Lets say you have a signal at a known frequency with a known amount of phase noise, and you pass the signal through a fixed attenuator, is there a way to calculate/estimate what impact the attenuation has on the phase noise of the output signal?

Greetings crg:

My guess is that the effect of the attenuator on phase noise would be similar to the intended signal and noise. You would need to perform a cascade analysis for the path where the phase noise is first injected and treat that component of the total signal as a separate signal that is summed with the intended signal.

The discrete phase noise component would be attenuated by an amount equal to the total Gain in the path from the phase noise injection point and the continuous phase noise component would be subject to the total Gain and Noise Figure.

If the phase noise is present on the signal at the system input, then it is treated to the same effects as the rest of the intended signal and noise.

If it is injected in an LO at a mixer, then the phase noise component is determined by a separate cascade comprised by the path the LO signal takes.

In the last case, for a real system the summation of the two components [(intended signal + noise] + [LO + phase noise)] would be a point-by-point vector addition, so a simple scalar summing could produce erroneous results.

Any comments by others?

_________________
- Kirt Blattenberger
RF Cafe Progenitor & Webmaster