this is very confusing ... it works. I did upgrade yesterday from gnuradio
3.7.3 to 3.7.5, but
surely enough both osmocom_fft -f 1575.420e6 -g 30 and a basic flowgraph from
gnuradio-companion
using a 2 MHz sampling rate, osmocom source and WxGUI FFT sink, seem to exhibit
stable CW measurements.
I did unplug the dongle and plug it back into the USB port to make sure the
settings had not been
altered by oscmocom_fft.
I had to change synthesizers because the Hittite one I was using is on an
experiment at this time,
so I fetched a R&S SMA100 which only goes down to -20 dBm and added 18 dB
attenuation. At such a
level and with 50 dB gain on the RF front end the signal is barely visible when
averaging
(alpha=0.133), but the PLL sure looks like it is locked and exhibiting the
expeted -80 ppm offset.
Sorry for the trouble, will make sure now the issue is indeed solved, and let
the measurement run until
monday morning.
Thanks, JM