What is a spectrum analyzer?

What is spectrum? Why do we need to measure spectrum? With these questions, we will learn what spectrum analyzer is.

At the most basic level, a spectrum analyzer can be thought of as a frequency-selective, peak-detecting voltmeter that is calibrated to display the effective value of a sine wave. It should be emphasized that although spectrum analyzers are often used to display power directly, they are not power meters. Of course, if a certain value of the sine wave (such as the peak or average value) is known and the value of the resistor used to measure it is known, the voltmeter can be calibrated to indicate power. The advent of digital technology has given modern spectrum analyzers additional capabilities. This guide describes the basics of spectrum analyzers while also explaining the new capabilities that these instruments can gain using digital technology and digital signal processing techniques.

Frequency Domain vs. Time Domain

Before introducing the spectrum analyzer in detail, readers may ask: "What is spectrum? Why do we need to analyze it?" We are accustomed to using time as a reference to record events that occur at a certain moment. This method is of course also applicable to electrical signals. Therefore, an oscilloscope can be used to observe the change of the instantaneous value of an electrical signal (or other signals that can be converted into voltage by appropriate sensors) over time, that is, to observe the waveform of the signal in the time domain with an oscilloscope.

However, Fourier theory1 tells us that any electrical signal in the time domain can be composed of one or more sine waves with appropriate frequency, amplitude, and phase. In other words, any time domain signal can be transformed into a corresponding frequency domain signal, and the energy value of the signal at a specific frequency can be obtained by measuring it in the frequency domain. Through appropriate filtering, we can decompose the waveform in Figure 1-1 into several independent sine waves or spectral components, which can then be analyzed separately. Each sine wave is characterized by amplitude and phase. If the signal we want to analyze is a periodic signal (as is the case in this book), Fourier theory dictates that the frequency domain spacing of the included sine waves is 1/T, where T is the period of the signal2

Figure 1-1. Composite time domain signal

Some measurement situations require us to examine all the information of the signal - frequency, amplitude and phase. However, even if we do not know the phase relationship between the sinusoidal components, we can still perform many signal measurements. This method of analyzing signals is called signal spectrum analysis. Spectrum analysis is easier to understand and very practical, so we will also introduce how to use a spectrum analyzer to perform signal spectrum analysis in the future.

In order to correctly transform from the time domain to the frequency domain, theoretically the value of the signal at each moment in the entire time range, that is, in the range of positive and negative infinity, must be involved. However, in actual measurement, we usually only take a finite time length.

1. Jean Baptiste Joseph Fourier, 1768-1830. He proposed that any periodic signal can be viewed as a superposition of a series of sine waves and cosine waves.

2. If the time signal occurs only once, then T is infinite and is represented in the frequency domain by a series of continuous sine waves.

According to Fourier transform theory, signals can also be transformed from frequency domain to time domain. Of course, this involves theoretically estimating the values ​​of all spectral components of the signal within the frequency range of positive and negative infinity. In practice, measurements made within a limited bandwidth obtain most of the energy of the signal, and the results are satisfactory. When Fourier transforming frequency domain data, the phase of each spectral component also becomes a crucial parameter. For example, if the phase information is not preserved when transforming a square wave to the frequency domain, the waveform transformed back may be a sawtooth wave.

What is spectrum?

So, what is the spectrum in the above discussion? The correct answer is: the spectrum is a group of sine waves, which, after appropriate combination, form the time domain signal under investigation. Figure 1-1 shows the waveform of a composite signal. Suppose we want to see a sine wave, but it is obvious that the signal shown in the figure is not a pure sine wave, and it is difficult to determine the reason by observation alone.

Figure 1-2. Relationship between signal in time domain and frequency domain

Figure 1-2 shows this composite signal in both the time domain and the frequency domain. The frequency domain plot depicts the amplitude of each sine wave in the spectrum as a function of frequency. As shown, in this case, the signal spectrum consists of exactly two sine waves. Now we know why the original signal is not a pure sine wave, because it also contains a second sine component, in this case the second harmonic. So are time domain measurements obsolete? The answer is no. Time domain measurements are better suited for certain measurement situations, and some measurements can only be made in the time domain. Examples include pulse rise and fall times, overshoot, and ringing, which are all included in pure time domain measurements.

Why measure spectrum?

Frequency domain measurements also have their advantages. As we have seen in Figures 1-1 and 1-2, frequency domain measurements are better suited for determining the harmonic content of a signal.

In the field of wireless communications, people are very concerned about out-of-band radiation and stray radiation. For example, in cellular communication systems, the harmonic content of the carrier signal must be checked to prevent interference with other communication systems with the same operating frequency and harmonics. Engineers and technicians are also very concerned about the distortion of the information modulated onto the carrier.

The interference caused by third-order intermodulation (two different spectral components of a composite signal modulate each other) is quite serious because its distortion components may fall directly into the analysis bandwidth and cannot be filtered out.

Spectrum monitoring is another important area of ​​frequency domain measurements. Government agencies allocate different frequency bands for various wireless services, such as broadcasting, wireless communications, mobile communications, police and emergency communications, and other services. It is critical to ensure that different services operate within their assigned channel bandwidths, which usually requires transmitters and other radiating devices to operate in adjacent frequency bands. In these communication systems, an important measurement for power amplifiers and other modules is to detect signal energy that overflows into adjacent channels and the interference caused by this.

Electromagnetic interference (EMI) is the study of unwanted radiation, intentional or unintentional, from different transmitting devices. What we are concerned about here is that whether it is radiated or conducted (generated through power lines or other interconnects), the interference caused may affect the normal operation of other systems. According to regulations established by government agencies or industry standards organizations, almost anyone engaged in the design and manufacture of electrical or electronic products must test the relationship between radiation levels and frequencies.

Figure 1-3. Transmitter harmonic distortion test Figure 1-4. GSM radio signal and spectrum emission mask showing the limit of unwanted radiation Figure 1-5. Signal emission measurement results compared to CISPR11 limit values ​​in two-tone test of RF power amplifier

Noise is often measured. Any active circuit or device will generate additional noise. Measuring the noise figure and signal-to-noise ratio (SNR) can describe the performance of the device and its impact on the overall system performance.

Figures 1-3 through 1-6 show several examples of these measurement applications using the X-Series signal analyzers.

Signal Analyzer Types

The original swept-tuned superheterodyne analyzer could only measure amplitude. However, as technology continued to evolve and communication systems became more complex, phase became increasingly important in measurements. Spectrum analyzers, while still known as signal analyzers, have evolved into their own class of instruments. By digitizing the signal, both phase and amplitude information can be retained and displayed after one or more frequency conversion steps. As a result, current signal analyzers, such as the Keysight X-Series, combine the features of analog, vector, and FFT (Fast Fourier Transform) analyzers. To further improve functionality, the Keysight X-Series signal analyzers incorporate a computer and are equipped with a removable disk drive, which allows sensitive data to remain in a secure area even if the analyzer is moved to an unsecured location.

Technology has also led to miniaturization of instruments. As a result, engineers can more easily perform outdoor measurement tasks, such as transmitter or antenna site surveys, with rugged portable spectrum analyzers such as Keysight FieldFox. In situations where a short stop is required to make a quick measurement, analyzers with zero warm-up time allow engineers to get to work as quickly as possible. By applying advanced calibration techniques, these handheld analyzers can perform field measurements with an accuracy within one-tenth of a dB of laboratory-grade benchtop spectrum analyzers.

Note: As computers became HP's primary business, HP created the independent Keysight Technologies in the late 1990s and moved its test and measurement business into Keysight. Therefore, many older spectrum analyzers are HP-branded but supported by Keysight.