摘要:

There is a large field of application of digital-computing techniques where reliability and simplicity are far more important than speed. A basic circuit is described which uses one transistor, one capacitor and three resistors, from which a complete digital-computing system may be constructed economically. A system using this circuit is extremely simple to design, construct and maintain, and should prove very reliable, although it cannot achieve the speed of operation of systems using more complex circuits. The way in which well-known computer circuits are constructed from this basic circuit is described, and an account is given of a complete logical system, employing 184 such circuits, which was constructed to demonstrate their application.

DOI:10.1049/pi-b-1.1958.0283

出版商:IEE    

年代:1958

数据来源: IET

摘要:

Apparatus is described by means of which the short-circuit current gain is measured directly. Results of such measurements are presented for commercial alloy-junction and surface-barrier transistors; corrections are applied to yield the internal diffusion-current gain. The effects of stray capacitances on the measurements are discussed. The cut-off frequency of the internal current gain is compared with values derived indirectly from other measurements. For alloy-junction transistors the behaviour is closely in accord with existing one-dimensional diffusion theory, with some reservations, but for surface-barrier transistors the agreement is less close.

DOI:10.1049/pi-b-1.1958.0284

出版商:IEE    

年代:1958

数据来源: IET

DOI:10.1049/pi-b-1.1958.0285

出版商:IEE    

年代:1958

数据来源: IET

DOI:10.1049/pi-b-1.1958.0286

出版商:IEE    

年代:1958

数据来源: IET

摘要:

In nucleonic work a common requirement is the counting of pulses occurring at relatively long intervals of 250 microsec or more. Cold-cathode decade tubes (e.g. Dekatrons) and electro-mechanical registers have proved their value in this field, but their performance and reliability depend to a large extent on the means employed to drive them. The use of transistors for this purpose should result in efficient and reliable circuits.In the circuits described the Dekatron is driven by a transistor blocking oscillator which, when triggered, produces a pulse of defined amplitude and width, followed by a similar pulse of opposite polarity. A secondary winding on the blocking-oscillator transformer applies these pulses, in the correct phase and amplitude, to the Dekatron guide electrodes. When decades are cascaded, the negative edge of the output cathode waveform, instead of the normal positive edge, is used for triggering the next stage. This system reduces the delay between input and output pulses to a few microseconds, which is the triggering time of the blocking oscillator, and so reduces the error when the Dekatrons are used for counting standard time intervals.For operating a mechanical register two transistors are cross-coupled in a monostable circuit. The register is in the collector circuit of one transistor, which conducts for 0.1 sec when triggered. Since only a small fraction of the supply voltage appears across the transistor when it is conducting, registers requiring several watts can be operated by low-power transistors.The transistor circuits produce accurate waveforms for operating the Dekatrons and registers, enabling the minimum resolving times to be realized.

DOI:10.1049/pi-b-1.1958.0287

出版商:IEE    

年代:1958

数据来源: IET

摘要:

The paper deals with the properties of half-wave magnetic amplifiers having finite control-circuit resistance or rectifier reverse conductance. The analysis is based on the concept of ‘reset factor’, which is related to the magnetic-amplifier Q-factor during the reverse half-cycle of excitation voltage.It is shown that maximum power gain of a half-wave magnetic amplifier is a function merely of the ratio of unsaturated load-winding reactance to load resistance. In most practical cases maximum gain is obtained with a very small number of control turns, so that nearly all the winding space can be used for the load winding. Curves are presented which may be used to determine optimum turns ratio and maximum gain for various combinations of control-circuit resistance and rectifier leakage.Experimental results for magnetic amplifiers using different core materials are presented. These results agree well with theory. Finally, the effect of various bias circuits on half-wave magnetic-amplifier performance is examined.

DOI:10.1049/pi-b-1.1958.0288

出版商:IEE    

年代:1958

数据来源: IET

摘要:

The sensitivity of a d.c. amplifier using transistors is mainly limited by the variation with temperature of the inter-electrode potentials and currents of the transistors, the variation between any two temperatures being termed the thermal drift.A typical low-frequency germanium transistor connected as a directly-coupled common-emitter amplifier may introduce current and voltage drifts, referred to the input, of 50 μA and 100 mV, respectively, when the temperature changes from 20 to 500°C. By using two such transistors in a balanced circuit, in which the drifts oppose, the net drifts are reduced to about 1 μA and 2 mV, respectively, if several interdependent component values are adjusted at different temperatures. By selecting transistors, and using greater care in balancing the amplifier, the drift can be reduced by one or two orders of magnitude, but such a procedure is usually unacceptable.The same transistor connected as a simple modulator introduces a fundamental current drift of only 3 μA but it still has a voltage drift of 100 mV. It is thus most suitable for use with devices having a high source impedance such as an ionization chamber for measuring radiation intensity. The current error can be reduced by a factor of 10 to 0.3 μA if two such transistors are used in a balanced modulator which requires only one adjustment at a single temperature.The use of the transistor in a common-emitter chopping circuit results in a current drift of 50 μA and a voltage drift of only 2 mV, while reversing the functions of emitter and collector reduces these drifts to 3 μA and 100 μV, respectively. The latter voltage drift is considerably less than that of most thermionic-valve amplifiers, but the current drift is still large by thermionic-valve standards, and limits the use of the chopper to low-impedance sources.A modification to this chopper eliminates the leakage current, but leaves the voltage drift unchanged at 100 μV. This voltage drift, acting on the transistor impedance, produces a current drift of only 3 × 10−8amp, which makes this chopper suitable for use with high-impedance sources.The chopper drift figures are obtained without recourse to selection or balancing of tcansistors, and are low enough to enable the transistor to contribute its many other advantages to such applications as sensitive null detectors and analogue computers.The current drift can be reduced still further by using higher-frequency transistors, which, in general, have lower leakage currents. For example, the surface-barrier SB100 transistor (fcα≃ 50 Mc/s) introduces a maximum current drift of only 3 × 10−9amp.The silicon junction transistor, promising even lower leakage currents, should enable direct currents of less than 10−11amp to be measured.

DOI:10.1049/pi-b-1.1958.0289

出版商:IEE    

年代:1958

数据来源: IET

摘要:

In transistor d.c. amplifiers, the sensitivity is mainly limited by drift of operating point caused by changes in ambient temperature. A modulated amplifier is necessary to obtain a high sensitivity without recourse to methods of drift compensation requiring the adjustment of several balancing controls. The paper describes a modulated system consisting of a transistor input chopper, a high-gain transistor a.c. amplifier, and a transistor output chopper. Gains are expressed throughout as transfer impedances (output-voltage/input-current), a convention which is well suited to transistor amplifiers. The a.c. amplifier uses four transistors and has a gain of 20 volts/μA over a band covering 60c/s–20kc/s. It contains one capacitor only, giving a single low-frequency time-constant, which simplifies stability problems when feedback is applied. A chopping frequency of 1·6kc/s is used, and the complete system gives an open-loop gain of 50 volts/μA with a bandwidth extending from direct current to 25 c/s. The peak output is ± 10 volts, and the current drift referred to the input is 4 × 10−9amp in the range 20° C–50° C. The voltage drift at the input is less than 100 μV. Despite the low input impedance the virtual-earth principle is still valid, and the amplifier should be useful in analogue-computer applications. The complete amplifier is suitable for amplification of signals from a wide range of source impedances—ionization chambers and thermocouples being typical extremes.

DOI:10.1049/pi-b-1.1958.0290

出版商:IEE    

年代:1958

数据来源: IET

DOI:10.1049/pi-b-1.1958.0294

出版商:IEE    

年代:1958

数据来源: IET

摘要:

A theoretical investigation has been made of the distortion produced in f.m. frequency-division-multiplex trunk radio systems by the simultaneous reception of a wanted and an unwanted carrier, not necessarily identical in frequency. This form of disturbance can occur both between different systems and within a single system.It is found that, unless there is substantial failure of the receiver limiter action, the distortion generated in a particular channel is accurately proportional to the relative level of the interfering carrier. Thus, it is possible to plot curves of distortion in the worst channel against frequency separation of wanted and unwanted carriers, from which, for a given separation, a permissible level of interfering carrier can be deduced. Curves are presented for 240- and 600-channel systems when the interfering carrier is either modulated or unmodulated. Numerical results are shown for various carrier-frequency separations that can arise under the C.C.I.R. frequency plan for systems of six r.f. channels each carrying 600 speech channels.The distortion due to this mechanism can be divided into intelligible and unintelligible crosstalk. Reasons are given for the expectation that the former will not normally be important.Some measurements of distortion due to an unmodulated interfering carrier, using components of a 600-channel system, show good agreement with the theory when the frequency separation is not too large. For separations of the order of 9 Mc/s, or greater, distortion levels substantially above the theoretical were noticed. The discrepancy is considered to be due to departure in the demodulator from the ideal behaviour assumed (i.e. frequency insensitivity of the limiter and adequate broadness of band of filters, etc., and of the discriminator).

DOI:10.1049/pi-b-1.1958.0295

出版商:IEE    

年代:1958

数据来源: IET