Table of contents

This year, 1968, is the 50th anniversary of the formation of the Institute of Physics. (The Institute of Physics and The Physical Society was formed from the amalgamation of the two bodies.) In order to celebrate this Jubilee we are publishing a series of articles concerning various aspects of instrumentation over the last 50 years.

The electromagnetic blood flowmeter is an important tool with both medical research and clinical applications. Its principles and development are reviewed; the subjects discussed include head design, transformer e.m.fs, catheter devices, electronic systems, baseline stability, calibration and commercial instruments.

The shapes of the sideband responses to a nuclear quadrupole resonance line obtained with a logarithmic super-regenerative spectrometer are discussed and compared with those obtained with a new type of spectrometer in which a normally incoherent super-regenerative oscillator is made coherent by injecting an external r.f. voltage into its tank circuit. The phases and amplitudes of the Fourier components of the r.f. output of the locked super-regenerative oscillator are derived and its line shape response is shown to depend on the frequency difference between the locking signal and the r.f. pulses in the super-regenerative oscillator. By controlling this difference, any desired mixture of absorption and dispersion line shapes is obtainable. Other advantages are that the frequency stability is improved, the coherence is independent of quench parameters, and sideband suppression and frequency measurement are both simplified.

A practical locked super-regenerative spectrometer is described in which mechanical ganging between the tank circuits of the two oscillators is supplemented by an automatic phase control negative feedback loop. An output signal can also be obtained from the automatic phase control system and it is shown that this enables separate absorption and dispersion line shape responses to be obtained simultaneously. The actual super-regenerative oscillator circuit requires no adjustments in order to maintain full sensitivity over wide tuning ranges. The performance of the spectrometer is illustrated by recordings of ³⁵Cl nuclear quadrupole resonance in KClO₃ and in the K form of (PNCl₂)₄.

The performance requirements for Zeeman modulators for nuclear quadrupole resonance spectroscopy are discussed and a simple circuit using thyristors is described which generates an antisymmetrical half-sine current waveform in a large-diameter solenoid. The undesirable effects of the coil inductance are eliminated by controlling the thyristor firing angle.

In scanning a spectral peak, significant distortion may result when the response time of the detection and recording system is comparable to, or greater than, the half-width of the peak. Such distortion appears as an increase in line width, decrease in pulse height, time displacement of input and output pulses and some asymmetry. Equations giving the output pulse width, apparent sensitivity and time delay are derived, by assuming a spectral peak of Gaussian shape. Values of the above parameters obtained from numerical solutions of these equations are presented in graphical form, in terms of the initial peak width and the time constant of the detection system.

The velocity of neutrons reflected almost directly backwards from a crystal is close to h/2dm, where h is Planck's constant, d the lattice spacing and m the mass of the neutron. The velocity can be determined accurately by measuring a time of flight over a known distance, without any very accurate determination of the angle of reflection, and such a measurement may therefore provide an accurate determination of d (or h/dm). The proposed experimental arrangement involves two parallel identical crystals in the form of thin disks, placed some metres apart, the first in an intense neutron beam, and the second reflecting into a detector. They are first adjusted for Bragg reflection, and then set in synchronous sinusoidal vibration. The vibration modulates the velocity of neutrons reflected by the first crystal and the velocity accepted for reflection by the second. The intensity of twice-reflected neutrons is a maximum whenever the velocity wave for neutrons arriving at the second crystal is in phase with the vibration of the second crystal. At a given vibration frequency, plots of intensity against the position of the second crystal would exhibit sharp peaks. Measurement of the positions of two such peaks about 10 m apart to within about 0·01 mm would provide a determination of d to within about 1 p.p.m. Aspects of the proposed experiment such as crystal material and size, the frequency and amplitude of oscillations, intensity, and background are discussed, and it is concluded that it is feasible.

A simple, one-piece, optical unit is described which has a pointing sensitivity matching conventional theodolites, but is smaller in size and may be cheap enough to build into structures for permanently defining reference axes. Visual and photoelectric measuring experiments are described.

White light incident upon a cell formed by two parallel semi-silvered surfaces undergoes an interference process within the cell such that the transmitted light has a number of discrete wavelengths missing. The component wavelengths which undergo constructive interference become visible in a spectrometer as regularly spaced coloured fringes (Edser-Butler fringes). The spacing of the surfaces can be calculated from the number of fringes in a given wavelength range with quite good precision. By spacing the surfaces with parts cut from a fibre a figure can be obtained for the maximum diameter of each fibre part across the width of the cell. It will be shown that in experiments with carbon fibres the diameter can be measured with an accuracy to better than ±0·6%.

Electron beam probe equipment is described that measures the potential of flat conducting or semiconducting surfaces with a resolution of 5 mV and a probe diameter of 10-20 μm. The operation of the probe is, in some ways, similar to that of the Vidicon camera tube and the use of a control loop ensures that the output from the system is an accurate replica of changes in surface potential. The probe can be scanned over an area of more than 2 cm diameter and is suited to the examination of planar semiconductor devices.

A simple pressure vessel is described embodying a single valve and seal which is suitable for the irradiation of organic substances at pressures up to 2000 lb in⁻² and temperatures up to 400°C.

A method is described of forming instantaneous products and squares of signals, and of subsequently averaging them over known, and variable, time intervals. This is useful in measurements of fluid turbulence, where the most commonly used instrument is the vacuo-junction. In the case of low frequencies, this device shows certain disadvantages, and a Hall effect multiplier is introduced here as an alternative which has been successfully used.

A simple ballistic technique for measuring small fractional changes of magnetization of a ferromagnetic material, and an apparatus in which this is applied to studying the temperature dependence of magnetization at low temperatures are described. The technique gives direct values of fractional magnetization change, and the sensitivity of the apparatus described is about 1 part in 10⁵ for gadolinium metal.

Methods of using Hall probes to measure dipole fields, and thus specimen magnetization, are analysed and tested. (I, H) loops can be plotted automatically for crystals with a total moment of 0·1 e.m.u., showing discontinuities of 0·01 e.m.u. A method of measuring the absolute value of I_s is described and tested on weakly magnetic crystals, the accuracy being to about 2% for a total moment of 0·1 e.m.u.

The selection, design, calibration and use of a simple thermistor gauge with existing free molecule pressure probes for measurement of macroscopic gas velocities in a rarefied gas are described. Also discussed are typical thermistor bridge circuitry, a vacuum calibration chamber, a particular calibration procedure, and data gathered by employing a thermistor gauge in rarefied air at 10 mtorr with a free molecule probe of length-to-diameter ratio 25.

A thin-film resistance thermometer is assumed to be at or near the stagnation point on the rounded tip of a towed probe. The velocity field that carries the temperature fluctuations in the free stream toward the film is replaced by one of the exact solutions of the Navier-Stokes equations for stagnation flow toward flat walls. Two-dimensional and axisymmetric stagnation flows are considered. The frequency response in stagnation flow is taken as the ratio of temperature-fluctuation amplitudes of the film (on the fluid-substrate interface) and the outer flow at infinite distance from the wall. For the case of one-dimensional, harmonic temperature distribution, the problem reduces to a Ricatti-type equation with an interface boundary condition. For parameter ranges of interest, one- and two-term series expansions in frequency are useful approximations of the exact stagnation-flow solutions. The stagnation-flow frequency-response results are expected to apply to stagnation-region films on glass probes in water. The available experimental data support this conclusion.

Most existing microrheometers generate extension and measure resulting load in the specimen. As the breaking load of a single eucalypt fibre is about 20 g, and its extension to failure is only about 50 μm, it is preferable to generate the load and measure the extension. In the unit described, the load is produced at a constant rate of approximately 0·1 g s⁻¹ by a coil and magnet arrangement. The extension is detected by a differential transformer type displacement pick-up with a resolution of 0·5 μm. The output from an inductance bridge to which the pick-up is connected and a voltage proportional to coil current are fed to an XY recorder to produce a load-elongation curve. Clamp type jaws hold two small paper tabs with the fibre glued across a narrow gap between them. To allow maximum freedom for self-alignment of the fibre, the moving jaw system floats in mercury.

Specimens of strain gauge wire clamped directly in the jaws showed that the instrument performed satisfactorily. Specimens glued with Araldite showed evidence of glue strain, as indicated by the modulus decreasing with free length, and a technique for assessing true modulus is presented. The instrument is designed so that it may be applied to work on rheology of wet paper webs, lateral compression testing of fibres, and also to provide sinusoidally oscillating stresses as well as time dependent unidirectional load programmes.

The design, manufacture and calibration of a range of thin-film, heated-element gauges to measure instantaneous velocity or skin-friction in air, water or blood is described. The probes consist of a platinum resistance thermometer which is heated electrically, with additional elements to detect reversed flow. The manufacturing technique of a flush-mounted skin-friction gauge, and a needle probe currently used in the diagnosis of heart conditions in man, is described in detail.

Modern electronic components, particularly integrated circuits, allow a new approach to the old problem of determining damping capacity. The new circuit enables the time taken for the amplitude of a freely vibrating specimen to half its value to be determined and the result is expressed either in the number of vibrations of the specimen or the elapsed time.

This article describes a new method of performing analogue division and its applicability in generating the general function V = (a₁V₁ + a₂V₂)^m/k/ (b₁V₁ + b₂V₂)^n/k subject to the condition 0 less-than-or-eq, slant V less-than-or-eq, slant 1. In particular a form of the circuit giving V₁(V₁ + V₂) is described.

A simple automatic instrument for the recording of time-lags in the range 1 ns upwards is described. It has been primarily designed for the measurement of breakdown time-lags, being largely immune from spurious triggering likely to be produced by stray surge currents. The instrument has been found useful in a number of applications where oscillographic recording is awkward. A brief discussion of the errors characteristic of this type of measurement is included.

An instrument for the digital measurement of short time intervals, with a channel width variable between 2 and 5 ns, is described. It uses the technique, conventional for longer intervals, of counting clock pulses derived from a stable oscillator of known frequency, and has the additional advantages of a short dead time and a long term stability.

The theoretical response of a pair of thermistors when used as the sensing elements in a total radiation pyrometer is calculated. The response is shown to be almost linear over a very wide range of temperature. Practical circuits to exploit these properties in a total radiation pryrometer and in a temperature control system based on a radiation pyrometer are given.

This note describes a modified `holder technique' for preparing thin metal foils for transmission electron microscopy. In this, profiling of the sample and determination of the exact moment of perforation are carried out relatively more easily than in the earlier techniques.

A cheap, simple device is described for maintaining the level of liquid nitrogen in Dewar vessels to within adjustable limits. It uses a germanium diode as the level sensing element.

The note describes a mechanical seal devised to seal a calcium fluoride window to a stainless-steel flange at temperatures in the range 500-1000°C. The assembly consists of a gold sealing ring, a compliant backing ring and a constant force device to provide the seating load. Differential pressures of ±500 Pa (±2 in water gauge) are withstood and the large expansion and low strength of the window are designed for.

A modified four-ball lubricant-testing machine in which each of the three fixed balls makes its own wear track on the rotating ball is described and discussed. The effect of this modification on the sliding contact surfaces is considered. The geometry of the system may be chosen so as to either minimize or maximize the differences among the three contacts in sliding velocity, normal load and apparent contact area.

A method is described for producing single crystals of ulvöspinel from the melt. The melting is done in a depression pressed in a platinum strip which is electrically heated. The subsequent cooling of the ulvöspinel through its melting point commences at a cold spot produced at the base of the depression by a sliver of platinum welded to the external surface. The whole system is enclosed in a carbon dioxide-hydrogen atmosphere whose partial oxygen pressure is in equilibrium with the ulvöspinel at the melting point.

A mechanism is described which automatically follows the mean vertical movement of three mercury manometers and indicates the position directly on a mechanical reversing counter. At the same time cam-operated switches set up a printing counter which is arranged to record at predetermined intervals.

A simple electronic spark cutter is described. The vertical height of a table on which the sample is mounted is positioned by a servo control to maintain a constant spark gap between the workpiece and molybdenum wire cutting tool. The wire is pulled between spools to combat erosion.

An internally heated autoclave with argon as the pressure medium for use up to 1600°C and 4500 b is described.

An inexpensive device is described for use with a potentiometric recorder for continuously varying the temperature of an environment chamber in accordance with any desired preset programme. This enables the diurnal temperature fluctuations encountered in field work to be simulated readily under laboratory conditions.

A 0-4 kv pulse amplifier has been constructed for field ion microscopy. The circuit employed was a triode amplifier, powered by a 5 kv full-wave voltage doubler, triggered by a General Radio 1217-B pulse generator and capacitively coupled to the field ion microscope via a 30 kv, 0.01 μF capacitor. The circuit produces either negative or positive field evaporation pulses. The usable minimum pulse time was found to be in the range 10-20 μs for this amplifier design.

An atomic beam oven is described which is suitable for high-resolution crossed electron-atom beam experiments. For an oven temperature of 420°C the associated magnetic field a distance of 1 in from the oven aperture is found to be less than 1 mG.

A cell-space adjustable slit allows the reference beam of a Cary 14R spectrophotometer to be attenuated by a factor of up to 1000. The attenuator transmittance has been measured and the results shown graphically as corrections applicable to observed sample transmittances. The smallest nominal values of recorder full scale deflection obtainable without excessive pen noise are 0·01% in the ultra-violet and visible, and 0·1% in the infra-red ranges of the instrument. The transmittance curves for shade 12 welding filters (luminous transmission factor of about 0·002%) have been measured without thinning the samples over the range 270 nm to 2·6 μm. The technique may also be used for double-beam spectrophotometry of other kinds of dense filters.

A simple modification to a Robinson n.m.r. absorption circuit is described which allows the oscillation frequency to be swept at an approximately uniform rate over a relatively wide range.

An instrument is described, designed for quasistatic tracing of hysteresis loops of ferroelectric capacitors. The instrument is equipped with circuits for compensation of insulation resistance and of the linear component of the capacitance of the ferroelectric sample; compensation of external spurious leakage currents is also provided for. With ferroelectric samples having a high and relatively constant insulation resistance (as for example triglycine sulphate) the error in loop-closing is typically below 0·5% of its amplitude even at tracing times up to 1000 s.

A goniometer for use in combination with the standard cooling equipment in the JEM series 6 and 7 electron microscopes is described. The construction allows smooth tilting up to ±20° with only minor lateral displacement of the part of the specimen under inspection.

A rotating-anode electropolishing unit of integral design is described and details are given of its use in the electrolytic polishing and etching of nickel. Circuit diagrams of transistor power supplies for both electropolishing and electroetching are also given.

The suspension strings used to support the longitudinal specimen absorb energy and so give a falsely high value for the natural damping capacity. The note describes an experimental procedure and method of extrapolation of the results which effectively eliminates this source of error.

A simple method of temperature control in the range - 70 to 100°C for proton magnetic resonance experiments is described. The method is based on a liquid cell cryostat containing a temperature-regulated non-hydrogenous liquid. Two probe-feedback networks are described for controlling the temperature: (i) a simple contact thermometer operating a relay switch; (ii) a thermistor amplifier circuit. The circuit for the latter is described. The method eliminates temperature gradients in the sample and provides long term stability to at least ±0·1 degC.

Difficulties were experienced in the operation of a high-temperature furnace attachment to a single-crystal goniometer. These are described, together with the modifications which were made to facilitate use of the attachment and to obtain accurate indication and control of the specimen temperature.

A simple photocathode cooling device using the Peltier effect is described. The cathode temperature achieved is sufficiently low for the thermal emission from Sb-Cs and tri-alkali cathodes to be comparable with other sources of dark current in the image tube.

Calibration curves, derived from absolute susceptibility measurements are given for four-inch Henry profile pole tips at gaps from 5 cm to 2·5 cm, for several magnetizing currents. The region of `constant force' permits accurate measurements on samples of limited size.

A novel application of the quadrant electrometer for the precision measurement of frequency is described. A null method has been developed employing only a standard capacitor and a resistor as the circuit components. The authors used the method for the calibration of an electronic oscillator in the frequency range 10 Hz-20 kHz.

In this Letter are discussed the general design principles and some of the transistor circuits required for coherent amplifiers and it is shown how a recently published circuit by Gerard can be improved.