Re: [Openbts-discuss] Jumps in the RX packet timestamps ? [solved + working with single RFX900 !]

[Sylvain M. <24...@gm...>]

Ok, apparently it was due to gnuradio-3.1

I changed the few things for OpenBTS to compile with the latest
gnuradio and now it works fine and reliably.

The conclusion is also that for pure desktop testing 1 RFX board can
work. In my setup :
 - I used a proper 50 ohm RF termination for TX and not just leave it
unconnected, to avoid the energy being reflected back into the RFX.
 - I put the TX power to the minimum (but even then it still largly
covers the entire room so for testing no worries).
 - I avoided some frequencies. Don't know why but every 4MHz I have a
parasite spike and if you're right on it that's not good ... check
with usrp_fft that both upstream and downstream ranges are clear.
 - I used the nokia 3310 in "BTS Test" mode to force it to scan only 1
ARFCN (and so it doesn't 'lock' to another bts clock)


Sylvain