摘要:

The precision with which standards of inductance can be measured and used at radio frequencies is limited, not only by the ordinary uncertainties of experimental technique, but also by the fact that the standards are necessarily circuit elements with terminations, while inductance is primarily a characteristic of complete circuits. The procedure followed at the National Physical Laboratory when the highest possible precision is required is described in both its experimental and theoretical aspects. The precise relation between the familiar equivalent network for a coil and the basic definitions is indicated, and practical details are given for the resonance method of measurement that has been adopted as standard practice for all work of this kind. It is successfully operated at all frequencies from the audio range, where it overlaps the bridge methods employed as standard practice in the lower ranges of frequency, to values little short of those of self-resonance of the coils. In the overlapping region the bridge and resonance methods agree to about 1 part in 104and with suitable elaboration the resonance method gives about the same accuracy for coils of any value between 1 H and 10 μH at any frequency within their working ranges, the limit for the smaller coils being about 4 Mc/s. With coils of the order of 1 μH measured values may show a standard deviation as small as 0.0001 μH, but from the general considerations outlined in the paper limits closer than 0.0005 μH will seldom be significant in any work with lumped circuits.

DOI:10.1049/pi-b-1.1956.0184

出版商:IEE    

年代:1956

数据来源: IET

摘要:

A review is given of the principal materials used for the construction of resistance standards. The difficulty of producing manganin and constantan commercially with the requisite small value of temperature coefficient at room temperature makes attractive the newer alloys, whose temperature coefficients can be controlled by simple heat treatment. One of these alloys, having the additional advantage of a resistivity three times that of manganin, has been studied at the National Physical Laboratory. It is composed of nickel, chromium, aluminium and copper and is known commercially as Evanohm. When the temperature coefficient, at a given temperature, has been reduced to zero by heat treatment, the curvature of the resistance/temperature characteristic is only one-tenth that of manganin. The stability of resistors constructed of this material has been investigated and has been generally found to be of the order of a few parts in 105per year. The investigations are continuing and it is hoped that better figures may be obtained for well-aged standards. The stability is not adversely affected—and may be improved—by operation at temperatures up to 120°C. Above 140°C there is usually an increase of resistance, but even at 400°C this increase is not rapid. Operation above 400°C is not recommended even for low-accuracy resistors.Further advantages of the material are a low thermal e.m.f. to copper, high mechanical strength and high ductility. The disadvantages are the necessity of hard soldering and the susceptibility of the resistance to change due to cold working, including vibration.The resistance standards under investigation for long-term stability are five 1-ohm standards, and one standard of each of the following values: 1 000 ohms, 100 000 ohms, 1 megohm and 10 megohms. An additional 100 000-ohm standard of an alternative design has recently been constructed.

DOI:10.1049/pi-b-1.1956.0186

出版商:IEE    

年代:1956

数据来源: IET

摘要:

For a relay-controlled servo mechanism represented by annth-order differential equation it is shown that one change-over and one only, at a unique time, is necessary to bring error and error rate to zero in the least possible time. Previous workers have used the sign of a non-linear function to control change-over, but such methods are suitable only for simple second-order systems with step inputs. To satisfy the need of a technique capable of wider application, e.g. random inputs, prediction of switching by means of a high-speed repetitive analogue computer has been demonstrated with a model experiment. Such a scheme has been shown to be practicable; its use is not limited to simple systems. In the specific example of a 10 kW control system the responses are compared of the on-off control using predicted change-over to those obtained (by simulation) of an orthodox servo mechanism, a linear but saturating servo. The optimized on-off control is always better for all amplitudes of step, ramp and parabolic function inputs, and on the average the same performance would be obtained with both systems using only about half the torque with the on-off control.

DOI:10.1049/pi-b-1.1956.0188

出版商:IEE    

年代:1956

数据来源: IET

摘要:

A new method is presented for the frequency-response analysis of non-linear elements. This involves evaluating the gain of one of the frequency components in passing through the non-linear element when the input to the element consists of two sinusoidal waves of differing amplitudes, phases and frequencies. A cubic characteristic is analysed fully as an example.The paper describes the use of this method for analysing four different types of problem arising in feedback systems containing one simple-type amplitude non-linearity. The method can be considered as an extension of the describing-function technique with a correspondingly wider field of application.

DOI:10.1049/pi-b-1.1956.0191

出版商:IEE    

年代:1956

数据来源: IET

摘要:

Calculations are given for the magnitude and form of the distortion introduced into a square wave by a network or system which exhibits uniform transmission except for increasing (or decreasing) phase delay in the low-mid-frequency region. The fractional peak distortion is found to be equal to twice the area under the curve relatingTnto frequency, whereTnis the delay relative to that at high frequencies. The waveform of the distortion is given for several simple shapes of curve forTn. This distortion is especially characteristic of vestigial-sideband systems, and occurs in television as a ‘pre-shoot’ before a transition and as a smear (in principle equal, but opposite, to the pre-shoot) after it.

DOI:10.1049/pi-b-1.1956.0194

出版商:IEE    

年代:1956

数据来源: IET

摘要:

A system is described for associating and collecting the intercepts of individual particles in a particle scanning system, where the information is presented as a function of the scanning voltages. A series of stores is used to segregate the intercepts, each store having its own memory system and provision for re-use on completion of the scanning of the particle with which it is associated; the stores can thus be used many times during a single frame scan. A method of adding the intercepts of each particle to obtain a measure of the area of the particle is described, but this must be regarded as only one of the possibilities of extracting information from the series of intercepts collected.

DOI:10.1049/pi-b-1.1956.0195

出版商:IEE    

年代:1956

数据来源: IET

摘要:

The amplitude modulation of microwaves produced by the magnetization of ferrites is non-linear and suffers from hysteresis. Hence, square-wave modulation is the only function in which the distortions are not objectionable.The paper describes a feedback method of applying the modulation signal, providing a linearity substantially that of the feedback crystal used to detect the modulated microwave signal, and reducing the effect of hysteresis by an amount approximating the feedback loop gain. Pure sine-wave modulation is achieved at low frequencies, in which the second-harmonic sidebands are more than 45 dB below the fundamental. Linear modulation, by sawtooth and square waveforms, is also achieved, in which the modulation envelope faithfully reproduces the applied signal.The employment of ferrite microwave modulators in engineering applications will involve techniques which are already standard for lower frequencies with conventional components, and the paper establishes the feasibility of using the powerful method of envelope feedback to control them.

DOI:10.1049/pi-b-1.1956.0196

出版商:IEE    

年代:1956

数据来源: IET

摘要:

The provision of wide-band noise sources for the decimetre wavelength region is important for many radio applications. Anoise signal of suitable power level for most purposes is obtained when the plasma region of a gaseous discharge is matched to a transmission line.Examination of the properties of a discharge plasma shows that a noise source with an output which is constant over several octaves can be obtained by matching a cylindrical discharge tube directly to a two-conductor line. Such matching can be achieved by using conductor pairs of various shapes. The factors which affect the operation of the matching element are considered, and a practical design procedure is outlined.These noise sources are simpler to construct and of better performance than those used previously at decimetre wavelengths.

DOI:10.1049/pi-b-1.1956.0197

出版商:IEE    

年代:1956

数据来源: IET

摘要:

The paper outlines the general requirements and the conditions of use of radiation monitors employed in γ- and β-ray survey in connection with geological prospecting. In such instruments, it has been usual to employ either filament valves or cold-cathode valves in amplifier and trigger circuits, and vibrators, filament-valve oscillators or high-voltage battery stacks in the power supplies. Arguments leading to the transistor as the preferable component in all cases are given, and typical transistor circuits are discussed in some detail.Both point-contact and junction transistors are discussed, and the superiority of the junction-type circuits for this type of application is demonstrated.

DOI:10.1049/pi-b-1.1956.0198

出版商:IEE    

年代:1956

数据来源: IET

摘要:

There is a wide choice of scaling devices which will operate at maximum counting rates up to several hundred kilocycles per second, but for counting rates in the region of megacycles per second, the choice is limited almost entirely to special thermionic scaling devices or circuits using thermionic valves. Such circuits tend to be rather complex and have a relatively high power consumption.The paper describes some scaling circuits using transistors which will resolve 0.4 microsec and hence count at a maximum rate of 2.5 Mc/s. The transistors are the normal point-contact type, and the circuits are simple, they have wide tolerances and are economical in power consumption.Features which contribute to the short resolving time are the prevention of bottoming of collector potential and the absence of capacitors. A typical scale-of-10 circuit uses seven transistors, seven pulse transformers and 14 crystal diodes.

DOI:10.1049/pi-b-1.1956.0199

出版商:IEE    

年代:1956

数据来源: IET