Electronic Measurement Systems

Theory and Practice

A.F.P van Putten

• Publisher: CRC Press
• Year: 2019
• ISBN:
  9781351453134
• Edition: 1
• Topic: Naturwissenschaften
• Subtopic: Physik

Electronic Measurement Systems: Theory and Practice, Second Edition is designed for those who require a thorough understanding of the wide variety of both digital and analogue electronic measurement systems in common use. The first part of the book discusses basic concepts such as system specification, architectures, structures, and components. Later chapters cover topics important for the proper functioning of systems including reliability, guarding/shielding, and noise. Finally, an unusual chapter treats the problems of the human aspects of the design of measurement systems. The book also includes problems and exercises.New to the Second Edition

• Extended section about signal structures, I/O bussystems, DAQ boards, and their architecture
  
• User programmable devices (UPLD's) and the use of microprocessor principles in instrumentation
  
• Novel approaches on reliability due to built-in testability becoming a major design feature
  
• A brief introduction to the related physics of each transducer energy domain to understand what the principle of operation is
  
• Discussion of the ADM method for drift elimination
  
• Introduction to the European Electro Magnetic Compatibility legislation and the ISO 9000 system
  
• Additional noise calculation techniques and noise in sensors
  
• Chapter on autozeroing transducers and sensor interfacing, paying particular attention to bridge circuits for modulating transducers

One

INTRODUCTION TO ELECTRONIC MEASUREMENT SYSTEMS

Ever since human beings began to think, we have exchanged information and have given measures to quantities in an attempt to understand our surroundings. We appear to have a basic curiosity about and a need to investigate our environment. We recognize this with our senses, but our natural capacities are limited, so we have developed tools to help us to fulfil our measurement needs. In a continual process towards perfection, we are still refining our tools to improve our understanding of all kinds of mechanism. Often models are shaped which describe more or less accurately the real world. This requires collection of information about the environment, a system or a process. Knowledge is related to information and, with collected information concerning a system or a process, we can increase our knowledge. This is a continuous story of interaction: model making and performing measurements and vice versa. Besides these aspects our society has become an electronic society and, generally speaking, the complexity of our society forces us to measure and to control energy and information in a large variety of applications.

1.1 INFORMATION

Information is anything that increases our image of knowledge of a system or process (MacKay), for instance, a type of arrangement or regularity which can be recognized in the environment. The letters on a page show a certain regularity and the message involved is recognized by the sequence in which the letters have been printed. For instance, information content is implemented in the order in which the letters appear on paper, and a limited quantity of information is conveyed. We can also say information is anything which reduces uncertainty of the source sending this information. We call this type semantic information and it is language related. If the same letters appear completely at random, all semantic information is lost and interpretation is no longer possible. The amount of semantic information has become zero. It is also said the left redundancy is zero.

* * *

Example 1.1

To illustrate the concept of information redundancy in a language consider the following sentence: ‘N t mrng t bkr bks brwn brd’. Although characters are missing you still can read this sentence; in other words in most sentences or words redundancy is present.

* * *

However, no quantitative measure for semantic information is known so far. This example explains why it is said that information always involves a certain amount of regularity. Put in a reverse way, every regularity contains a limited quantity of information. Here order and disorder or information and chaos are each other’s opposites.

A second way to define information is based on a technical relationship for information. It is based on the number of independent degrees of freedom in a source of information. This is called structural information and it is expressed in logons. In its turn, each logon can be expressed in metric or selective information. Usually, metric information is connected to energy and it is expressed in a dimensionless number called mettons. Selective information is related to the content of a message consisting of m possible symbols each with a probability of occurrence.

Note that uncertainty is an event that can occur, and information is the event that has occurred. For an information source the following diagram reviews the different concepts used in information theory.

The base of the logarithmic unit is two. An example can illustrate the concepts.

* * *

Example 1.1

A source produces symbols from an alphabet with 26 different characters with the following probabilities: p = 1/8 for 1 character; p = 1/16 for 8 characters; p= 1/32 for 9 characters; p=1/64 for 4 characters; p= 1/128 for 4 characters. Then the selective information is

$$H=-\frac{1}{8}\text{log}\frac{1}{8}-\frac{8}{16}\text{log}\frac{1}{16}-\frac{9}{32}\text{log}\frac{1}{32}-\frac{4}{64}\text{log}\frac{1}{64}-\frac{4}{128}\text{log}\frac{1}{128}=4.375\text{bit/char}.$$

The maximum amount of selective information is found for a symmetric distribution, hence H_max = log 26 = 4.700 bit/char. Then the redundancy, Red. = (H_max−H)/H_max = (4.7−4.375)/4.7 = 0.0691. Note that when H = H_max then Red. = 0!

* * *

When a certain event occurs a limited amount of information is received. The interrelationship between information and chaos was first shown by Shannon, who has connected the idea of information to entropy. In accordance with the second law of thermodynamics in a closed system the entropy increases towards a maximum, which corresponds to a minimum amount of energy in that system. Hence, entropy is defined as a measure for chaos, or disorder. Information can be quantified accordingly, based on the uncertainty of possible events and their probability of occurrence. On the basis of this concept a smallest amount of selective information is defined. We will not go into further detail here, but reference is made to the relevant literature. In this respect another important definition to distinguish is the concept of knowledge. Knowledge is anything that reduces uncertainty of a system or process. It is the same concept as information. However, we can distinguish knowledge in a different respect. Firstly, we can distinguish perceptual knowledge as parametric knowledge, which is based on by whom, where and when the information is interpreted. Secondly, we can distinguish knowledge revealed from natural laws and not dependent on human interpretation, for instance Newton’s law on gravitation or Ohm’s law. If an experiment is repeated under the same conditions the same results are always found. Finally, we can distinguish knowledge known by intuition. It is the way artists know how to make something, or how sometimes an invention is born. It is just there and the explanation ‘why’ follows (sometimes) later on. A nice example is the discovery of superconductivity in 1911 by Kamerling Onnes explained in 1957 by Bardeen, Cooper and Schiefer (the BCS theory).

Another important point to consider is that information can be bound to six different energy forms. In particular, we have electrical, thermal, radiant, magnetic, chemical and mechanical energy. In general, energy is defined as anything that can perform labour. We will state here that information is always bound to a certain type of energy carrier or mass. For instance, electrical information is connected to electrons, magnetic information is connected to magnetic dipoles. Information can also be printed on paper. Radiant information is connected to photons, mechanical information is connected to mass, and so on. In the electrical energy domain the electrons are the smallest information carriers. How information is connected to electrons is not well understood.

Today everyone is familiar with the fact that information can be measured, handled, stored, retrieved and represented, just depending on its application and requirements demanded of it. Also, information can be converted from one energy form into another by so-called transducers. For measurement purposes this feature is of utmost importance as we will see.

Our world is a very complex one with a tremendous amount of different types of information available, coming from all kinds of process and physical mechanism, often so complex that it is not possible to understand all collected information directly. A good example of such a process is the climate, which is very complex indeed. If statements have to be made about the weather over a particular area and at a specific time, a lot of data are collected concerning temperature, pressure, humidity and wind velocities and directions at different heights. Finally, we make a weather forecast, which always involves a certain amount of uncertainty. This uncertainty is a basic characteristic for every statement which is based on data collected by measurement techniques. The reason for this may be that the process is too complicated, the way the data are collected is imperfect or the amount of data collected is insufficient. In the above example this uncertainty is quite obvious, but errors are inherent in every measurement technique. The result is that we create an image of a part of the world with a limited amount of information. However, despite this implicit imperfection measurements remain the heart of science and technology and without them no progress can be made.

The first thing to do when performing measurements is to characterize the type of process that is necessary to investigate and to determine the physical parameter(s) which are to be measured. Also, any possible interference disturbing the measurement must be isolated from the measurements. This means that criteria must be formulated for every measurement concerning type, response time, accuracy, etc. After the process has been characterized, decisions can be made concerning which type of measurement will be applied and then collection of data can begin. After the collection of the data we have to analyse, synthesize and evaluate the measurements. Finally, an interpretation and check of the results must follow. Collecting data corresponds to receiving information about the system or process. As more information is received, uncertainty concerning the system is decreased.

In this respect again two different kinds of information have to be distinguished. Firstly, in the process under consideration, the type of information must be recogniz...