmitters for parallel operation, including the principle of ?drive suppression? for protection purposes, are discussed. The design of the circuits and operational procedure developed for the Droitwich high-power medium-wave station, which has a power output to the aerial of 400 kW carrier provided by two 200-kW class B modulated transmitters in parallel, are described together with operating experience. Mention is made of a second station, the Ottringham high-power long-wave station of the B.B.C. European service, in which an aerial power of 800 kW was obtained by means of four similar 200-kW transmitters operated in parallel, and which will be fully described in another paper. Appendices contain some data on the design of the modulated-amplifier output circuit of the transmitters used, and a description of an r.f. impedance monitor for fault protection on the high-power r.f. chain.

The response of a linear diode-voltmeter to single and recurrent r.f. impulses of various shapes

Abstract: An analysis is given of the response of a linear diode-voltmeter to r.f. impulses, of the indication of a meter measuring the rectified voltage, and of the damping effect of the diode on the preceding circuit.The response to a short single impulse of arbitrary shape is evaluated in terms of the time-integral or area of its envelope. The ratio of this area to the peak amplitude is termed the effective duration of the impulse.The response to recurrent impulses whose spacing is small compared with the discharge time-constant is expressed in a general manner and graphed for the cases of rectangular and triangular impulses. The “effective charge time-constant” of the voltmeter depends on impulse and circuit parameters. The rectification efficiency for triangular impulses is found to lie between 0.75 and 1 times that for rectangular impulses of the same effective duration and periodicity. It is inferred that for most practical impulses the rectification efficiency will lie between the values shown graphically for triangular and rectangular impulses when expressed in terms of effective duration of the impulse.The maximum indication of a critically damped meter to the rectified voltage produced by the application of a single impulse is evaluated in terms of the ratio of the meter and discharge time-constants.The damping imposed by the diode rectifier on a sharply tuned circuit is calculated for single and recurrent impulses and shown to be considerably greater than for continuous waves.

The measurement of peak voltage at a frequency of 600 Mc/s by means of a modified probe circuit

Abstract: If the probe current of a coaxial-line standing-wave detector is passed through a pair of rectifiers, a simple relation may be obtained between the mean current through either of the rectifiers, the capacitance of the probe to the inner conductor, and the frequency and peak amplitude of the voltage on the line. At a frequency of 600 Mc/s, errors are introduced into voltage measurements made with this circuit, owing mainly to the shunting of the rectifiers by stray capacitance. These errors may be reduced to 5?10% by using low-resistance silicon crystal rectifiers at their maximum current rating, and the magnitude of the errors can be deduced from a simple test made at low frequency. The shunting effects of stray capacitance across the rectifiers may be compensated, for to a large extent by a variable inductive stub, tuned to resonate with the stray capacitance at the operating frequency. The compensation is not perfect, however, and the use of a stub is a disadvantage with large rectifier currents. Measurements of 600-Mc/s pulsed voltages made with this probe circuit were found to agree, within the limits of experimental error (±3%), with the corresponding values deduced from measurements of the power dissipated in a calorimeter which terminated the coaxial line.

An Improved Ionospheric Height Recorder

Abstract: Automatic equipment is described for measuring the virtual height of the reflecting regions of the ionosphere. The system comprises a transmitter, which scans a wide frequency range some twenty-five times per second, and a receiver which is tuned slowly through the same range. Each time the frequency of the transmitter passes through the pass-band of the receiver, a pulse is produced which is displayed on an oscillograph, together with other pulses produced by waves reflected from the ionosphere. The advantages of this technique are enumerated. The equipment enables a record, covering the range 2?16 Mc/s, to be obtained in a few seconds. The apparatus is fully automatic and produces records as required without manual operation. The equipment has been designed with a view to duplication, is robust and easy to service, and is semi-portable.