Table of contents

We describe here the design and performance of a current sensing noise thermometer using a low T_c DC SQUID as the front end amplifier. The DC SQUID is used to measure the thermal noise current in a resistor and the temperature is then obtained from the Nyquist formula. The thermometer is fast, absolute and precise and is usable over a wide temperature range below 4.2 K, in principle down to well below 1 mK. The excellent energy sensitivity of the DC SQUID, operated at fixed temperature, enables the use of a relatively large noise resistor, in the mΩ range. This requires relatively short averaging times when measuring the spectrum of noise fluctuations. We have shown that it is possible to determine absolute temperature with a precision of 1% in a measuring time of 10 seconds with an amplifier noise temperature, T_N, of the order of 30 µK, and to an accuracy better than 0.3%. The percentage precision is independent of temperature for temperatures much greater than T_N. Our method of heat sinking the noise resistor ensures proper cooling of the electrons. We incorporate a fixed point device for checking the gain calibration. We have cooled the thermometer successfully to below 1 mK, achieving a minimum electron temperature of 300 µK. We present the results of a preliminary comparison with a ³He melting curve thermometer (MCT) above 4.5 mK, and with a platinum NMR thermometer down to the lowest temperatures.

We present a new capacitive sensor for displacement measurement in a surface-force apparatus which allows dynamical measurements in the range 0-100 Hz. This sensor measures the relative displacement between two macroscopic opaque surfaces over periods of time ranging from milliseconds to, in principle, an indefinite period, at a very low price and down to atomic resolution. It consists of a plane capacitor, a high frequency oscillator and a high sensitivity frequency-to-voltage converter. We use this sensor to study the nanorheological properties of dodecane confined between glass surfaces.

An original experimental technique for simultaneous electrical conductivity and thermo-e.m.f. measurements in a wide temperature range (up to 2000 K) under high pressures (up to 50 MPa) has been developed. Special ceramic measuring cells were designed and constructed for investigation of chemically aggressive liquid semiconductors. Different sources of errors connected with convective flows, jamming signals, systematic deviation of measuring devices and especially diffusion of the liquid sample into the body of the ceramic are analysed and method of elimination are proposed. New results for some liquid metal-chalcogen alloys in comparison with those obtained earlier illustrate advantages of the proposed method.

The authors present a numerical one-dimensional model which computes the soil water content profile, along transmission line probes, from time-domain reflectometry signal traces recorded with a reflectometer (TDR). Taking a discrete time-domain approach, they first developed a direct iterative model which computes simulated signal traces from the dielectric constant distribution of the propagation medium. The results tallied well with signal traces recorded in simple cases. The inverse model was developed in a second stage, to compute water content distribution along the transmission line probes from a reflected signal trace. The results are shown both for controlled media and field conditions and are compared to direct water content measurements. Filtering of the reflection coefficients had to be incorporated to compensate for the high sensitivity to noise of the iterative computation and the simplicity of the propagation hypothesis. The inverted profiles are very close to the direct measurements, validating this model for measurements in low loss soils.

Experiments were performed using the electronic sine-wave voltage-perturbation test to systematically study the frequency responses of near-wall hot-wire probes subjected in turn to varying magnitudes of convective velocity and different effects of wall influence. In addition, quartz-substrate hot-film gauges with various thicknesses of quartz coating were also investigated. Results of the high cut-off frequency obtained using the sine-wave test (f_sine) were found to be in fair agreement with those obtained using the square-wave test (f_S) both for hot-wire and for hot-film sensors. The sine-wave test response curve exhibited a distinct bulging effect for the hot-film gauges. For the hot-wire sensors, a much weaker bulging effect was also observed. In contrast to f_S and f_sine, the low frequency response characteristic corresponding to the location of the bulging effect (f_bulge) compared much more favourably with the dynamic frequency response (f_D) obtained by Khoo et al and Chew et al using a known near-wall fluctuating flow field. Freymuth's theory for non-cylindrical hot-film sensors incorporating the Bellhouse-Schultz model was applied to predict the responses of the hot-film wall gauges when they were subjected to electronic sine-wave testing and dynamic perturbation testing under different parametric conditions. Although it is one-dimensional in nature, the model is capable of predicting most of the trends observed in the present study and previous works by Khoo et al (1998a) and Chew et al (1998a).

A novel laser Doppler sensor, which allows the direct measurement of velocity profiles over the length of the measurement volume is presented. The measuring principle is based on a chromatic coding of the detected tracer particles, arising due to the dispersion of the two employed laser wavelengths. Velocity measurements with a spatial resolution up to 60 µm inside the measuring volume are demonstrated.

A fine needle probe for determining the thermal conductivity of penetrable materials such as fluids, fruit and animal flesh has been developed. The present probe is constructed by inserting twenty strands of copper wire with electrical insulation coating into a fine stainless steel needle. The copper wire serves both as a heating unit and as an electrical resistance thermometer. The effects on the thermal conductivity measurement caused by the thin needle wall have been analysed. It is found that the effects can be negligibly small if the instruments and the measurement procedure are adequately designed. The usability of the as-constructed apparatus for penetrable materials has been tested to measure the thermal conductivities of liquid, fruit and animal flesh. The accuracy of the present measurement was estimated to be within ±3%.

In roughness measurement a standardized Gaussian filter according to ISO 11562 is used to separate roughness from waviness. If the uncertainty of the measured profile is known, the uncertainty of the roughness profile as well as the waviness profile can be obtained by calculating the propagation of the measurement uncertainty in case of the filtering process. It turns out that, even in the case of uncorrelated measured data, the filtered data are always correlated. This correlation must be taken into account when calculating roughness or waviness parameters.

This paper presents an uncertainty analysis for roundness measurements on a glass hemisphere. The measurements were carried out with the spindle measuring instrument Talyrond 73. For the correction of spindle errors, a multi-step error separation technique with ten steps was used. Various input quantities were investigated such as the form deviations of the hemisphere, the characteristic of the probing system and thermal drift effects. Finally, a detailed uncertainty budget was developed. It is shown that the standard uncertainty for a single profile point is 1.5 nm. This results in a standard uncertainty of u(RON_t) = 3 nm for the roundness deviations of the hemisphere.

A simple correction of residual non-linearity of inverse capacitance displacement transducers is presented. The transducers were designed for independent probe position monitoring of a scanning probe microscope and/or linearization of the closed-loop control of a SPM scanner. The linearity of the transducers prior to compensation was tested in an SPM head using optical interferometry and the fine-tuning and efficiency of non-linearity suppression using a scaled-up model of the sensor capacitor, mounted on a micrometer.

A proof-of-principle, digital signal processing system is described which can perform deconvolution of audio-bandwidth signals in real time, enabling separation and precise measurement of pulses smeared by a given impulse response. The system operates by convolving a time-domain expression of an inverse filter with the original signal to generate a processed output. It incorporates a high-level user interface for the design of the inverse filter, a communications system and a purpose-designed digital signal processing environment employing a Motorola DSP56002 device. The user interface is extremely versatile, allowing arbitrary inverse filters to be designed and executed within seconds, using a modified frequency sampling method. Since the inverse filters are realized using a symmetrical finite impulse response, no phase distortion is introduced into the processed signals. A special feature of the design is the manner in which the software and hardware components have been organized as an intelligent system, obviating on the part of the user a detailed knowledge of filter design theory or any abilities in processor architecture and assembly code programming. At the present time, the system is capable of deconvolving signals sampled up to 48 kHz. It is therefore ideally suited for real-time audio enhancement, for example, in telephony, public address and long-range broadcast systems, and in compensating for building or room acoustics. Recent advances in DSP technology will enable the same system structure to be applied to signals sampled at frequencies ten times this rate and beyond. This will allow the real-time deconvolution of low-frequency ultrasonic signals used in the inspection and imaging of heterogeneous media.

A non-contact technique for obtaining accurate profiles of optical quality surfaces with micrometric accuracy has been developed. The technique is based on the Ronchi test principle, that is, on the study of the interaction of a wavefront reflected on the surface to be profiled with a square-wave transmittance ruling. From the resultant fringe pattern and some basic geometrical optics principles it is possible to measure the local normal to the surface being tested at a set of given points. This local normal map may then be integrated, yielding the surface profile. By use of a theoretically expected surface shape, the main parameters of the surface may then be determined by surface fitting of the measured data to that expected surface shape. Results of the profilometric measurements both of a spherical and of a toroidal surface are presented. The measured profiles are validated by comparison of the radii of curvature obtained using a high precision radioscope with the ones obtained by surface fitting of the measured profiles to their expected surface shapes. Additionally, subtracting the best-fit theoretical surface from the measured profile allows the observation of surface deviations from the theoretical shape to within some tenths of a nanometres.

A transducer is developed for the continuous measurement of biaxial pole strains in sheets as they are bulge formed through circular and elliptical dies by the action of lateral hydraulic pressure. The transducer measures simultaneously the two principal chord lengths and the pole height to enable calculation of the required strains. It employs the principle of an elastic arch whose feet remain in contact with the plastically deforming sheet. Two such arches, one at right angles within the other, admit a vertical displacement probe. The prototype design shows linear calibrations between output bridge voltage and the feet displacement on a flat surface. The linear calibration is preserved by maintaining contact between feet and the sheet as its principal curved surfaces are being formed. This is achieved by connecting the feet to securing masts with rings so giving a hinge with the required rotation. The transducer provides the three outputs required for a data logger with strain conversion based upon the geometry of intersecting chords. Good agreement is found between transducer strains and a theoretical prediction. The latter assumes that the principal axes of the bulge deform into circular arcs to contain the die rim and the pole point at a given height above the die surface.

Nd:YAG rod wavefront distortion has been quantified in a Twyman-Green interferometer which employs the multiplicative analogical moiré phase-shifting method for demodulation. The measuring procedure is based on the superimposition of a transmission grating located outside the interferometric module with a phase modulated high-frequency Twyman-Green spatial carrier. Later, a set of fringe patterns, provided by low-pass filtering of the phase-shifted multiplicative moiré patterns obtained, is combined in a generic phase-shifting algorithm (GPSA). The interesting test information is achieved once the previous procedure has been carried out with and without the transparent object located in the measuring area. The wavefront distortion (WFD) is calculated by the subtraction of these results provided by the GPSA, or directly if a differential phase-shifting algorithm (DPSA) is employed. The WFD obtained is in accordance with characteristics of the test object. The comparison of the results obtained with both families of algorithms provides information about the sensitivities to error sources of the DPSA employed. A first approximation to the analysis of significant error sources of DPSAs is presented.

There are many techniques available for measuring the dielectric properties of materials. The one-port reflection coefficient method is one of the simplest techniques for this application. In this paper a large coaxial cell is described with emphasis on how it was designed, constructed and calibrated over the frequency range from 50 MHz to 1 GHz. This cell is much larger than commercially available ones and suitable for the dielectric property measurements of a wide range of liquid and solid materials. An improved technique that employs a minimization scheme to extract the dielectric properties of the sample is presented. Measurements have demonstrated that this system is efficient and accurate for determining the dielectric properties of various materials over a broad frequency band.

A simple low-cost Mössbauer spectrometer based on an electrically programmable logic device and programmed in the digital design language VHDL is described. The spectrometer is fast, with a minimum channel dwell time of about 12 µs, set by the speed of transfer of data to the host PC, and has 512 channels that can easily be extended if required.

This design note describes an optoelectronic technique able to obtain the surface topography. The sensor is based on a new fringe projection system, which uses a phase diffraction grating. The proposed system is simple, compact, robust and easy to use.

As lasers and photonic devices and systems become more important in electrical and information technologies, there is definitely a need to introduce this subject into the undergraduate curriculum of electrical engineering students as early as possible. Since most of the laser and electro-optical devices require extensive background in optics, this presents a problem for the electrical engineering textbook. Fundamental optics is not taught in the traditional electrical engineering curriculum. Many laser and electro-optic courses review some basic electromagnetic theory and move right into laser theory and device principles. Students often struggle to fill in the elementary optics background that is needed. This book is written to fill the gap in optics in standard laser textbooks by systematically presenting the relevant topics in optics that are important for laser science and devices.

The book covers optics quite well, probably because of the fact that the earlier version of the text had the title `Optics'. From geometric optics to wave optics, interference to diffraction, basic theory to optical instruments, and polarization to coherence, the authors give a balanced discussion of optical phenomena and their mathematical formulation. The mathematics has been kept at an elementary level.

The modern parts that include lasers and various photonic and electro-optic devices are a bit too sketchy in comparison with the optics part (chapters 1 to 14). Descriptions of laser systems, nonlinear optics, fiber amplifiers and ultrashort pulse lasers, to name just a few, are either too brief or totally lacking.

Color photographs are excellent features throughout the book. Solutions to numerical examples and problems will definitely help students using the book for self-study.

In conclusion, this is a self-contained book for an undergraduate text on lasers and photonics that does not require instructors and/or students to search for supplemental topics in optics.

Chi H Lee

After a short introduction on generalities about probability distributions, the book consists of 40 chapters describing in alphabetical order individual distributions commonly encountered in applications. It is a substantially revised edition with, in particular, new diagrams and a longer treatment of the Weibull distribution in its two, three and five parameter forms. With two exceptions the distributions are univariate, i.e. one-dimensional.

Typically for each distribution there is an introductory paragraph about potential applications, the formula for the distribution, the main properties of the distribution, usually some diagrams illustrating the shape of the distribution, and some notes on relations with other distributions and on how its parameters might be estimated from data. In some cases a note on simulation of values is included. The book ends with some computational references, some statistical tables and a short bibliography.

The book contains a large amount of information clearly set out in concise form. The introductory notes on the motivation for the individual distributions are sometimes perfunctory and inevitably are no substitute for a careful discussion of the domain of potential applicability of, for example, the exponential distribution or the negative binomial distribution. They are of little or no help in dealing with the question: `Here is a particular kind of problem calling for a simple form of probability distribution: what kind of distribution might be useful?' However, the book could certainly be a valuable reference source on such specific if rather narrow questions as: `Given involvement in the use of the negative binomial distribution, what are its moments?'

A criticism of the book is the failure to give references for further reading about individual special distributions. The short general bibliography is no help on where to look for particular points.

D R Cox