Again a simulation is required using ADS.A resistor of 100Mohm is connected toa VCO with gain set to 20MHz/V, and the spectrum is shown below.
I have filtered the noise from the resistor using a three pole C-L-C passive filter.
The simulated rms noise voltage at 9.85mV, agrees reasonably closely with the calculated value of 9.101mV, and is expected to be higher as the noise BW of a filter is usually higher than the 3dB BW.
The noise power spectrum is calculated using same FM equation as in the first section and assumes the modulation is at a low level.
The simulated noise sidebands at -65.9dBc agree reasonably well with the calculated value of -67.8dBc.