Hi,

I have a basic doubt regarding mixer ( Gilbert cell ) noise figure.
People say,
Typically NF ( SSB ) = 2*NF ( DSB )

I want to know if the above holds only for thermal noise or for flicker noise also.
Since the flicker noise at the mixer output is due to mechanisms entirely different from the thermal noise folding back from the signal band and the image band, I would think that the flicker noise contribution should be same in both NF ( SSB ) and NF ( DSB )  computation.

But when I plot the spot NF ( DSB ) and NF ( SSB ) of a  mixer in cadence , I find a 3 dB difference in the flicker noise region ( low freq. ) also.


Another question :

In considering the noise contribution from mixer transconductor and load in computing   NF ( DSB ) and NF ( SSB ),
I would think that the load should contribute equally in both the cases and
the transconductor should contribute twice in the SSB case.

Am I right ?


Prabhu