Hi All, Some of you may recall that I am trying to sample at 2.4M and count beeps between samples (usually 80 Beeps per minute, sometimes about 47 beeps per minute).
80 beeps per second equates to a beep every 1.3sec. Anyway I have saved a 10 sec numpy array to disk, then processed it in small chunks, lets say about 0.2 seconds worth of chunks. This works perfectly. The issue is when I try to do this in realtime from the RTLSDR I am getting more like 120 beeps per minute which is wrong. I suspected the sampling rate at 2.4M might be to blame so I went as low as possible to 250k and the beeps per minute got to 86 (closer). I still need to be certain if I am dropping samples, and the code that acquires data is threaded and buffered, so quite complicated. I assume I can add some time stampss to check for missing samples against expected samples? I change the sample rate variable to 250k, but I dont think I needed to change any other calculations, since freq offset stays the same? Code is here : https://github.com/bigalnz/test_fft/blob/read_from_sdr/test_fft.py Thanks Al -- "Beat it punk!" - Clint Eastwood
