IEEE© 1997 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

By choosing to view this document, you agree to all provisions of the copyright laws protecting it.


A Course on Instrumentation: The Signal Processing Approach

Alfonso Carlosena and Rafael Cabeza


Abstract - This paper describes a new course on instrumentation given to telecommunications engineering students at the Universidad Pública de Navarra. This course has been introduced within the context of the renovation of curricula in Spain, in which instrumentation has become a mandatory subject. Instrumentation is thus presented as a relevant discipline on its own in which many technologies and techniques converge. A system approach is used in the course to emphasize signal and system theory concepts to students.



I. Introduction

In 1983 the Spanish Parliament approved the Law for the Reform of Universities (LRU), which was intended to produce a major change in the structure of universities and in the curricula they offer to the students, as well. The Spanish universities started to adapt to the new structure while maintaining the curricula unchanged until 1987, when the government decided on a general framework for defining the new official degrees in the universities. The guidelines given can be summarized as follows [1]:

Mandatory subjects form the basis, and cover most credits, of a given curriculum. While they must be followed by any student wishing to obtain a particular degree (i.e. physics, computer science,...), two types of mandatory courses can be distinguished: core and obligatory courses. The former refers to courses common in all universities offering the same degree, while each university defines the latter. The rationale for this approach is to unify an important portion of the curriculum for a particular degree all over Spain so that students can move easily from one university to another. At the same time, universities have some freedom to define their own curricula. Each university also specifies optional courses and therefore permits the students some choice in defining their own specialization or emphasis within the selected degree. Finally a number of credits is allotted to subsidiary courses about subjects that may not be directly related to the curriculum.

The Universidad Pública de Navarra (State University of Navarra) was established in 1987 and therefore was created within the new framework. One of the degrees offered to students (since 1989) is telecommunications engineering. At the Universidad Pública de Navarra, it has been offered according to the new guidelines published in 1988, while at other older universities the same degree was offered according to the previous scheme. The older universities thus faced an unavoidable process of adapting to the new guidelines. Note that during this transition time, most degrees offered in Spain preserved the name, and even the basic orientation, of the former ones but changed to the new scheme in terms of subject distribution, teaching hours, etc.

The traditional Spanish degree in telecommunications engineering has been close in concept to electrical engineering degrees in Anglo-Saxon countries, but with some remarkable differences. In particular, basic subjects closer to mechanical engineers, such as control theory and the science of materials, are not present. In contrast, emphasis is put on signal and system theory, electronics, transmission media, computer architecture and networks and, of course, communication theory. While communications (or telecommunications) engineering was considered in many universities as a specialization of electrical engineering, it is not surprising that the area is today so broad that such a curriculum may be even considered as a basic discipline [2].

The renovated curricula in telecommunications engineering brought along some relevant changes with respect to the previous scheme, one of the most remarkable novelties being the inclusion of a core subject called "Electronic Instrumentation". Most Spanish universities were offering instrumentation courses to their students taking a degree in telecommunications engineering, but such subjects were elective and their duration and contents varied widely depending on the university. The inclusion of the new "Electronic Instrumentation" course as mandatory is a recognition that instrumentation itself is a relevant part of an engineer's background, and not simply an auxiliary discipline. While this approach has been common in most European universities, it is not, with very few exceptions [3], in institutions within the USA.

Taking into account the above considerations, this paper reports the experience at the State University of Navarra with this new subject. We had the challenge to define, from scratch, the new course and auxiliary laboratory, and in a very short time to implement it in a newborn university. The concept of instrumentation is so broad that a careful selection of topics had to be accomplished while taking into consideration aspects such as time limitations, students' background, and skills required for a telecommunication engineer, but never forgetting that an overspecialized education is no longer considered adequate. We believe that our experience can be helpful to other colleagues facing the always challenging task of defining new courses on instrumentation, which are being incorporated, more and more, in many curricula.



II. Course Context

The courses that entitle a student to obtain a degree are distributed in Spain, with very few exceptions, over five academic years. The total number of credits required is about 370 in telecommunications engineering, which means an average of about 70 credits per year (the exact figure depends on the institution). A minimum of 60 semester hours (six credits) is assigned to Electronic Instrumentation by law, and its location must be in either the fourth or fifth academic year. The State University of Navarra decided not to increment the number of credits and to place the course in the second term of the fourth year. The six credits have been further divided into two halves for lectures and practical (laboratory) work.

When the student faces the course on instrumentation she (or he) has received a sound basis in circuit theory (12 credits), analog and digital electronics (35), signal and system theory (both continuous and discrete) (15), communication and information theory (18) and computer architecture and networks (12), in addition to basic courses in mathematics and physics. Besides the lectures in the classroom, they have also attended many sessions in the laboratories, using typical low cost instrumentation (analog oscilloscopes, signal generators,... and even spectrum analyzers) and standard packages such as Matlab or Pspice running on personal computers. They also have some background in measurement error theory. In parallel with the course on electronic instrumentation, our university offers an elective course called Transducers and Signal Conditioning, which is obviously directly related to instrumentation.

The above data give information about the students' background but are not enough to define the course contents and orientation. The always difficult decisions to disregard some topics and to limit the length and emphasis of the ones included is particularly difficult in the case of instrumentation due to the expanse of the subject and also the lack of experience. The following considerations help to bound the problem and finally find a rational solution.



III. Course Content and Signal Processing Concepts

The detailed syllabus of the course is shown in table 1. It is divided into four major sections, each one containing a number of lessons of one hour duration. The title of each lesson suggests the topics treated. The sum of the hours is 27, which are complemented with 3 hours of discussion in which specifications and characteristics of some selected commercial instruments are analyzed. In this way the 30 hours allotted for classroom sessions are covered.

The first, introductory, section is devoted to a general overview about instrumentation, serving as motivation for the remainder of the course and justifying the approach taken. The second part deals with instruments for time domain signal analysis, exemplified by the oscilloscope in its two versions: analog and digital, with particular emphasis on the second. This section is followed by the longest section of the course, on frequency domain analysis, which is further divided into filter analysis (analog and digital), Fourier (dynamic) analysis and sweep techniques. Interactions and combinations of the techniques mentioned are shown, as well as their extension for system characterization, exemplified by network analyzers.

Some time is devoted to the increasingly interesting time-frequency techniques (sometimes called the modulation domain in the instrumentation context) arising from multiresolution filtering or short time Fourier transforms. The fourth section concentrates on the two most relevant instrumentation oriented buses: IEEE-488 and VXI which serve to emphasize the idea of synergy between different modules, described in the previous sections, and computers (whatever their physical form) to build more complex or specific instruments. PC buses are also shown as a widely used solution to build instruments from an inexpensive platform.

Table 2 illustrates how the course emphasizes system and signal processing concepts. The first column displays an instrumentation issue and the second the basic concepts which are invoked. The list of issues is not intended to be exhaustive, nor are the concepts associated exclusively with a particular instrumentation topic. The table simply indicates the broad range of concepts that can be covered and thus emphasized within the subject of instrumentation.



IV. Description of Practical Sessions

The laboratory used for the course described in this paper consists of 14 stands composed of the following items:

All instruments include an IEEE-488 interface. The oscilloscope was selected because of its dual analog and digital nature, which allows a straightforward comparison between the two working modes. The signal generator provides a number of built-in signals from typical periodic sine or square waves to FM, and in addition has the capability to produce a user defined signal, a very interesting feature in our context. The combination of both instruments gives the capability to generate signals, to excite a system for instance, and also to acquire the desired response(s) with the two oscilloscope channels. The I/O card incorporates both capabilities in one device, although not simultaneously. The instruments and card, connected to the computer through the GPIB bus and the PC bus, respectively, can be combined in different ways: from a simple oscilloscope built up with the PC and the I/O card, to a dual channel network analyzer made with the signal generator and the digital oscilloscope.

LabView has been selected for many reasons. First of all, it can be considered as the "de facto" standard language for instrumentation programming. Second, a student version is available so that many can make use of it at home. Third it is graphical and intuitive. Only a short training time is required before students can write their first simple programs. Fourth, communication with most commercial cards and the GPIB bus is implemented. And fifth, the language fits hand-in-glove with our systems approach. LabView contains an endless number of built-in signal generation and processing functions that allow straightforward construction of any complex system.

The design of a network analyzer is one experiment typically suggested to the students as a project that serves to test their achievement. They can freely select between different possible techniques (sweep, dynamic,...), different exciting signals (multisine, step, noise ..), and different instrument and card combinations. They must decide on an approach, develop a complete design based on that approach, and compare the advantages and limitations of the selected approach with those of other approaches developed by their colleagues. Prior to assignment of the project, students are trained in the use of the different devices found in the laboratory and of course in the graphical language on which LabView is based. At this point, we have found one of the most important limitations of the course: it is difficult to fit the real number of hours needed to the theoretical course duration.

This academic year, the network analyzer based on a multisine excitation signal was most frequently selected by the students. The distinguishing characteristic of this kind of analyzer is the input signal, which is composed of different harmonics of a fundamental frequency. By choosing this fundamental frequency, the student is able to determine the frequency range of measurement. The relative phases between harmonics are selectable by a control on the front panel of the virtual instrument. Other parameters that can be adjusted are: the number of frequency points, the peak-to-peak amplitude of the input signal and the sampling frequency of signal generator. The difficulty of this project is medium, as can be seen in its block diagram, but the concepts involved in its resolution are quite fundamental.

On the other hand, knowledge acquired during practical sessions is very useful in later work, especially in M.Sc. projects. Many students use LabView as a tool in her/his M.Sc. work and many of them use her/his own network analyzer to carry out experimental measurements in the laboratory.



V. Conclusions

The course described in this paper has been given for three years now and our experience is very positive. According to students' opinions, they find the course very challenging because they are forced to work hard on an almost real design problem, in contrast to other courses where they solve canned problems or experiments leading to a single solution. Regarding the course contents, they show an almost unanimous preference for the last part of the course, where they find the possibility of combining different devices to build an entirely new instrument fascinating. For us, their teachers, it has been very illustrative to discover the points or topics in basic disciplines that are not well assimilated by the students in previous courses. One that stands out above the others is the way in which digital instruments represent inherently continuous signals, namely, the connection between continuous- and discrete-time system theory. This shortcoming has been also pointed out by other authors [12].



Acknowledgments

The Gobierno de Navarra and the CICYT are gratefully acknowledged for financial support.



References

[1] S. Dormido, M.G. Hartley, Editorial, International Journal of Electrical Engineering Education, 28, 4-5, 1991

[2] D.C. Coll, "Communications Engineering: A New Discipline for the 21st Century", IEEE Transactions on Education, 37, 2, 151-157, 1994

[3] J.A. Orr, B.A. Eisenstein, "Summary of Innovations in Electrical Engineering Curricula", IEEE Transactions on Education, 37, 2, 131-135, 1994

[4] P. Stein, "The Unified Approach to the Engineering of Measurement Systems for Test and Evaluation-A Brief Survey", IEEE Instrumentation and Measurement Technology Conference, KeyNote Address, Brussels, Belgium, June 1996

[5] A. Barwicz, "System Approach to Electrical Measurements", IEEE Instrumentation and Measurement Technology Conference, 397-402, Irvine, California, May 1993

[6] R. Pallás, "Instrumentation in Education and Education on Instrumentation", Mundo Electrónico, 154, 71, 75, 1985 (In Spanish)

[7] L. Finkelstein, "General Principles of Formation in Measurement Science and Technology" in P.H. Sydenham (ed) Handbook of Measurement Science, Vol 3, 1417-1431, Wiley 1992

[8] J. Schoukens, M. Vanden Bossche, R. Pintelon, "Expert Opinion: the Key to Better Measurement is Software", IEEE Spectrum, 30, 1, 55, 1993

[9] D.C. Hanselman, "Signals and Linear Systems: A Teaching Approach Based on Learning Style Concepts", IEEE Transactions on Education, 35, 4, 383-386, 1992

[10] F.S. Barnes, Editorial of Special Issue on Advances in Electrical Engineering Education, IEEE Transactions on Education, 37, 2, 1994

[11] L. Finkelstein, "Design Orientated Teaching of Measurement and Instrumentation", Measurement, 4, 89-92, 1986

[12] J.R. Deller, "Tom, Dick and Mary Discover the DFT", IEEE Signal Processing Magazine, 11, 2, 36-50, 1994



Table I

Course Syllabus

I-INTRODUCTION

II- INSTRUMENTATION FOR TIME DOMAIN ANALYSIS

III- INSTRUMENTATION FOR FREQUENCY DOMAIN ANALYSIS

IV- INSTRUMENT INTERCONNECTION. VIRTUAL INSTRUMENTATION



Table II

Instrumentation and Signal Processing Concepts

Analog Bandwidth

Frequency Response
Rise Time vs. Cutoff Frequency

Digital Bandwidth

Sampling Theorem. Aliasing
Theoretical vs. Practical Sampling Freq.

Vertical Resolution

Quantization Levels = #bits
Quantization Noise

EBR (Effective Bit Resolution)

Conversion Nonlinearities

Equivalent Sampling

Subsampling
Nonuniform Sampling

High Resolution Modes

Oversampling
Noise Modulation

Signal Representation

Reconstruction Filtering
Interpolation

Filter Bank Analysis

Filter Concepts
Shape Factor, Transient Time
Frequency Resolution

Octave Filter Banks

Multiresolution Analysis
Multirate Signal Processing

Dynamic Analysis

DTFT, DFT, FFT
Windowing, Leakage, Picket Fence

Real Time Analysis

Short-Time FT
Time-Frequency Analysis

Band Selectable Analysis

Destructive and Non-destructive Zoom
Decimation in Time. Discrete Mixing

Swept Spectrum Analyzers

Mixing, Heterodyning, IF Filters
Up- and Down-Conversion
Phase Noise, Residual FM

Dual (Dynamic) Channel Analysis

Frequency and Impulse Response
Auto- and Cross Correlation



Contact Information

Alfonso Carlosena
Universidad Pública de Navarra
Dpt. Ingeniería Eléctrica y Electrónica
Campus de Arrosadía
E-31006 Pamplona,
Spain
Phone: +34 48 169 329
Fax: +34 48 169 204
E-mail: carlosen@upna.es

Rafael Cabeza
Universidad Pública de Navarra
Dpt. Ingeniería Eléctrica y Electrónica
Campus de Arrosadía
E-31006 Pamplona, Navarra
Phone: +34 48 169 329
Fax: +34 48 169 204
E-mail: rcabeza@tsc.upna.es



Biographies

Alfonso Carlosena was born in Navarra, Spain. He received the M.Sc. with Honors and the Ph.D. in physics in 1985 and 1989, respectively, from the University of Zaragoza, Spain. From 1986 to 1992 he was an assistant teacher in the Department of Electrical Engineering and Computer Science of the University of Zaragoza. He spent the winter semester of 1989-90 and 91-92 as an invited researcher at the Institute of Signal and Information Processing (ETH-Zürich). Since October 1992 he has been an associate professor at the Universidad Publica de Navarra where he also serves as head of the Technology Transfer Office. His research interests are in the area of classical circuit theory, analog signal processing and active devices, CAD tools and instrumentation.

Rafael Cabeza was born in Soria, Spain. He received the Licenciado en Fisica degree in 1990 from the Universidad of Zaragoza, Spain. In 1996, he received the Ph.D. in telecommunication engineering from the Universidad Publica de Navarra. Since 1992 he has been working at the Dpt. de Ingenieria Electrica y Electronica of the Univeridad Publica de Navarra. His main research interests are circuit theory, analog integrated circuits and instrumentation.