Key Spectrum/Signal Analysis Settings

There are a wide variety of Signal/Spectrum Analyzers on the market that use a wide range of methods and technologies. Most modern spectrum analyzers use Analog-to-Digital Converters (ADCs) to convert the time-domain signals input at the port of the spectrum analyzer to frequency domain data using Fast Fourier Transforms (FFTs). Some spectrum analyzers, typically those that operate at several gigahertz to tens of gigahertz use a heterodyne architecture that provides frequency downconversion of higher frequency signals prior to digital conversion and processing. Others use direct digital conversion, or digital-IF, to convert RF/Microwave signals directly into digital data to be processed. The complexity and capability of the conversion and digital signal processing electronics directly impacts the bandwidth, speed, and other performance factors of the spectrum analyzer.

Other factors that impact the performance of a spectrum analyzer are the exact methods of sampling and the amount of memory and digital processing available to process incoming signals. Deciding these factors often leads to trade-offs with resolution and measurement sensitivity, which is why some spectrum analyzer software allows for the ability to control, at least to some extent, these features.

With FFT analyzers, the data that is captured and processed must be taken during a sample. If some portions of the energy of a signal exist outside of a capture window, then that signal will not be accurately represented. There are methods of staggering or even overlapping FFT windows, to enable the accurate capture of signals that might be extremely brief, brief enough to be beyond human perception. With overlapping FFTs, very short signals can be captured, but the burden of processing and stitching together the data from the overlapping FFTs requires substantial data transfer rates and digital signal processing capability. Storing such signal data, depending on the bandwidth, may also be substantial.

These factors relate to the real-time bandwidth of a spectrum analyzer. Essentially, the real-time bandwidth is the frequency range that a spectrum analyzer can capture and process in real-time without losing data. This is not to be confused with the resolution bandwidth (RBW), which actually relates to the sample capture time and the frequency resolution. For example, a smaller RBW range results in higher frequency resolution, but longer capture time. The result of this is generally greater measurement sensitivity with the tradeoff of longer sweep times.

Nor should real-time bandwidth be confused with the video bandwidth (VBW), which provides a smoothing of amplitude of the capture signals using digital filtering and digital signal processing methods. For some spectrum analyzers, the processing done by the VWB filters can be user controlled to provide log, power, or voltage detection and additional processing features.

Some spectrum analyzers also have attenuators and amplifiers in the spectrum analysis signal chain. Depending on the setting of these attenuators, the signal noise can be reduced, or stronger signals can be measured. Generally, increasing the gain within the signal chain increases the noise, possibly leading to increased nonlinearities, but may also enable detection of weaker signals.