On Mon, Jan 27, 2020 at 06:25:48PM +0100, Daniel Estévez wrote: > El 26/1/20 a las 21:38, Marcus Müller escribió: > > Seconding what Brian says: > > Math says *any* signal up to a bandwidth of 6 MHz can be represented by > > 6 MS/s. So, either, your signal isn't that bandlimited, or you forgot > > to tell us an important requirement (or you might have your math > > wrong).
You'd need 12M real samples/second or 6M complex ones. An Airspy R2 will provide 20M real samples/second (12 bits), frequency range is 24 MHz to 1.7 GHz. > Math also says that the time of arrival of a 6 MHz waveform can be > determined with an accuracy of 2ns, if the signal SNR is large enough. Or the measurement time is large enough. Some 25 years or so ago I implemented ESA tone ranging as an SDR application. It's used to find the distance from a ground station to a satellite by measuring the round-trip time. This in turn is done by phase measurements of signals which are phase modulated on the carrier. The highest frequency modulation was 100 kHz, and this provided a resolution better than a nanosecond. No need to upsample at all - upsampling doesn't give you any new information. Ciao, -- FA
