We use an oscilloscope to view variations of input voltage with time. We use a spectrum analyzer to view variations of input voltage with frequency. Measurements in the frequency domain are possible with traditional architectures such as the super heterodyne, swept-tuned spectrum analyzer. Earlier such instruments were made purely with analog components. Modern instruments have evolved with digital elements such as Analog to Digital Converters (ADCs), Digital Signal Processors (DSPs) and microprocessors.
However, the basic principle of working remains the same, and for observing controlled, static signals, it is the best suited. The signal analyzer makes amplitude vs. frequency measurements. The signal of interest is down-converted and swept through the passband of a Resolution Bandwidth Filter (RBW). The signal then passes through a detector that calculates the amplitude of each frequency point in the selected span.
Although this method provides a high dynamic range of measurements, it can only calculate the amplitude data for one frequency point at a time. This assumes that while the analyzer is making one sweep, there is no significant change to the signal being measured. Therefore, the measurements are suitable for relatively stable input signals that remain unchanging. If there were fast changes in the signal, statistically there would be a probability that some signals were missed.
This spectrum analyzer architecture therefore, does not provide a reliable way to discover transient signals, which leads to a prolonged time and effort for troubleshooting modern RF signals; this lead to the development of the Real Time Spectrum Analyzer.
To analyze signals in real-time, therefore, the analysis must be carried out fast enough so that all signal components in the frequency band of interest are accurately processed. For this, two things are necessary. First, the sampling of the input signal must be fast enough and satisfy the Nyquist criteria, which implies the sampling frequency must be more than two times the bandwidth of interest.
Second, all computations must be performed fast enough and continuously so that the output of the analysis always matches and keeps track of the changes in the input signal.
The architecture of the Real-Time Spectrum Analyzer or RSA is so designed that it can overcome the measurement limitations of the simple Spectrum Analyzer (SA). The RSA addresses the challenges that transients and dynamic RF signals pose to signal processing by the SA. The RSA does this by real-time digital signal processing before the storing the results in memory.
This way, the user is able to see and capture transient events invisible to other types of instruments, and he can trigger on such events allowing them to be selectively captured in the memory. The data in the memory can then be separately analyzed extensively in many other domains with the help of batch processing. The real-time DSP engine is also helpful in signal conditioning and calibration of many types of analysis.
Modern RSA architecture uses a combination of both analog and digital signal processing for converting RF signals into measurements that are calibrated and time-correlated. Input RF signals are converted into analog IF signals that are filtered by a bandpass filter before they are digitized.