Measuring very close and relatively weak signals with a spectrum analyzer
When measuring signals with a spectrum analyzer, setting the resolution bandwidth (RES BW) correctly is the key to reducing measurement errors and capturing two very close and relatively small signals. We know that the resolution is related to the bandwidth of the IF filter (IF filter) of the spectrum analyzer. The narrower the bandwidth, the higher the resolution. In addition to the bandwidth of the filter, shape factor, filter type, residual FM, and noise sideband are also factors that should be considered when setting the resolution bandwidth. Below, we will analyze each item in turn.
First of all, we noticed that the test signal cannot be displayed as an infinitely thin line on the spectrum analyzer display, and it has a certain width itself. When tuned through the signal, its shape is the display of the spectrum analyzer's own resolution bandwidth (IF filter). Thus, if the filter bandwidth is changed, the width of the display response is changed. The resolution bandwidth is specified as the 3dB bandwidth of the filter. The 3dB bandwidth tells us how to separate two equal-amplitude signals that are very close to each other. Generally speaking, if the interval between the two signals is greater than or equal to the 3db bandwidth of the selected resolution bandwidth filter, the two equal-amplitude signals can be distinguished. .
Example 1: Two equal-amplitude signals are separated by 10kHz. It is easy to separate them with a resolution bandwidth of 10kHz, because there will be a sag of about 3db between the two signals. It is clear that this indicates that there is more than one signal.
Form factor (also known as bandwidth selectivity) is a measure for the spectrum analyzer to resolve adjacent unequal amplitude signals. For a given resolution filter, the shape factor is the ratio of 60dB bandwidth to 3dB bandwidth. Usually the spectrum analyzer IF The filter has a form factor of ≤15: 1. When we measure two adjacent signals of unequal amplitude, if the small signal is very close to the large signal, the small signal will be masked by the large signal. In other words, the small signal may be buried in the skirt of the large signal. For two signals that differ in amplitude by 60db, the separation should be at least half or less than half of the 60dB bandwidth (approximately 3dB drop). The meaning of the 15: 1 form factor is that the 60db bandwidth corresponding to the 1kHz resolution bandwidth is 15kHz, and half of it is 7.5kHz. If the interval between the large and small signals is 10kHz, the resolution bandwidth of 1kHz can be used to distinguish them.
Example 2: In the above example, two equal-amplitude signals are separated by 10 kHz. Obviously, it is problematic to distinguish signals with a resolution of 10 kHz. However, it is possible for distortion products (small signals) that are separated by 10 kHz and the amplitude drops by 50 dB. It is masked. If the shape factor of the 3kHz filter is 15: 1 and the bandwidth of the filter dropping by 60db is 45kHz, the distortion product (small signal) will be hidden under the skirt of the test signal (large signal). Resetting a narrow-band filter with a resolution bandwidth of 1 kHz, the bandwidth corresponding to 60 dB is 15 kHz, and distortion products (small signals) can be easily observed because half of the 60 db bandwidth is 7.5 kHz, which is less than the interval of 10 kHz between the frequency bands. Therefore, In this example, the resolution bandwidth required for measurement should be no greater than 1 kHz.
Another factor that affects the resolution is the remaining frequency modulation of the spectrum analyzer. The remaining frequency modulation is related to the frequency stability of the spectrum analyzer's local oscillator. The resolution bandwidth of a spectrum analyzer cannot be made so narrow that it can be observed. Its own instability, if it can do so, then we will not be able to distinguish between the remaining FM of the spectrum analyzer and the added signal, and the remaining FM makes the displayed signal ambiguous, so that within the specified remaining FM The two signals cannot be distinguished. This means that the remaining frequency modulation of the spectrum analyzer determines the minimum permissible resolution bandwidth, and likewise, it determines the minimum interval of equal amplitude signals. Using a local oscillator constructed with phase lock as a reference source can reduce the residual frequency modulation and also reduce the minimum allowable resolution bandwidth. The reason why the high-performance spectrum analyzer is expensive is because it has a better phase-locking system, with lower residual frequency modulation and a smaller minimum resolution bandwidth.
Noise sideband (also known as phase noise) is a modulation sideband. As a noise sideband on the base of the signal response, it shows instability. This noise may obscure the near-end (near carrier) low-level signal. In other words, considering only the bandwidth and form factor, we may see it, but these noise sidebands affect the resolution of high-level near-end low-level signals. Sideband noise generally only occurs in the narrowest resolution bandwidth, because the response time required by the narrower filter is longer. When the scan time is too fast, the resolution bandwidth filter of the spectrum analyzer cannot fully respond, that is, the amplitude decreases 3. The frequency moves up. In order to maintain the correct reading status, we should follow the scan time settings below.
Scanning time ≥ k × scanning interval / (resolution bandwidth) 2
From the above formula, we can see that the same scan interval (span), narrow resolution bandwidth (RES BW) means a long scan time (SWP). Here, for flat-top filters, 10 ≤ k ≤ 20, and for synchronously tuned analog filters, 2 to 5 ≤ k. For spectrum analyzers using digital signal processing, k≤1, which indicates. The digital resolution bandwidth has an excellent form factor and measurement speed. Generally, the spectrum analyzer can automatically interlock the scanning time, and automatically select the fastest allowable scanning time according to the selected scanning interval and resolution bandwidth. If the manually selected scanning time is too fast, a message will appear on the display.
Finally, it should be pointed out that there will always be a certain amount of clutter in the measurement. Usually, the clutter is very small and can be ignored. However, when measuring low-level signals, the clutter problem is serious. Usually, the clutter occupies a wide frequency band. If the measured value contains the clutter level, then the measurement error will depend on the clutter level. The resolution bandwidth is narrow, allowing less clutter to pass, and the clutter layer on the display drops. A 2x change in resolution bandwidth will cause a 3dB change in clutter level. The reduction of the clutter level enables the measurement of smaller signals that were originally masked by the clutter. If after adjusting the resolution bandwidth, the signal is still close to noise (clutter), then the attenuation and video bandwidth (VBW) can be adjusted to improve the visibility of the signal. For attenuation and video bandwidth, this article does not make a specific analysis.
In summary, the appropriate narrow resolution bandwidth, the appropriate small input attenuation, and the appropriate video bandwidth (video bandwidth ≤ 0.1-0.01 resolution bandwidth) are the two very close or relatively weak signals that we use a spectrum analyzer to reduce the measurement. The basic method of error.
Stainless Steel Bakery Trolley
Stainless Steel Bakery Trolley,Bakery Rack Trolley,Bakers Rack Trolley,Bread Rack Trolley
Jiangmen Junerte Stainless Steel Kitchenware Co.,Ltd , https://www.junertejm.com