Handbook of Modern Sensors: Physics, Designs, and Applications 5th Edition

Handbook of Modern Sensors: Physics, Designs, and Applications 5th Edition

English | 2016 | ISBN: 978-3-319-19302-1 | 758 Pages | PDF | 25 MB

This book presents a comprehensive and up-to-date account of the theory (physical principles), design, and practical implementations of various sensors for scientific, industrial and consumer applications. This latest edition focuses on the sensing technologies driven by the expanding use of sensors in mobile devices. These new miniature sensors will be described, with an emphasis on smart sensors which have embedded processing systems. The chapter on chemical sensors has also been expanded to present the latest developments.
Digital systems, however complex and intelligent they may be, must receive information from the outside world that is generally analog and not electrical. Sensors are interface devices between various physical values and the electronic circuits that “understand” only a language of moving electrical charges. In other words, sensors are the eyes, ears, and noses of silicon chips.
Unlike other books on sensors, the Handbook of Modern Sensors is organized according to the measured variables (temperature, pressure, position, etc.). This book is a reference text for students, researchers interested in modern instrumentation (applied physicists and engineers), sensor designers, application engineers and technicians whose job it is to understand, select and/or design sensors for practical systems.


Functional Approximations

Approximation is a selection of a suitable mathematical expression that can fit the experimental data as close as possible. The act of approximation can be seen as a curve fitting of the experimentally observed values into the approximating function. The approximating function should be simple enough for ease of computation and
inversion and other mathematical treatments, for example, for computing a derivative to find the sensor’s sensitivity. The selection of such a function requires some
mathematical experience. There is no clean-cut method for selecting the most appropriate function to fit experimental data—eyeballing and past experience perhaps is the only practical way to find the best fit. Initially, one should check if one of the basic functions can fit the data and if not, then resort to a more general curve-fitting technique, such as a polynomial approximation, e.g., as described below. Here are some most popular functions used for approximations of transfer functions.

Linear Regression

If measur ements of the input stimuli during calibr ation cannot be made consi stently with high accuracy and large rando m errors are expected , the minimal number of
measur ements will not y ield a sufficien t accuracy . To cope with rando m errors in the calibrat ion process, a method of least squares could be employed to find the slope and intercept. Since this method is described in many textbooks and manuals, only the final expressions for the unknown parameters of a linear regression are given here for reminder. The reader is referred to any textbook on statistical error analysis. The procedure is as follows:


If tolerances of a sensor and interface circuit (signal condi tioning) are broad er than the require d o verall accuracy , a calibrat ion of the sensor or, prefera bly, a com bination of a senso r and its interf ace circui t is requi red for minim izing err ors. In other words, a calibrat ion is required whe never a highe r accur acy is required from a less accurat e senso r. For exam ple, if one needs to measure temperat ure with accuracy, say 0.1  C, while the availa ble sensor is rated as having accuracy of 1  C, it does not mean that the sensor cannot be used. Rather thi s particula r sensor needs calibration. That is, its uniqu e transfer function shoul d be determ ined. This proce ss is cal led calibra tion.

Sensors for Mobile Communication Devices

All these possible options are workable from the engineering standpoint, yet from a convenience and practicality perspective, it appears that option b is the most attractive
for coupling with a generic MCD that is used by a consumer. Positioning sensors inside a protective case allows hiding them ergonomically and inconspicuously and
making the sensors instantly available whenever needed without additional actions by an operator. The sensing “smart case” communicates with an MCD either by wires or preferably wirelessly, for example through NFC or Bluetooth.