You're on the wrong path, guys. I'd suggest paying more attention to the subject instead of doing wild experiments and inform yourselves using available and proven information. Paul's frequencies are waaay off. Even if corrected with the error value, they have nothing to do with the original Davis frequencies. The frequency sweep WILL find bogus frequencies, because the SDR stick is not a precise narrowband receiver. The chips are originally made for digital terrestrial and analog radio reception for several hundred kHz to low MHz bandwidths.
Consider this: the Davis uses a bandwidth of 25 kHz and an FSK deviation of 19.8 kHz (double-sided). Hopping slots are allocated 502 kHz apart. If you skip when sweeping by 40 kHz, no way in life will you find the proper frequencies, ever. I guess the SDR FSK decoder code is not very well tuned to the task at hand, and while I admit I haven't programmed any SDR task, even though I know the theory behind it. To re-iterate: there is no need to "find" the proper frequencies, they are available in the RFM69 Arduino code. You do need to find and fix your offset on your sticks before starting rtlsdr, though. Or use precise ~1ppm devices. Your scanned frequencies will be local to your setup, they most probably will not work reliably for anyone else! On Saturday, March 16, 2019 at 7:18:26 PM UTC+1, [email protected] wrote: > > Hi Paul, > > The idea is to first get a raw list of US hop frequencies. > The second step is to correct the found frequency with the reported > freqError. > Then we look at the sorted list of ‘corrected’ frequencies and calculate > the differences between each two adjacent frequencies, so we will detect > any likely missed frequencies. > At last interpolate values for the ‘missed’ frequencies. > When all frequencies are more or less right, we can add them to the > program and start the normal hopping proces. > At last we will use the average freqError per channel to fine-tune the > frequency list. > > Luc
