Amplifier effect on phase noise

I have designed a PLL synthesizer. Now, I have to increase the power of this signal from synthesizer output. I will be using a saturated buffer amplifier. Can someone tell me how does this affect on the phase noise or does it? Any references?

Saturated is good. Linear region region is best. Operation in a partially saturated mode (like around the 1 dB compression point) is the worst for phase noise, because of the am to pm conversion.

Fully saturated, there could be more 1/f noise, so depending on your close to the carrier phase noise spec, you could have some trouble, but probably not.

Amplifiers do contribute to overall phase noise but this contribution is usually considerably lower than the phase noise of the phase-locked oscillator.
An oscillator generates noise in addition to the nominal carrier. Noise also in principle is the starter to oscillation. In a phase-locked system, the control-loop feedback response determines the PN vers.distance from carrier.
If you are interested, refer to "www.picosecond.com" where amplifier PN is specified in special wideband amplifiers used in optical communication.

No one know how an amplifier at compression or saturation affect the phase noise. In many cases they increase overall phase noise at the output. But in some case they don't affect it. For example, using HEMT amplifiers in saturation it is possible to multiply frequency of LN OCXO without phase noise excess above 20lgN. In PLL synthesizer an saturated amplifier will affect on VCO PN floor which isn't interesting in many cases. Close-to-carrier phase noise is too big in comparison with additive amplifier PN.