How to Correctly Set the Bandwidth of a Spectrum Analyzer

In measuring some CATV system indicators, spectrum analyzers are often used. In order to make the measurement results accurate, the use of spectrum analyzers often involves the problem of setting the resolution bandwidth. To understand this problem, you need to know some basic principles of spectrum analyzers. Figure 1 is a basic principle block diagram of a spectrum analyzer. The intermediate frequency in the figure (the input signal is generated by the sum frequency or difference frequency with the local oscillator signal) and the local oscillator are controlled by the ramp generator. Under the control of the ramp generator, the local oscillator frequency will change linearly from low to high. In this way, when displaying, the ramp voltage generated by the ramp generator is added to the X-axis of the display, and the detector output is connected to the Y-axis after passing through a low-pass filter. When the ramp generator scans the local oscillator frequency, the spectrum of the input signal will be automatically drawn on the display. The low-pass filter at the output of the detector is called a video filter, which is used to smooth the response during analysis scanning.

1. Resolution bandwidth

In spectrum analyzers, frequency resolution is a very important concept. It is determined by the bandwidth of the intermediate frequency filter, which determines the resolution bandwidth of the instrument. For example, the bandwidth of the filter is 100KHZ. Then the frequency of the spectrum line has an uncertainty of 100KHZ, that is, if two spectrum lines appear within the bandwidth frequency range of a filter, the instrument cannot detect these two spectrum lines, but only displays one spectrum line. At this time, the spectrum line level (power) reflected by the instrument is the superposition of the level power of these two spectrum lines. Therefore, measurement errors will occur. Therefore, for two closely related spectrum lines, their resolution depends on the bandwidth of the filter.

We take the measurement of carrier level as an example to compare the resolution bandwidth settings of the instrument. Figure 2 shows the spectrum curves of resolution bandwidths of 30KHZ, 300KHZ, and 3MHZ (from bottom to top) (the input is a single carrier signal). When setting the resolution bandwidth, we consider whether the instrument has enough bandwidth to fully respond to the input signal. The correct method is to widen the bandwidth of the filter. When the signal carrier amplitude is no longer increased on the screen, it means that the intermediate frequency filter has enough bandwidth to respond to the input signal. In the figure, we can see that when the resolution bandwidth changes from 300KHZ to 3MHZ, the displayed signal amplitude does not change, which means that the 300KHZ bandwidth is sufficient. In addition, when the resolution bandwidth is set between 300KHZ and 3MHZ, the signal amplitude does not change for a single carrier. However, when actually measuring the image carrier level of the CATV system, the resolution bandwidth cannot be set to 3MHZ. This is because in reality there are sound carriers of adjacent channels near the image carrier (1.5MHZ apart), and the 3MHZ bandwidth cannot filter out the energy of the adjacent sound carrier. In this way, the energy of the adjacent sound carrier will be added to the image carrier being measured, making the measured level higher than the actual value.

2. Video Filters

The filter after the detector in Figure 1 is called a detector filter or video filter. It is a low-pass filter that reduces noise variations in the detector output, reveals some signals that have been masked and are close to the background noise, and helps stabilize the measurement if the noise power is being measured.

There are often DC and AC components at the output of the detector. The DC component represents the energy within the intermediate frequency bandwidth, so the DC component can be extracted through the video filter to remove some AC components, which can provide a more stable noise-free output. Figure 3 is a signal diagram of the detector output under different video bandwidths. Figure 3a uses a broadband video filter, and Figure 3b uses a narrowband video filter. It can be seen from the figure that the noise fluctuation is large when the broadband filter is used, and the fluctuation is significantly reduced when the narrowband filter is used. The average noise value of the two is the same, that is, the filter will not reduce the average noise level, but can reduce the peak level of the noise. Therefore, it can expose low-level signals that cannot be seen with a wider video filter. However, in some cases, such as analyzing some special noise signals, we need a wider video filter bandwidth for observation and analysis, so we can set the bandwidth of the video filter according to different situations.

The relationship between the bandwidth of the video filter and the resolution bandwidth is: the noise before detection can be reduced by a narrower resolution bandwidth, thereby reducing the noise output level of the detector; the noise after detection is smoothed and reduced by a narrowband video filter to reduce noise fluctuations, but the average power level of the noise cannot be reduced.