An Overview of Frequency-Domain Measurements

Whether you are measuring a commercial radio wave, doing FFT analysis on an audio signal, or performing mass spectroscopy for chemical analysis, analyzing signals in the frequency domain and performing spectral analysis yields important information not otherwise available.

Introduction

The time and frequency domains, which are related to each other, are the most common ways of viewing, manipulating, and analyzing any signal of interest. You can use fast Fourier transforms (FFTs) and inverse FFTs to convert from one domain to the other. In this paper, we take a look at the basics of frequency-domain analysis and its common applications.

To understand the different types of applications that require frequency-domain analysis, we begin with a look at the types of applications that fall within different frequency bands. While the current usable spectrum extends into the hundreds of GHz thanks to applications such as satellite broadcasts and fiber optics, much of the world still operates at frequencies under 100 MHz.

Frequency: Application:
0 to 20 kHz Audio
100 kHz to 1 MHz AM radio
1 to 30 MHz Shortwave and military radio, ultrasound, mass spectroscopy
30 to 100 MHz FM radio, radar, lidar, high-energy physics

Some common frequency-domain measurements are frequency response, cross power spectrum, power spectrum, power spectral density, zoom FFT, adjacent-channel power, and occupied bandwidth. To acquire signals in these various applications, National Instruments offers high-performance, modular hardware that tightly integrates with LabVIEW and specific toolkits such as the Sound and Vibration Toolkit. For example, NI 4472 devices offer 24-bit dynamic signal acquisition (DSA) and are specifically designed to provide outstanding dynamic range and resolution for measuring audio signals. The NI PXI-5620 digitizer provides outstanding dynamic range and distortion-free performance for signals up to 32 MHz.

Frequency Response

You can display the frequency response of a system two different ways – a Bode plot or a Nyquist diagram. Both display the same information, but present it in different ways. In general, the frequency response is simply the response of a system to sinusoidal inputs at different frequencies. A linear system outputs a sinusoid of the same frequency but different magnitude and phase. The frequency response is the magnitude and phase differences between the input and output sinusoids across the frequency range of interest.

Power Spectrum

The power spectrum is a representation of the magnitude of the various frequency components of a signal. Evaluating the power spectrum is a way to isolate periodic features or noise. By looking at the power spectrum, you can see how much energy (power) is contained in the different frequency components of the signal.

Zoom FFT

When you are performing frequency-domain analysis, a zoom FFT is a useful measurement tool for zooming in on a narrow frequency band. If you are analyzing signals that require high sampling rates and resolution, you may find that the data size of a wide FFT is unmanageably large. You can instead use a zoom FFT to focus on a narrowband channel where transmission is occurring, thereby performing only the calculations needed to obtain the FFT data in the frequency range of interest.. Zoom FFTs are useful in many applications where narrowband signals are prevalent. For example, many digital communications standards use bandwidths of 25 MHz or less, making them good candidates for this type of analysis. Other applications include Doppler radar, mechanical stress analysis, and ultrasonic analysis in the medical field.

Adjacent-Channel Power

Many technologies allocate adjacent channels for information distribution from different providers, such as cell phones, TV, radio, and cable. In these and other applications, it is important that transmission from one channel does not cross over to an adjacent channel, which noticeably degrades the quality in the other channel.

Depending on the technology standard you are measuring, there are different criteria for adjacent-channel power measurements. For example, the CDMA wireless standard requires transmissions to fit within a 4.096 MHz bandwidth. Moreover, adjacent-channel power, measured at 5 MHz offsets, must be at least 70 dB below the in-channel average power.

Reader Comments | Submit a comment »