Spectrum Analyzer [002]

I've had current (red line), average (green fill) and max (blue fill) correctly drawing for a while, but it's taken some time to get the scale (in dBm) right. And I'm still not accounting for gain associated with the window function and the fft. So, ignoring that for the moment, the base line in the screenshot represents -100 dBm, and the top, 0 dBm.

The screenshot shows a 190 KHz span centered at 1040 KHz, so those nine fairly evenly spaced larger spikes are all broadcast radio stations spaced about 20 KHz apart (10 KHz is the standard), and then maybe some more distant or lower powered stations in between.

The number in the top-right corner indicates the number of "blocks" of samples received from the radio. Each block contains 2,048 measurements (ie, the instantaneous amplitude) of the span. Those measurements are received at a rate of ~190,000/second. There's a ton of noise though, so after a window function cleans it up, it is run through the fft, converted to dBm, averaged (over the period of 100 samples) and smoothed (for drawing). And then presto! A pretty graph.

Can't wait to actually get to demodulate something. But first I need to get the waterfall and density graphs drawing. And maybe actually draw an indication of the actual values (ie, dBm and frequency).