Photon noise
black (no reflection)
p (watts ) = nsn2 hf
TK
Area A
Ster.
number of states
in cavity in B
Calculation of the number of electromagnetic modes = m:
2kT
1
1
W Hz 1m 2ster 1
Recall, for hf << kT, kT W Hz mode
2
2kT
2
2 modes
= 2(f c )
Therefore
kT =
f (T )
2
2
2
m ster
Therefore
Lec 09.6- 1
1/11/01
m = 2(f c )2 A propagation modes
L1
Photon noise
Therefore
m = 2(f c )2 A modes
p (watts ) = nsn2 hf
Each mode has 2B degrees of freedom
(2B samples sec-1 times sec) (Nyquist sampling)
Each energy state ( = hf) has 2 degrees of freedom
(sin t, cos t)
Therefore number of states ns @ hf in B:
degrees states
ns = (# modes in A )
mode degree
= (m )(2B )(1 2 ) = 2(f c )2 A(2B )(1 2 )
Lec 09.6- 2
1/11/01
L2
Photon noise
ns = 2(f c )2 AB
p = nsn2 hf
photons state = n =
n2 = n + n
1
ehf kT 1
2
f
Therefore p = 2 AB n + n (hf )2 2 W " quantum limit"
c
If boxcar h(t) has = 0.5 sec, 1-Hz post-detection bandwidth,
yields units of W Hz -1 2 (NEPR )
2
2
f
NEPR (f ) = 4 A B n + n (hf )2 W Hz 1 2
c
Noise-equivalent power due to radiation noise
Lec 09.6- 3
1/11/01
L3
Photon noise
2
2
f
(
)
NEPR f = 4 A B n + n (hf )2 W Hz 1 2
c
Noise-equivalent power due to radiation noise
(B )
NEPR for a blackbody, all frequencies
NEPR = 4A
hf 2 c
12
2
n + n df
12
SB 4
NEPR = 4 A(4kT )
T
where n(f ) =
1
ehf kT 1
[WHz 1 2 ]
Where SB Stefan-Boltzmann constant = 5.67 10-8 Wm-2K-4
and Hz-1 refers to the detector output bandwidth
Lec 09.6- 4
1/11/01
L4
Photon noise
12
SB 4
NEPR = 4 A(4kT )
T
WHz 1 2
Where SB Stefan-Boltzmann constant = 5.67 10-8 Wm-2K-4
Recall: blackbodies radiate
Pr = A SB T 4 watts (small ) , Pr = ASB T 4 if = 2
Therefore:
NEPR = [Pr 16kT ]1 2 WHz 1 2
for = 2
Therefore minimize T,A
NEPR = 4 ASBkT 5
A
Lec 09.6- 5
1/11/01
= 4
4 10 16 WHz 1 2
for A = 10 6 , T = 4K
L5
Bolometer noise analysis
Bolometers first
convert photons
to heat
(ster)
Area A
hf
p
R = f(T)
Rbias >> R
B (Hz) vo
Responsivity S
heat, Gt conductivity
Tb
cold bath
R, Rb produce Johnson noise
Radiated photons have shot noise, i.e. radiation noise
Phonon noise arises from shot noise in phonons
carrying heat to the cold bath
NEP = 4kTbR S2 + 16AkT5SB + 4kGt Tb2
Johnson noise
Lec 09.6- 6
1/11/01
W Hz-12
Photon shot noise Phonon shot noise
L6
Optical Superheterodynes
s(t) = signal
Radiometer
output
photodetector mixer
(photons s-1)
i.f . f
P photons/sec
(laser L.O.)
Let signal s( t ) =
vm(t)
( )2
h(t)
vo
communication signal
2
2S cos s t E(t ) , s (t) = S
local oscillator p( t ) =
2P cos o t
Both are coherent lasers
D = dark photons/sec
Mixer output (L.O. 0) = S + D counts/sec, where quantum efficiency
( 2So cos s t ) (volts)
= constant [(S + P + SP cos i.f .t ) + D] where i.f. = s o
v m ( t ) = constant
Lec 09.6- 7
1/11/01
We want P >> S, SP >> D
N1
Optical Superheterodyne CNR
(
= constant [(S + P +
vm ( t ) = constant 2So cos st
)2
(volts)
) ]
SP cos i.f .t + D where i.f. = s o
We want P >> S, SP >> D
Let " constant" = 1 so v m ( t ) units are counts sec
v mix ( t ) v m [signal] + v m [noise]
v m [signal] SP cos i.f .t
Conveys information in S(t), i.f .
counts sec
v m [rms noise] 2P
=
Hz
2D if D >> P )
v o = 2SP cos 2 i.f .t
Lec 09.6- 8
1/11/01
P
, = 0.5 sec for Hz
<
Conveys information in s(t) for ~
2
<< i.f .
N2
Optical Superheterodyne CNR
Let " constant" = 1 so v m ( t ) units are counts sec
v mix ( t ) v m [signal] + v m [noise]
v m [signal] SP cos i.f .t
Conveys information in S(t), i.f .
P
counts sec
=
0.5
sec
for
v m [rms noise] 2P
=
,
Hz
Hz
2D if D >> P )
v o = 2SP cos 2 i.f .t
vo
rms noise
2PB
<
Conveys information in s(t) for ~
2
<< i.f .
B for P >> S, P >> D
Define CNR Carrier-to-Noise Ratio for v m ( t ) = 2SP 4PB
B = S 16B
We assume > 1/B so provides additional noise smoothing
Lec 09.6- 9
1/11/01
CNR ~
< S 4
(best we could do is CNR 1
for 4 photons/bit)
N3
Optical Superheterodynes, Comparisons
1) Radio total-power radiometer:
CNR = TA TRMS = TA
B TR for TR >> TA
radio expression
Optical superheterodyne CNR =
S 16B = (kTAB hf ) 16B
N
photons sec
Therefore TR = 4hf/k if CNR (radio) = CNR (optical)
This is 4 times radio quantum limit if i.f. noise etc. is negligible
(Note: PA kTAB in optical)
Thus optical superheterodynes can approach quantum limit
Lec 09.6- 10
1/11/01
N4
Optical Superheterodynes, Comparisons
2) Optical non-superheterodyne if D >> S; then
v o sig = S (gain normalized )
v onoise = 2DB
CNR = S 2D
versus CNRS.H. = S 16B
Therefore a superheterodyne is better if Bi.f . < D 8
i.e. the worse D is, the higher B can be before L.O. shot
noise dominates (assuming no mixer or i.f. excess noise)
Lec 09.6- 11
1/11/01
N5
Antennas
Basic Characterization
Professor David H. Staelin
Lec 09.6- 12
1/11/01
Uses of Antennas
In
Antennas
Out
Signal
Processor
Transducer
Electromagnetic
Environment
Transducer
Antennas couple electromagnetic radiation and transmission lines
for transmission and reception
We have studied: hf << kT
hf >> kT
hf kT
All bands use antennas
Lec 09.6- 13
1/11/01
Radio
Optical
IR
A1
Antennas Characterization
y
Side Lobes
G(f,,)
PT
Transmitter
Power
PTR [W ] =
4 Pd
P(f , , ) W ster -1 radiated power
Main Lobe
" total radiated power"
Radiation efficiency:
R PTR PT
Gain (over isotropic):
P(f, , )
= RD(f , , )
G(f , , )
PT 4
Directivity (over isotropic):
P(f, , )
D(f , , )
PTR 4
Antenna pattern:
Lec 09.6- 14
1/11/01
t (f , , )
G(f, , ) D(f, , )
1
Go
Do
A2
Antenna Example
P Wm 2 = G(, )
PT
4R
P(Wm-2)
Target
R(meters)
Moon
3 108
10-5
Jupiter
1012
10-12
Antares
3 1016
10-21
MIT Haystack antenna @ =1 CM; Go 73 dB
Assume it radiates 1MWatt radar pulses
Assume kTB 1.4 10-23 10K 1Hz 10-22
Watts(say TR 10 and we use 1Hz CW radar)
Then P(Wm-2) received on Antares is comparable
to receiver noise power kTB
Lec 09.6- 15
1/11/01
A3
Receiving Properties of Antennas
Characterized by Effective Area A(f, , ):
I(f,,)
Pr(f)
Power spectral density received:
Pr (f ) = A (f , , ) [I(f , , ) ]f (W )
[(
[m2 ]
flux density S Wm 2Hz 1
)]
f, are source bandwidth, solid angle
Lec 09.6- 16
1/11/01
Recall: Radiation intensity I(f,,) received
from blackbody at temperature T is:
2kT
I(f , , ) =
Wm 2Hz 1ster 1
2
A4
Receiving System Example
Recall 1-MW radar on Antares 10-21 W/m2 on earth (GT = 73 dB)
Radio
Antares
Earth
R = 3 1016 m
Received power = A (f, , ) [I(f, , )d] B = 10 17 W from Antares
say 10 4 m2
S
10 21Wm 2Hz 1
Suppose kTB = 10-22W (recall above) then SNR = 105
Audio at 104Hz SNR = 10 (commercial opportunity?)
Lec 09.6- 17
1/11/01
A5
Relation Between A(f, ,) and G(f, ,)
Go
Ao
G(,)
A(,)
G(, ) A (, )
=
We later prove (using reciprocity) that
Go
Ao
Lec 09.6- 18
1/11/01
C1
Receiving Properties Deduced from
Reciprocity and Thermodynamics
d
Principle of detailed balance
TB(f, , )
P(f, , )
TK
Zo
Zo
Reciprocity + thermal equilibrium says that within d,
power out = power in
Antenna radiates P(f, , ) W Hz -1ster 1 into d, so
power out = P(f, , )ddf = (kTdf 4 )Gd(watts )
Antenna receives from d:
power in =
Lec 09.6- 19
1/11/01
1 2kTB (, )
df d A (f, , )[watts ]
2
2
One
polarization
Wm-2Hz-1ster-1
C2
Receiving Properties Deduced from
Reciprocity and Thermodynamics
Antenna radiates P(f, ,) [W Hz-1 ster-1] into d, so
power out = P(f, ,) d df = (kTdf/4) Gd (watts)
Antenna receives from d:
1 2kTB (, )
power in =
df d A (f, , )[watts ]
2
One
2
polarization
Wm-2Hz-1ster-1
In thermal equilibrium T = TB(f, , ); then equating
radiation and reception (detailed balance) yields
G(f , , ) =
4
2
A (f , , )
This assumes hf << kT and that powers superimpose, i.e., that the
TB (1, 1 ) signal E(t ) is uncorrelated with that for TB (2 , 2 )
Lec 09.6- 20
1/11/01
C3
Antennas Used to Provide a Radio Link
Pt
Wm 2 at receiver
Pr
Pr =
r
Gain = GT
Gr =
Ar
Pt
4r
Gt A r Watts
isotropic
m2 effective area
2
Note: Pr as r 0!, so this relation requires r > r minimum
Let Pr = Pt at rmin and At = Ar = D2 (m2) [D aperture diameter in practice]
Then
Gt A r
4 r 2min
= 1=
A t Ar
2r 2min
D4
2r 2min
Therefore rmin = D2 in practice we want r > 2D2
This zone where r ~
> 2D2 is called the " far field" of the aperture
Lec 09.6- 21
1/11/01
C4
Definition of Antenna Temperature TA(K)
kTA W Hz
(m2 ) (Wm 2Hz1ster 1)
) = 4 A(, ) I(, )d
Received power
spectral density
for a specific polarization
2kTB 1
Since I =
for thermal radiation, single polarization
2
2
Therefore
Lec 09.6- 22
1/11/01
TA =
1
2
A (, )TB (, ) d
1
=
G(, )TB (, ) d
For TB
uncorrelated
in angle
C5
Observing Small Thermal Sources TB(,,f)
Limiting case:
TA S
Go TBS A o TBS
1
G(, ) TB (, ) d =
=
=
4
4 S
2
Go/2
Go
small source, << B
TB, s
average source TB source solid angle
TAS is due to source (assume zero background)
Physical interpretation (for S << B):
Define A , " beam solid angle" so that Go A
Go
Gd = 4r
Go = 4r A
Then TA S S r TB
A
Coupling coefficient r
s, TBS
TAS
Lec 09.6- 23
1/11/01
Geometric coupling ratio =
S
A
C6
Ways to Characterize Small Thermal Sources
Limiting case:
TA S
Go TBS A oTBS
1
G(, ) TB (, ) d =
=
=
4 S
4
2
1. TB (, , f ) (for each of 2 polarizations)
2. TBS average brightness temperatur e
3. S(f ) Wm 2Hz 1 =
I(f , , ) d f(antennas )
if source small, S << A
Units of S : 1 " jansky" = 10 26 Wm 2Hz 1
Lec 09.6- 24
1/11/01
C7
