I would like to determine the NF of an amplifier when in compression. I have an input signal of 11.87dBm, with a phase noise floor of -163dBc/Hz. The output of the amplifier is 17.9dBm with a phase noise floor of -162.5dBc/Hz.
Using the equation NF=10Log(SNRin/SNRout) (same as Phase noise IN [dBc] - Phase noise OUT [dBc]) I get NF=0.46.
This cannot be correct though as the specified NF of the amplifier is 4 when operating linearly.
I would like to determine the NF of an amplifier when in compression. I have an input signal of 11.87dBm, with a phase noise floor of -163dBc/Hz. The output of the amplifier is 17.9dBm with a phase noise floor of -162.5dBc/Hz.
Using the equation NF=10Log(SNRin/SNRout) (same as Phase noise IN [dBc] - Phase noise OUT [dBc]) I get NF=0.46.
This cannot be correct though as the specified NF of the amplifier is 4 when operating linearly.
BY definition, noise figure of an amplifier is measured by introducing two different input noise levels, and measuring the output noise power. If the tested amplifier is linear, then one can calculate its NF by Y-factor method.
Once the tested device becomes non-linear, the above is not valid. Calculating makes not much sense as you can see.
If your amplifier is multi-stage, then you can design the stages so that the saturation occurs in the latter stage(s) and the front stage, determining the NF, is not affected. Usually a good amplifier gets saturated in such way. Under saturation, the output power meter reads a small or no change as a response to input noise variation, and the NF of such amplifier makes no sense.
Noise Floor is a different thing than Noise in the signal.You confuse Noise in the signal with Noise Floor which is determined by measurement equipments/set-up/system.