R3267 Series Spectrum Analyzer Operation Manual (Vol.1)
A.3 Glossary
A-9
A.3
Glossary
Average Noise Level
This sensitivity represents spectrum analyzer’s capability of detecting the smallest signal and is directly re-
lated with noises generated from a spectrum analyzer itself. The sensitivity, however, varies depends on the
resolution bandwidth used. In general, the maximum input sensitivity of a spectrum analyzer is expressed as
average noise level when the instrument is used with its minimum resolution bandwidth.
Bandwidth Accuracy
The bandwidth accuracy of the resolution bandwidth filter is expressed by the deviation from the nominal
value of the 3 dB lowered point. This deficiency has almost no effect when measuring normal signals at a
continuous level, but it should be taken into consideration when measuring the level of a noise signal.
Bandwidth Selectivity
The bandpass filter normally attenuates using a Gaussian distribution instead of the so-called rectangular
characteristic. Consequently, if two adjacent signals of different sizes are mixed, the smaller signal hides at
the tail of the larger signal (Figure A-6).
Therefore, the bandwidth at a certain attenuation range (60 dB) should also be defined. The ratio between the
3 dB width and 60 dB width is expressed as the bandwidth selectivity (BW60 dB/BW3 dB).
Figure A-1 Bandwidth Selectivity
Bandwidth Switching Uncertainty
Several resolution bandwidth filters are used to obtain an optimal resolution in signal spectrum analysis ac-
cording to the scan width. When switching from one resolution bandwidth filter to another while measuring
one signal, an error is generated for the differences in loss. This error is defined as the bandwidth switching
uncertainty.
(a) Separation of signal at different levels
(b) IF filter selectivity
