Phase noise evaluation digital algorithm for precision oscillations

The accuracy increase for digital algorithms of noise phase modulation calculations is urgent for some measurements and especially for processing of signals which have been obtained by means of computer aided design software. Calculation accuracy limits an application of digital algorithms for precision oscillation research. This paper considers accuracy of phase noise calculation common methods when precision oscillation is described by a single sequence (without quadrature components) of digital samples in time domain and presents the author's algorithm for this purpose. The algorithm is based on signal points tracking by means of the signal interpolation in the vicinity of these points to determine the time moment when oscillation crosses an ordinates' axis. The calculation accuracy can be previously set and can provide an error less then 0.01 rd without large computation expenses. The algorithm output sequence has a minimum dimension for researched phase fluctuation description (two samples within oscillation period). Therefore it allows one to obtain phase noise spectral density floor equal to a minus 160 dBc/Hz. The calculated sequence of output phase noise samples allows one to monitor both short-term phase fluctuation and slow drift of a phase.