You may want to take a look at http://sourcelink.cadence com solution 1832229 (SLK) Why Is There A Positive dBc/Hz Phase Number Displayed When Plotting Pnoise?
ADE figures the "phase noise" by taking the amplitude of the oscillation from the noise free pss simulationand dividing it into the noise levels calculated from a pnoise simulation. The "output noise" plot is just the raw, unscaled, noise levels at the oscillator output, that is the raw results of the pnoise simulation.If you plot both on a log plot, then they should differ by a constant, and that constant should be the level of the oscillator output.ADE produces the "phase noise" output, which is normalized by the carrier amplitude. However, that normalization does not generally give 0dBc at 1Hz. For that to happen the noise level would have tobe exactly the same as the oscillation level at 1Hz, which need not be the case.
Phase noise levels greater than 0dBc do not make sense, since the noise has more power than the actual oscillation, so they decided to disregarddata for values less than the cross-over frequencies.The cross over frequency is simply the frequency where the ADE output is 0dBc. In other words, the frequency where the power in the carrier is the same as the calculated noise power.
The bottomline: disregard/discard phase noise values greater than 0dBc, as they are non-physical.