Technical data

130 Chapter 8
Concepts
Resolving Closely Spaced Signals
Concepts
Resolving Closely Spaced Signals
Resolving Signals of Equal Amplitude
Two equal-amplitude input signals that are close in frequency can
appear as a single signal trace on the analyzer display. Responding to a
single-frequency signal, a swept-tuned analyzer traces out the shape of
the selected internal IF (intermediate frequency) filter (typically
referred to as the resolution bandwidth or RBW filter). As you change
the filter bandwidth, you change the width of the displayed response. If
a wide filter is used and two equal-amplitude input signals are close
enough in frequency, then the two signals will appear as one signal. If a
narrow enough filter is used, the two input signals can be discriminated
and appear as separate peaks. Thus, signal resolution is determined by
the IF filters inside the analyzer.
The bandwidth of the IF filter tells us how close together equal
amplitude signals can be and still be distinguished from each other. The
resolution bandwidth function selects an IF filter setting for a
measurement. Typically, resolution bandwidth is defined as the 3 dB
bandwidth of the filter. However, resolution bandwidth may also be
defined as the 6 dB or impulse bandwidth of the filter.
Generally, to resolve two signals of equal amplitude, the resolution
bandwidth must be less than or equal to the frequency separation of the
two signals. If the bandwidth is equal to the separation and the video
bandwidth is less than the resolution bandwidth, a dip of
approximately 3 dB is seen between the peaks of the two equal signals,
and it is clear that more than one signal is present.
When the Agilent CSA spectrum analyzer span is > 0 Hz, the sweep
time is set automatically to keep the analyzer measurement calibrated.
When the resolution bandwidth is < 1 kHz, there will be large increases
in the sweep time as you decrease the RBW in a 1, 3, 10 sequence.
Fortunately, the Agilent CSA allows you to also set the RBW in 10%
increments, thereby allowing you greater flexibility in trading off sweep
time and resolution.
For the shortest measurement times, use the widest resolution
bandwidth that still permits discrimination of all desired signals.
For example, in a 10 MHz span, the sweep time with a 300 Hz RBW is
1.02 s, and the sweep time with a 100 Hz RBW is 8.01 s. If the 300 Hz
RBW does not provide sufficient resolution, and the sweep time with a
100 Hz RBW is too long, you could try the 200 Hz RBW. The sweep time
with a 200 Hz RBW is 1.4 s, over 5 times faster than the sweep time
with a 100 Hz RBW.