Validity of cross-spectrum PN measurement

Abstract

This paper reports the anomaly of instruments that uses cross-spectrum method for the measurement of oscillator phase noise. Although, cross-spectrum techniques facilitate high sensitivity by correlation and averaging, it can give disingenuous measurement data depending upon many unknowns (phase-inversion, multi-mode spectrum associated with negative index dynamics, odd order harmonics, impedance/phase mismatch, and thermal energy) that are not taken into account. Most of the time errors in PN (phase noise) measurement is within the limit but this is not true for ultra low phase noise crystal oscillator and oscillator that uses negative index (ε <; 0, μ <; 0) resonator and filter element. We report a new caution of using metamaterial resonator as a filter between DUT and instrument as an attempt to reduce the noise floor. Exciting design choices with ε <; 0 or μ <; 0, however, impedance is not what we are used to. This discussion is imperative for emerging negative index resonator oscillator circuit at microwave and millimeter wave frequency.