Re: Airspy Mini Direwolf APRS IGate RPi


David Ranch
 


Hello Nick

I did some additional testing and it would seem that it is something within the rx_fm and perhaps how it handles certain interfacing aspects.  I ran essentially the same test using the RTLSDR driver and piped the output into sox and then compared(by ear) the same test while using the rtl_fm and even when attempting to adjust the gain the audio just seems way too low whether it’s the rtlsdr dongle or the airspy mini.

Hmm.. I would think AF gain would be the simplest item to correct.  I would keep digging as I'm sure that will be fixable.  Btw.. are you using PulseAudio on your system?  If so, consider looking at your pavucontrol mixer settings as well.


I also had tried using airspy_rx but I had issues figuring out what options/arguments to use to process an narrow band fm signal that can then be used by direwolf.  If anyone knows how to do that and could provide some help with that I would appreciate it.

airspy_rx is a different animal as it's more of a real command-line SDR than a purpose built narrow-FM demodulator application.  There are multiple ways to use it such as this CSDR example:

    https://github.com/ha7ilm/csdr/issues/13


I was able to get that approach working with an AirSpy R2 but it took some adaptation as my system didn't support the specific destination audio codec.  Even while working, the audio wasn't smooth enough for suitable APRS support:

   $ airspy_rx -t 0 -r /dev/stdout -a 2500000 -f 144.390 -g 18 | csdr fir_decimate_cc 60 0.005 HAMMING | csdr bandpass_fir_fft_cc 0 0.1 0.05 | csdr realpart_cf | csdr agc_ff | csdr limit_ff | csdr fractional_decimator_ff 1.0416 | csdr convert_f_u8 | aplay -c 1 -f U8 -r 48000


It sounds like rx_fm might be a better solution for your targeted narrow-FM application as it's reported the overall CPU load might be lower with this approach.

--David
KI6ZHD

Join airspy@groups.io to automatically receive all group messages.