Re: [Discuss-gnuradio] Re: usrp wfm problem

[Matt Ettus]

Quoting Marcus Leech <address@hidden>:

> It would be useful to have a function that returned the as-aliased
> frequency for
>   any given input frequency (up to the maximum input frequency of the A/D).
>   Trivial, I suppose, but less opportunity for brain farts that way.


A good idea.

Anyway, to elaborate on what I said before, it is helpful to think in
"normalized frequency" terms.

The sampling rate is normalized to 1.  The nyquist freq is 0.5.  Thus, for real
samples, the signal is from frequencies of 0 to 0.5.  For complex samples, it
is from -0.5 to +0.5.

In the case we talked about before, 64 MHz => 1.  433 MHz = 6.765625.  Drop the
whole number part and you are left with 0.765625.  Since that is over .5, you
reflect it in 0.5.

1 - 0.765625 = .234375

.234375 * 64 MHz = 15 MHz

Since we had to do that "reflection", the spectrum is reversed.  This isn't a
problem for FM, but needs to be taken into account for most modulations.

Matt