As mentioned earlier, the frequencies were sniffed in original VP2 consoles (you can find the posts about that in wxforum threads). They're the GFSK center frequencies, no IF. The "RFM69 frequencies" just mean the ones programmed into the firmware of small Arduino clone based receivers using an RFM69 transceiver module. Also noted that with a simple SDR receiver the packets are appearing at the expected frequency. It's possibly a simple overlook with handling frequencies or frequency correction in the SDR demodulator code. I think any help or ideas come handy.
On Mon, Mar 18, 2019 at 12:56 AM Greg Troxel <[email protected]> wrote: > [email protected] writes: > > > I found a correlation formula between the RFM69 frequencies and my EU > > trial-and-error frequency set. > > > > rtldavis frequency = (RFM69 frequency -40000000) * 1,048319523 > > > > These calculated rtldavis frequencies DO work! > > > > Excel gave the following equation: y = 1.0439x - 4E+07, but the factor > was > > not precise enough. > > > > The results: > > > > RFM69 rtldavis > > 868066711 868078500 > > 868181885 868199238 > > 868297119 868320041 > > 868412292 868440779 > > 868527466 868561518 > > It would be interesting to really understand what's going on. It seems > like the "RFM69 frequencies" are what is programmed into the > transmitter, and I wonder if that is the intended transmit frequency or > if there is some IF involved. It would be interesting to also sniff the > RFM69 receive chip frequencies at the console. > > And, to measure the actual frequencies with test equipment, or something > that can be relied upon to be reaosnably accurate because of > calibration. > > When I saw the formula above, I suspected a 40 MHz IF and was suspicious > of the 1.048ish factor. But I wonder what that would look like if it > were fit with different assumptions about the base formula. > > Overall, something seems wrong in code somewhere, as a change of 12 kHz > offset to a 35 kHz offset over 500 kHz does not seem plausible. > > I'll try to run this and see if I can spot anything. > >
