> Thinking about it this way,
> I would have to agree with an integration range from 0 to f0.
> If we were only to cover half that range,
> we would neglect the portion that occurs at negative frequencies
> (and also folded to the range f0/2 - f0).
No, integration range has to be from 0 to PSSfund/2.
Since PSD of Spectre is 1-sided view spectrum
which is twice of 2-sideded view spectrum.
For example, integration range is from 0 to fmax
for normal noise analysis. This is because PSD is 1-sided view.
Here you don't have to integrate from -fmax to +fmax.
Cadence will revise TDnoise explanation in SpectreRF manual soon.
You will see the reason why both analysis range and
integration range are from 0 to PSSfund/2.
If you integrate from 0Hz to PSSfund,
then you will include the same noise many times,
so an integral range should be limited to 0Hz~PSSfund/2.
About noise folding, see below.
(3) Folding and Alias.
When inputing noise into a linear time periodic varying (LTPV) system,
because of frequency translation, the output noise at f has contribution
from input noise at f+n*PSSfund (-infinity<n<+infinity).
This is noise folding. It is physical.
When sampling the above output noise from LTPV at t0+n*T, and do
PSD for it, the PSD repeats every PSSfund.
This is aliasing. It is not physical.