文档介绍：AppendixA
Fundamentals of Two-Port Noise Theory
Any noisy two-port can be replaced with a noiseless two-port with two input noise sources de-
pending only on the noisy two-port itself. The equivalence of both is valid for any source im-
pedance. This equivalent two-port is shown in Fig. . The two noise sources are the noise
2 2
current source, in between the positive and negative input and the noise voltage source, en in
series with either input. The source impedance is represented by Gs and jBs which are the re-
sistive and ponent of the source admittance. The noise of the source impedance is
2
represented by is.
Since the input noise current in may be partly correlated with the input noise voltage, en,the
input noise current is split up according to
in = ic + iu, ()
where ic and iu represent the correlated and uncorrelated part of the input noise current respec-
tively. The correlation allows to rewrite ic as
ic = Ycen, ()
where
Yc = Gc + jBc ()
is plex correlation admittance. Reconfiguring the different noise sources into their Norton
equivalent allows to add them together through superposition:
in,tot = is + ic + iu + Ysen = is + iu + (Yc + Ys)en. ()
2
en
noiseless
G jB i 2 i 2
S S s n two−port
Figure : Equivalent noise model of a two-port and its signal source.
174 Fundamentals of Two-Port Noise Theory
This yields three noise current sources which are mutually uncorrelated. The total average
2
squared noise current, in,tot is foundas
2 2 2 | | 2
in,tot = is + iu + Yc + Ys en. ()
where the cross averages are zero. The noise factor can be calculated by dividing this total
squared noise current by the squared noise current of the source:
i2 + |Y + Y |2e2
F =1+ u c s n . ()
2
is
This can be rewritten as
G + |Y + Y |2R
F =1+ u c s n , ()
Gs
where
e2 i2 i2
R n G u G s . ()
n 4kT ∆f u 4kT ∆f s 4kT ∆f
Gc, Bc, Gu and Rn are four noise pletely describing the noise