Table of Contents
Advertisement
extracting the peak value of each frequency point frame by frame, or
averaging the results of all the frames. In this manner, the result shown on the
display no longer represents the instantaneous situation of the signal, thus it's
no longer the "truth". As a matter of fact, the "truth" of a varying signal can
only be described with a time domain waveform or a mathematical formula.
Why? Because a decent frequency resolution requires a large amount of FFT
sampling points (like, 2048) to achieve, and during the sampling and
processing time, the original signal has probably changed. Yet, compared to
the sweeping spectrometer, normally the real-time spectrometer is much
closer to the "truth".
What's more, there is a limit for frequency resolution of displaying, as well.
Each frame of FFT produces more than a thousand data points, and with
splicing, these data points can even reach a magnitude of hundreds of
thousands. Due to the fact that the maximum displaying capability of KC908's
horizontal axis is 800 pixels, each pixel has to represent multiple frequencies.
How to define "represent" is another major concern of detection. In KC908,
the solution is displaying the peak value of all those frequencies a single pixel
represents, to ensure no signal is missed, which is the basic principle a
reconnaissance equipment should live up to. However, this method of
presentation may make certain frequency/phase modulated signals look like
stable signals. Thus, while analysing a single signal, it's advised to reduce
Advertisement
Table of Contents
