Biomedical Engineering Reference
In-Depth Information
with,
g
m
ω
0
C
t
R
s
+
R
g
R
s
+
R
g
+
R
s
+
2
|
G
m
(
jω
0
)
|
=
C
t
2
,
D
(
jω
0
)
=
,
(6a)
g
m
L
s
C
t
g
m
L
s
R
g
+
R
s
+
R
g
+
jω
0
(
L
g
+
L
s
)
g
m
jω
0
C
t
.
T
(
jω
0
)
=
×
(6b)
g
m
L
s
C
t
R
s
+
R
g
+
Replacing these terms in (5) and with some simplifications,
F
will be
1
x
,
R
γω
0
C
t
α
2
g
d0
R
R
s
F
=
+
(7)
where
α
2
δ
5
γ
Q
s
)
C
gs
α
2
δ
5
γ
C
gs
C
t
g
m
g
d0
,
x
=
(1
+
C
t
+
1
−
2
|
c
|
,
α
=
1
ω
0
C
gs
R
s
,
ω
τ
0
C
gs
g
m
C
t
,
Q
s0
=
ω
τ
=
C
t
=
(8)
Q
s0
C
gs
R
s
1
ω
0
C
t
R
,
Q
s
=
C
t
R
=
R
=
R
s
+
R
g
,
g
m
C
gs
.
C
t
=
C
gs
+
C
aux
,
R
=
R
s
+
R
g
,
and
ω
τ
0
=
Based on above equations, the noise figure is optimized with gain and power
considerations.
The testchip is designed, fabricated, and tested in a 0.13
m complementary
metal-oxide semiconductor (CMOS) process. Figure
17
shows the measured S-
parameters of this circuit. As shown, excellent input and output matching as well as
output-to-input isolation are achieved. Peak value of 18.1 dB is measured for
S
21
with
excellent gain flatness. Figure
18
a shows the measured group delay of the designed
LNA. Figure
18
b shows the measured IIP3 of this circuit. The intercept point in
μ
1
dBm shows very good linearity. Also, Fig.
18
c shows the noise figure of this circuit,
with a minimum value of 3.12 dB and an average of 3.81 dB.
−
Ultra-Wideband Mixer
Despite all the advantages of UWB systems, their design is typically very challeng-
ing due to the fact that they should demonstrate good performance for the entire
bandwidth. Power consumption, conversion gain, linearity, silicon area, and noise
figure are major concerns in the design of UWB mixers.
For Gilbert cells, which are the most common topology of mixers, an increase in
the bias current of RF stage will typically result in a better noise performance as well
Search WWH ::
Custom Search