Spectral Measurements

Case A: Bandwidth exceeds that of

available amplifiers
T
A
(f)
S(f)
channels
f
1
f
N
receivers and antennas
detector noise, detect directly or
split frequencies and then detect
before amplification or detection
f
1
f
N
f
1
f
N
f
Receivers-G1
1) Extreme bandwidth: use multiple
2) If signal large compared to
3) Use passive frequency splitters
Spectral Measurements
Case B: Bandwidth permits amplification
f
1
f
N
further frequency splitting
Case C: Bandwidth permits digital spectral analysis
(
f Resolution
- ~ ) f ( V
2
M
2
N
) f ( ) (
N N

(Permits ~100 more B per cm
2
silicon)
(Reference: Van Vleck and Middleton, Proc. IEEE, 54, (1966)
)
G2
1) Amplify before either detection or
1) If computer resources permit, compute
N 2B
: transform point N per multiplys N log N

2) Or 1-bit (or n-bit) samples) (N
spectra M average
Examples of Passive Multichannel Filters
IN
f
1
f
n
f
2
Z
o
2. Waveguides
f
1 resonant
cavities at f
f
virtual
short
/4
f
N
RCVR RCVR
f
1
f
3
f
2
filters
passive
channel-dropping filters
Z
o
at f
1
f
G3
1. Circuits
Examples of Passive Multichannel Filters
3. Prism
-
bound electron(s)
red
blue
f
o
f
prism
5. Cascaded Dichroics
plane wave
<f
1
<f
2
f
n
>f
1
>f
2
>f
n-1
(f
1
>f
2
>f
n
)
4. Diffraction grating
(f)
G4
Digital spectral analysis example: autocorrelation
0 0
f
()
(f)
[W Hz
-1
]
analog signals
Possible analog implementation:
1) max lag =
max
= NT
2) sample lag, T sec
3) finite integration time >>
max
) (

v

B
RF
0
f
f
RF

0
f
f
IF
delay line



LO
local oscillator
v(t)
) T (

v
) (

v
) NT (

v

NT =
max
BB
RF

G5
: on based is
T 2
Resolution of autocorrelation analysis
) ( W ) ( ) (

y v
=
1)
W( )
0


M
= NT
-
M
0
f
0
f
B

v
(f) W(f)
~1/2
M
Hz
Thus

v
(f)
) f ( W ) f ( ) f (

v v
M
=
7 7 7 7
W(f)
f
0
G6

m
<
Aliasing in autocorrelation spectrometers
) f ( I ) f (
) t ( i ) (
) f (

) (

v
v
v
v


=
=
7 7 7 7
2)
i(t)
T
0
t
I(f)
0
f(Hz)
-1/T
Aliasing is spectral overlap
-1/T
0
1/T 2/T
B
) f (

v

3) Finite averaging time adds noise to
) f (

, ) (

v v

G7
1/T 2/T
Autocorrelation of hard-clipped signals
A/D
( )
2

c
o
u
n
t
e
r
delay line

LO
x(t)
v
o
(t)
1
v(t)
) (

v

hard clipping
A/D 1
+1
-1
0
t
+1 if v(t) > 0
Receivers-I1
Analysis of 1-bit autocorrelation
( ) ( )
0 -
0
x t
2 1 2 2 1 1



<
+

[ ] = =
2 1 x
) (
( )
( )
2 1
x x
2 1
e
1 2
1
2
2
2
2
1


+














1 2 v
t t ), ( ) ( where =

( ) ( ) [ ] ( )



=
0
0
2 1 2 1
0
2 1 2 1 x
2 2
( )










= + =




0
0
0 0
2 1 2 1
1 2 1 4
I2
J GRVZM are x , x where
x 1
x 1
x sgn , x , x t x Let
x sgn x sgn E
1 2
x x 2
2 1
dx dx x sgn x sgn
2 1



2 1
x x
dx dx x , x p dx dx x , x p

2 : Note dx dx x , x p
Power spectrum for 1-bit signal
Change
variables
x
2
x
1

dr
rd
r
x
1
= r cos
x
2
= r sin
dx
1
dx
2
= rdr d
( )
( )

=
2
0 0
1
1
2 r
2
2
x
1 e
1 2
1
2
r
d d 4 ) (
2
2
( )
( )

=
2
0
2
1
1 2
1
d 4
I3



2 sin
2 1


2 1
2 sin
Power spectrum for 1-bit signal
( )
( )

=
2
0
2
1
1 2
1
d 4
( )
( )
1 sin
2 2
1
4 1 d
sin 1
1
4
1
4
1
0
x

+


=

=



( ) ( )

=
x v

2
sin

( ) ( ) ( )
T
x
)

Where =
0 b
0
p(b)
p(a)
b

(see Burns & Yao, Radio Sci., 4(5) p. 431 (1969))

2 Let
I4
a


2 1
2 sin
2 1




- v(t sgn v(t) sgn
exact not b if
bias has : Note
Spectral response & sensitivity: autocorrelation receiver
;
B
f
1
f
T
) f (
eff
rms

channel bandwidth
(S. Weinreb empirical result,
MIT EE PhD thesis, 1963)
Apodizing weighting functions:
f
1.099
0.87
0.69
N
s
N f
s
N
s
first
sidelobe
-7 dB
-16 dB
-29 dB
Note trade between spectral resolution, sidelobes in (f) and T
rms








=

= N ;
N
T
1
f
M
s



uniform
raised cosine
blackman
0
0
0
I5

6 . 1
f 60 . 0
f 13 . 1
taps #
Spectral response & sensitivity: autocorrelation receiver
If N delay-line taps, how many spectral samples N
s
?
1
W()

M
Say uniform weighting of ():
Then B = N
s
f = N
s
/2
M
) where spectral resolution f 1/2
m
for orthogonal channels from boxcar W()
( ) ( ) B 2 N
M s
= = = =
W(f) for adjacent channel
M

f
In practice: raised cosine widens f by 1/0.6 1.7, so N
s
N/1.7
W(f)
I6
(1
taps # N rate nyquist at 2B 1 T B NT 2
2 1
Types of power
Delivered
Available
Exchangeable
( ) { } { }
j t
e
v t R t sin t

= +
V
g
g
Z
L
Z
R
g g
+
-
+
-
V
{ }( )


=


g
Z P
P VI R
2
1
P
L
D available
D e delivered
Receivers-K1
Receivers Gain and Noise Figure
Ve Re V cos Im V
+ j X
Z if i.e., , P max
Delivered and Available Power
{ }( )


=


g
Z P
P VI R
2
1
P
L
D available
D e delivered
R
L

g
e
Z R -
g
e
Z R
P
D
0 Z f
g
e
<
( ) - finite P P
g L
Z Z
D


=
K2
0
g
=
0
g
e
Z R -
g
e
Z R
P
A
P
D
R
L
0 Z f
g
e
>
Z if i.e., , P max
R : I
option power
le exchangeab
Z Im
R : I
Definition of Gain
g
Z
L
Z
G
1 2
G
power
(= G
p
)

1 2
D D
P P
G
available
(= G
A
)

1 2
A A
P P
G
transducer
(G
T
)
power

1 2
A D
P P

G
exchangeable
(=G
E
)
1
2
E
E
P
P
G
insertion
(= G
I
)

1
2
D
D
P
P
with
amplifier
without
amplifier
Note: G
A
, G
E
dont depend on Z
L
do depend on Z
g
(via P
E2
)
K3
Definition: Signal-to-Noise Ratio (SNR)
First define:
N
1
=
N
2
=
S
1
=
S
2
=
1
z
WH

2 2 2 1 1 1
N S SNR ; N S

V
g
1 2
g
Z
L
Z
F
G
( )
g
E
=
K4
exchangeable noise power spectrum @ Port 1
same, at 2
exchangeable signal power spectrum @ Port 1
same, at 2
SNR Define
Z f G Recall
Definition: Noise Figure F
kT N
N S
N S
SNR
SNR
F
o o 1
2 2
1 1
2
1

[Ref. Proc. IRE, 57(7), p.52 (7/1957); Proc. IEEE, p.436 (3/1963)]
S
2
= G
E
S
1
(see definition of G
E
)
N
2
= G
E
N
1
+ N
2T
transducer noise
( )
N
1
N GN GS
N S
F
1 1 1
1 1
+ =
+
=
( )
E
G

o
R
o
R
1
T
T N
1 F =

=
K5
K 290 T , where ,
G N
T 2
T 2
G let
T 2
G kT
G kT
G N
excess noise figure
receiver noise temperature
Receiver Noise Example
( 6 F T
F ( 2 F T
) F ( 1
T
T
1 F T
R
R
o
R
R
= =
= = =
= = + = =
Examples:
Excess noise corresponds to
receiver noise temperature T
R

+
1

T
R
= (F 1) T
o
290K
noiseless
G, F
T
A
+
1

T
A
N
2T
T
A
T
R
K6
dB) 5 . 7 ~ F K 1500
dB) 3 K 290
dB 0 K 0
F , G
F , G