[time-nuts] USRP B210 phase-meter

[Anders Wallin]

a short writeup of our phase-meter/digital-DMTD tests with an Ettus B210 is
now on researchgate:
https://www.researchgate.net/publication/336825605_Software_Defined_Radio_Based_Phase_Meter_for_Frequency_Metrology

an ADEV of 2e-14 at 1s in 0.5 Hz bandwidth is fairly straightforward to
achieve, for a 100MHz signal split into ch1 and ch2.
There are issues with cross-talk between the channels - looking to solve
that later.

The raw data comes in at 1MS/s and everything works as expected when
downsampling by three decades to about 1kS/s. Then ADC noise is something
else (1/f ?) than white at fourier frequencies below about 1 kHz - so there
isn't much 'process gain' achieved when downsampling further to 100S/s or
10 or 1 S/s.

Anders