Spectrum and Network Measurements (2nd Edition)
This updated edition of the industry's classic text combines the theory, practice, and latest technology of spectrum and network measurements in electronic systems to offer a comprehensive and easy way to understand frequency domain measurements. Spectrum and Network Measurements, 2nd Edition has been completely updated to take into account the latest technology, particularly focusing on the shift from analog to digital in communication systems, plus an important new chapter on EMC measurements of radiated and conducted emissions has also been added. Using simplified block diagrams, this book offers a thorough coverage of critical concepts, such as decibels, Fourier analysis, noise effects, impedance matching, intermodulation distortion, reflection measurements, and SParameters. By picking up where the majority of electrical engineering programs stop, this title takes basic EE theory and translates it to the world of electronic measurements. This enables the reader to understand the basic theory of signals and systems, relate it to measured results, and apply it when creating new designs.
Inspec keywords: network analysers; fast Fourier transforms; pulse measurement; spectral analysers; electromagnetic compatibility; electric noise measurement; transmission lines; filtering theory; twoport networks; distortion measurement
Other keywords: fast Fourier transform analyzer; Fourier theory; spectrum measurement; distortion measurement; EMC measurement; modulation measurement; transmission lines; filtering; Decibels; twoport network; swept spectrum analyzer; network analyzer; noise measurement; vector network measurement; pulse measurement; averaging
Subjects: Network and spectrum analysers; General electrical engineering topics; Integral transforms; Textbooks; Electromagnetic compatibility and interference; Electric noise and interference measurement; Filtering methods in signal processing
 Book DOI: 10.1049/SBEW506E
 Chapter DOI: 10.1049/SBEW506E
 ISBN: 9781613530146
 eISBN: 9781613530368
 Page count: 359
 Format: PDF

Front Matter
 + Show details  Hide details

p.
(1)

1 Introduction to Spectrum and Network Measurements
 + Show details  Hide details

p.
1
–10
(10)

2 Decibels
 + Show details  Hide details

p.
11
–21
(11)
Decibels are used to specify ratios of powers and voltages in a logarithmic fashion. Absolute levels can also be specified by supplying suitable reference values. Decibels are commonly used for gain and loss calculations in electronic systems. Generally, spectrum and network analyzers display measurement results with their displays calibrated in decibels. The popularity of the decibel in such applications is due to its ability to compress logarithmically widely varying signal levels. For example, a 1 V signal and a 0.1 mV signal can both be represented on a display with 100 dB of range. To show these two signals simultaneously with reasonable clarity on a linear scale is impractical. Decibels also are useful for gain and loss calculations, where multiplication operations are transformed into (easier) additions.

3 Fourier Theory
 + Show details  Hide details

p.
23
–41
(19)
The most common way of representing signals is in the time domain. Another representation of a signal is via the frequency domain, which is inherent in spectrum measurements. In the frequency domain, the signal is described in terms of its frequency content, plotting the amount of power present at each frequency. A complete frequency domain representation includes both the magnitude and phase of the signal. The frequency domain is related to the time domain by a body of knowledge generally known as Fourier theory, named for Jean Baptiste Joseph Fourier (17681830). This includes the series representation known as the Fourier series and the transform techniques known as the Fourier transform. Discrete (digitized) signals can be transformed into the frequency domain using the discrete Fourier transform (DFT).

4 Fast Fourier Transform Analyzers
 + Show details  Hide details

p.
43
–85
(43)
The fast Fourier transform (FFT) can be used to implement a spectrum analyzer by digitizing the input waveform and performing an FFT on the time domain signal to obtain the frequency domain representation. What seems to be a simple measurement technique often turns out to be much more complicated in practice. Given reasonable computational power, usually in the form of a digital signal processor (DSP), field programmable gate array (FPGA) or custom integrated circuit, the FFT analyzer can provide significant speed improvement over the more traditional swept analyzer. The classic FFT analyzer (also called a dynamic signal analyzer) covers the frequency range from DC up to a few hundred kilohertz. These analyzers are typically applied to audio and mechanical measurements.

5 Swept Spectrum Analyzers
 + Show details  Hide details

p.
87
–105
(19)
The traditional method for implementing a spectrum analyzer is the swept heterodyne block diagram. Similar to a radio receiver, the spectrum analyzer is automatically tuned (swept) over the band of interest. This type of analyzer has been gradually replaced by the fast Fourier transform (FFT) analyzer at low frequencies, but the swept analyzer remains the dominant technology in the radio frequency range and above. In recent years, the swept analyzer has been combined with the FFT analyzer to provide the advantages of both techniques.

6 Modulation Measurements
 + Show details  Hide details

p.
107
–135
(29)
Ever since the early days of radio, modulation techniques have played an important part in electronic communications. Typically, a lowfrequency voice or data signal is used to modulate some characteristic of a carrier signalusually the amplitude, phase, or frequency. Over time, the digital forms of modulation have become dominant. Digital modulation often is implemented using vector modulation of the carrier signal, having an inphase (I) and a quadrature (Q) modulation component. Communication systems represent an intentional use of modulation, but there are also incidents of unintentional modulation, such as power line sidebands on an oscillator output or residual frequency modulation on an amplitudemodulated signal. Whether the modulation is intentional or not, a spectrum analyzer can be used to characterize and measure it.

7 Distortion Measurements
 + Show details  Hide details

p.
137
–149
(13)
Many electronic circuits and systems are considered to be linear time invariant (LTI), as introduced in Chapter 1. For a sinusoidal input the output of an LTI system is also sinusoidal with perhaps a different amplitude and phase. In the time domain, the output waveform is a sinuosoid, the exact same shape as the input waveform. In the frequency domain, we expect to see at the output the same frequency that was at the input (and only that frequency). Any other frequencies that are generated due to the input signal are considered to be distortion.

8 Noise and Noise Measurements
 + Show details  Hide details

p.
151
–165
(15)
In frequency domain measurements, electronic noise shows up in two distinctly different ways. The first case is when the measurement is affected by the presence of unwanted noise, with noise being a nuisance. For example, we could be measuring the distortion of an amplifier with the amplifier's noise degrading the measurement. The second case occurs when the noise present in the system is the parameter to be measured. In that same amplifier, we may want to measure the noise at the output. Many of the same principles apply to both cases, but it is important to know whether the noise is the measurement or whether it degrades the measurement. The electronic noise present in our measurements may come from the device under test (DUT) that is being measured or may be generated internally by the analyzer. In the general case, the analyzer internal noise must be significantly lower than the noise of the DUT. However, techniques that compensate for the noise in the analyzer can lower the measurement floor of the analyzer.

9 Pulse Measurements
 + Show details  Hide details

p.
167
–176
(10)
Pulsed waveforms are an important class of signals in systems such as radar and digital radio. Pulsed signals can present a more difficult measurement problem than continuous waveforms. The resolution bandwidth used in a measurement can affect the displayed spectrum. With a smallresolution bandwidth, the displayed spectrum has discrete spectral lines, but with wider wideresolution bandwidths these line spectra are smeared together and the spectrum appears to be continuous. The principles associated with the pulsed waveform are also applicable to pulsed radio frequency signals. The envelope of the spectrum is the same and depends on the pulse width, but the spectrum is centered on the radio carrier frequency.

10 Averaging and Filtering
 + Show details  Hide details

p.
177
–193
(17)
Many analyzer measurements have considerable amounts of noise present in them. The noise is often undesirable but may actually be a desired component of the measurement. Two basic techniques are used to reduce the noise: filtering and averaging. Filtering usually takes the form of an analog filter. However, it can also be implemented in digital form, whereas averaging is always done digitally. The two concepts are closely related and are treated here in a unified manner. Both filtering and averaging can be classified as either predetection (before the detector) or postdetection (after the detector). Predetection averaging/filtering reduces the noise present in a measurement, while postdetection averaging/filtering reduces the amount of fluctuation in the noise.

11 Transmission Lines
 + Show details  Hide details

p.
195
–217
(23)
Transmission lines are commonly used to connect test and measurement instruments to the device under test. Transmission lines are used to control the effects of inductance and capacitance, which are unavoidable in highfrequency systems. Coaxial cables are the most common transmission lines, providing shielding of the signals being measured. Measurement error can be introduced due to impedance mismatch at either end of a transmission line. These errors must be understood and minimized to ensure an accurate measurement.

12 Measurement Connections
 + Show details  Hide details

p.
219
–239
(21)
Connecting an instrument to a device under test (DUT) invariably involves disturbing that device. When making precision measurements, it is desirable to minimize loading and other effects so that the measurement is not corrupted by the measuring instrument. Probes, attenuators, impedance matching devices, and filters are used to couple the signal of interest into the instrument in the most efficient and accurate manner.

13 TwoPort Networks
 + Show details  Hide details

p.
241
–251
(11)
Twoport network theory provides the theoretical basis for network measurements. Twoport network theory can be expanded to Nport theory for networks having more than two ports, while oneport measurements are a subset of twoport measurements. The simplest of twoport measurements is the gain or transfer function of the device. This assumes a fairly simple model of the device under test (DUT). More complete twoport models such as impedance parameters provide a more complete view of device behavior, while scattering parameters present a twoport model that is consistent with transmission line theory and measurements.

14 Network Analyzers
 + Show details  Hide details

p.
253
–265
(13)
Network measurements can be divided into two types: transmission through the network and reflection at the network's input or output port. Full twoport network analysis normally requires the use of a multichannel network analyzer and a scattering parameter (Sparameter) test set. Simpler measurements, such as transmissiononly measurements, can be performed with less sophisticated equipment.

15 Vector Network Measurements
 + Show details  Hide details

p.
267
–296
(30)
Network measurements characterize the transmission through the device under test (DUT) and the reflections at each port. The device under test can have any number of input and output ports, but we'll focus on classic twoport measurements. For distortionless transmission through a device, the output signal must be identical to the input signal, perhaps delayed in time and scaled in amplitude. This implies the device that must have a flat amplitude response and a linear phase response. Group delay is the derivative of phase, which provides a useful way to view time delay through a device. The use of scalar network measurements has decreased over time, and most measurements are now vector, including magnitude and phase. Vector error correction is a powerful technique for reducing measurement error, especially at high frequencies.

16 EMC Measurements
 + Show details  Hide details

p.
297
–314
(18)
In most countries, electronic products are required to meet established standards for electromagnetic compatibility (EMC). These regulations attempt to allow the wide array of electronic devices to live together in electromagnetic peace and harmony. In this chapter, we'll take a look at using spectrum analyzers to measure two important aspects of EMC: radiated emissions (signals radiated from the device under test [DUT]) and conducted emissions (signals conducted via the power cable from the device under test).

17 Analyzer Performance and Specifications
 + Show details  Hide details

p.
315
–324
(10)
Spectrum and network analyzer specifications are the instrument manufacturer's way of communicating to the user the level of performance that should be expected from a particular instrument. Understanding and interpreting instrument specifications enable the instrument user to predict how the instrument will perform in a specific measurement situation, including the accuracy of the measurement. The form and style of the specifications are usually related somewhat to the block diagram and measurement techniques internal to the instrument. These specifications may appear to be more complex than necessary. However, oversimplifying an instrument data sheet can force the manufacturer to understate the performance level of an instrument to cover all possible cases in a single specification. In general, the details present in analyzer data sheets provide a better understanding of instrument performance, so that the user can obtain the best measurement possible.

Appendix A: TwoPort Vector Error Correction
 + Show details  Hide details

p.
325
–327
(3)
For the most accurate network measurements, vector error correction is employed, as discussed in Chapter 15. In this Appendix, we will examine the twoport model in more detail. Figure A1 shows the forward error model for the twoport error correction. For simplicity, only half of the error terms for the twoport model are shown: the ones relevant to forward measurements. The model shown is sufficient for errorcorrected S_{11} and S_{21} measurements. There is a corresponding error model for the reverse measurements, S_{22} and S_{12}. S_{11M} is the measured version of S_{11A}, which is the actual S_{11} value for the device under test (DUT). The error terms included in the equation are all from the left column of Table A1, which means the forward error model is sufficient to describe the measured result for S_{11}.

Back Matter
 + Show details  Hide details

p.
(1)
Supplementary material

Errata

Click on the icon below for the latest errata for this book.

PDF
97.0224609375Kb

PDF
