Tracked down a weird quirk in the rtl-sdr library this evening.
I was having problems with lower frequencies working, but only at
sample-rates >300e3. There's a piece of code that tries to optimize
selection of IF frequency and filters based on sample-rate. For
some sample-rates, a lower IF is used, presumably to give "headroom"
at the filter edges.
The *problem* with this heuristic is that at lower input frequencies,
the sum of IF + input-frequency is too low for the PLL to lock--seems like
it runs into trouble below 27.675MHz, which means that you need to
pick an IF that's high enough to guarantee that the PLL will lock.
In my tests, I modified the code to detect this case and arrange for a
higher IF frequency. This seems to work OK, but I wonder if
the maintainers of the rtl-sdr library (are there any anymore?) have
considered this issue...
Now, this isn't strictly gnuradio related, but the likely people
probably hang out here from time to time...
_______________________________________________
Discuss-gnuradio mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
