P i i l f C i iPrinciples of Communications
Chapter 3: Analog Modulation(continued)
Textbook: Ch 3, Ch 4.1-4.4, Ch 6.1-6.2
1
3 2 Eff t f N i AM S t3.2 Effect of Noise on AM SystemsBaseband system (a basis for comparison of various modulation systems):
No carrier demodulationNo carrier demodulationThe receiver is an ideal LPF with bandwidth W Noise power at the output of the receiverNoise power at the output of the receiver
0
002
W
n W
NP df N W−
= =∫Baseband SNR is given by
n(t)
2
0
R
b
PSN N W
⎛ ⎞ =⎜ ⎟⎝ ⎠
+ LPF
n(t)
)(tsm
2010/2011 Meixia Tao @ SJTU 2
+ LPF
E lExample:Find the SNR in a baseband system with a bandwidth of 5 kHz and with W/Hz. The transmitter power is 1kW and the channel attenuation is
140 / 2 10N −=
1210−power is 1kW and the channel attenuation is
Solution:
1210
12 3 9Solution: 12 3 910 10 10 WattsRP − −= × =
910 20RPS −⎛ ⎞⎜ ⎟ 14
0
2010 5000
R
bN N W −
⎛ ⎞ = = =⎜ ⎟ ×⎝ ⎠
1010log 20 13dB= =
2010/2011 Meixia Tao @ SJTU 3
Eff t f N i DSBSCEffect of Noise on DSBSC n(t)
+ BPF demod
n(t)
( )s t )(0 tm( )s t+ BPF demod
)(tni )(0 tnmodulated signal Input to the demodulator
( ) ( ) cosc cs t A m t tω=
( ) ( ) ( )( ) cos ( )cos ( )sin
i
c c c c s c
r t s t n tA m t w t n t w t n t w t
= +
= + −
Here ni(t) is a Gaussian narrow-band noise
0 / 2( ) cN f f W
S f⎧ − ≤
= ⎨
2010/2011 Meixia Tao @ SJTU 4
( )0 otherwiseinS f = ⎨
⎩
In the demodulator, the received signal is first multiplied by a locally generated sinusoid signal
( ) cos( ) ( )cos cos( )( ) cos cos( ) ( )sin cos( )
c c c c
c c c s c c
r t w t A m t w t w tn t w t w t n t w t w t
φ φφ φ
+ = +
+ + − +
1 1( )cos ( )cos(2 )2 2c c cA m t A m t w tφ φ= + +
[ ]1 ( )cos ( )sin21
c sn t n tφ φ+ +
Assume coherent detector we have
[ ]1 ( )cos(2 ) ( )sin(2 )2 c c s cn t w t n t w tφ φ+ + − +
0φ2010/2011 Meixia Tao @ SJTU
Assume coherent detector, we have5
0φ =
Then the signal is passed through a LPF with bandwidth WThen the signal is passed through a LPF with bandwidth W
[ ]1( ) ( ) ( )2 c cy t A m t n t= + ( ) ( ) ( )
c i in n c n cS f S f f S f f
for f W
= − + +
≤where
The output SNR can thus be defined as2
21 A PP A PS⎛ ⎞
for f W≤
2
0
41 24
oc
c mo c m
o nn
A PP A PSN P WNP
⎛ ⎞ = = =⎜ ⎟⎝ ⎠
2A PSince the received power of DSBSC in baseband is The output SNR can be rewritten as
42
c mR
A PP =
0
R
oDSB b
PS SN WN N
⎛ ⎞ ⎛ ⎞= =⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠
DBSSC does not provide any SNR improvement over a simple baseband systems
2010/2011 Meixia Tao @ SJTU 6
0oDSB b⎝ ⎠ ⎝ ⎠ simple baseband systems
Eff t f N i SSBEffect of Noise on SSBModulated signal:Input to the demodulator
( ) ( ) cos ( )sinc c c cs t A m t w t A m t w t= ± )
0 / 2 / 2( ) cN f f W
S f⎧ − ≤⎨
[ ] [ ]( ) ( ) cos ( ) ( ) sinA m t n t w t A m t n t w t= + + ± −)
0( )0 otherwise
cn
f fS f = ⎨
⎩( ) ( ) ( )ir t s t n t= + where
Output of LPF:
[ ] [ ]( ) ( ) cos ( ) ( ) sinc c c c s cA m t n t w t A m t n t w t= + + ±
1( ) ( ) ( )cAy t m t n t= +
Therefore, the output SNR is
( ) ( ) ( )2 2 cy t m t n t+
1 22
0
141
c mo c m
A PP A PSN P WNP
⎛ ⎞ = = =⎜ ⎟⎝ ⎠
2010/2011 Meixia Tao @ SJTU 7
0
4o
co n
nN P WNP⎝ ⎠
But in this case, 2
R c mP A P=
Thus,
PS S⎛ ⎞ ⎛ ⎞ SNR in an SSB system is
0
R
oSSB b
PS SN WN N
⎛ ⎞ ⎛ ⎞= =⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠
yequivalent to that of a DSBSC system
2010/2011 Meixia Tao @ SJTU 8
Eff t f N i C ti l AMEffect of Noise on Conventional AMModulated signal
[ ]( ) 1 ( ) cosc cs t A am t w t= +
Input to the demodulator (coherent detector)( ) ( ) ( )ir t s t n t= +
After mixing and low pass filter[ ]( ) ( ) cos ( )sinc c c c s cA A am t n t w t n t w t= + + −
[ ]11( ) ( ) ( )2 c c cy t A A am t n t= + +
Removing DC component
[ ]1( ) ( ) ( )y t A am t n t= +
2010/2011 Meixia Tao @ SJTU 9
[ ]( ) ( ) ( )2 c cy t A am t n t= +
2 21 ⎡ ⎤In this case, the received signal powerNow, we can derive the output SNR as
2 21 12R c mP A a P⎡ ⎤= +⎣ ⎦
2 22 2
14
1 2
c mc m
A a P A a PSN WN
⎛ ⎞ = =⎜ ⎟⎝ ⎠ 0
22
1 24
1
oAMc
N WNn
A P
⎜ ⎟⎝ ⎠
⎡ ⎤
Modulation efficiency
22
20
12
1
cm
m
bm
A a Pa P Sa P N W N
η⎡ ⎤+⎣ ⎦ ⎛ ⎞= ⋅ = ⎜ ⎟+ ⎝ ⎠
SNR in conventional AM is always smaller than that in baseband.
2010/2011 Meixia Tao @ SJTU 10
P f f E l D t tPerformance of Envelope-DetectorInput to the envelope-detector
[ ]( ) ( ) ( ) cos ( )sinc c c c s cr t A A am t n t w t n t w t= + + −
Envelope of r(t)
[ ]2 2( ) ( ) ( ) ( )V t A A am t n t n t= + + +
If signal component is much stronger than noise [ ]( ) ( ) ( ) ( )r c c c sV t A A am t n t n t= + + +
( ) ( ) ( )V A AAfter removing DC component, we obtain
( ) ( ) ( )r c c cV t A A am t n t≈ + +
( ) ( ) ( )c cy t A am t n t= +
At high SNR, performance of coherent detector
2010/2011 Meixia Tao @ SJTU 11
and envelop detector is the same
P f f E l D t tPerformance of Envelope-DetectorIf noise power is much stronger than the signal power
( ) ( )22 2 2( ) 1 ( ) ( ) ( ) 2 ( ) 1 ( )V t A am t n t n t A n t am t= + + + + +( ) ( )( ) 1 ( ) ( ) ( ) 2 ( ) 1 ( )r c c s c cV t A am t n t n t A n t am t+ + + + +
2 ( )A t⎡ ⎤Ignore 1st term
( ) ( )2 22 2
2 ( )( ) ( ) 1 1 ( )( ) ( )
c cc s
c s
A n tn t n t am tn t n t
⎡ ⎤≈ + + +⎢ ⎥+⎣ ⎦
2 2( ) ( ) ( )V t t t+
( )2
( )( ) 1 1 ( )( )
c cn
A n tV t am tV t
⎡ ⎤≈ + +⎢ ⎥
⎣ ⎦1 1
2εε+ ≈ +
( ) ( ) ( )n c sV t n t n t= +
( )
( )( )( ) 1 ( )
( )
n
c cn
V tA n tV t am tV t
⎣ ⎦
= + +
2The system is operating below the threshold, no meaningful SNR can be defined
2010/2011 Meixia Tao @ SJTU 12
( )nV t SNR can be defined.
E iExercise C id h h i WSS M( ) i hConsider that the message is a WSS r.p M(t) with autocorrelation function . It is given that .
2( ) 16sinc (10000 )MR τ τ=( ) 6m t =that .
We want to transmit this message to a destination via a channel with a 50dB attenuation and additive white noise
max( )
12with PSD . . We also want to achieve an SNR at the modulator output of at least 50dB. What is the required transmitted power and the channel
120( ) / 2 10 W/HznS f N −= =
What is the required transmitted power and the channel bandwidth if we employ the following modulation schemes?
DSB-SCSSBAM with modulation index = 0.8
2010/2011 Meixia Tao @ SJTU 13
3 3 A l M d l ti3.3 Angle ModulationAngle modulation is either phase or frequency of the carrier Angle modulation is either phase or frequency of the carrier is varied according to the message signal
The general form of an angle modulated wave isg g
where f = carrier freq, θ(t) is the time-varying phase and
[ ]( ) cos 2 ( )c cs t A f t tπ θ= +
where fc carrier freq, θ(t) is the time varying phase and varied by the message m(t)
The instantaneous frequency of s(t) is
1 ( )d tθ1 ( )( )2i c
d tf t fdtθ
π= +
2010/2011 Meixia Tao @ SJTU 14
R t ti f FM d PM i lRepresentation of FM and PM signalsFor phase modulation (PM), we have
( ) ( )pt k m tθ = where kp = phase deviation
For frequency modulation (FM) we have
( ) ( )p p pconstant
For frequency modulation (FM), we havewhere kf = frequency deviation constant
1( ) ( ) ( )2i c f
df t f k m t tdtθ
π− = =
The phase of FM is
deviation constant2 dtπ
The phase of FM is
0( ) 2 ( )
t
ft k m dθ π τ τ= ∫
2010/2011 Meixia Tao @ SJTU 15
Distinguishing Features of PM and FMDistinguishing Features of PM and FMNo perfect regularity in spacing of zero crossingNo perfect regularity in spacing of zero crossing
Zero crossings refer to the time instants at which a waveform changes between negative and positive values
Constant envelop, i.e. amplitude of s(t) is constant
R l ti hi b t PM d FMRelationship between PM and FM
[ ]integrator
m(t) Phasemodulator
∫ dttm )(FM wave
[ ]∫+ dttmktfA pcC )(2cos π
)2cos( tfA cC π
modulator
Discuss the
differentiator Frequencymodulator
m(t) )(tmdtd
[ ])(22cos tmktfA fcC ππ +
PM wave
Discuss the properties of FM only
2010/2011 Meixia Tao @ SJTU 16
)2cos( tfA cC π
E l Si id l M d l tiExample: Sinusoidal ModulationSinusoid Sinusoid modulating wave m(t)
FM wave
)(tmdtd
PM
dt
PM wave
2010/2011 Meixia Tao @ SJTU 17
E l S M d l tiExample: Square ModulationSquare Square modulating wave m(t)
FM wave
PM wavePM wave
2010/2011 Meixia Tao @ SJTU 18
FM by a Sinusoidal SignalFM by a Sinusoidal SignalConsider a sinusoidal modulating waveConsider a sinusoidal modulating wave
Instantaneous frequency of resulting FM wave is)2cos()( tfAtm mm π=
Instantaneous frequency of resulting FM wave is
)2cos()2cos()( tffftfAkftf mcmmfci ππ Δ+=+=
where is called the frequency deviation, proportional to the amplitude of m(t), and independent of fm.
mf Akf =Δ
Hence, the carrier phase is
( )0
( ) 2 ( ) sin(2 )t
i c mft f f d f t
fθ π τ τ πΔ
= − =∫ ( )0
sin(2 )
i c mm
m
ff tβ π=
∫
2010/2011 Meixia Tao @ SJTU 19
Where is called the modulation indexmffΔ=β
ExampleExample
Problem: i id l d l ti f lit d 5V dProblem: a sinusoidal modulating wave of amplitude 5V and frequency 1kHz is applied to a frequency modulator. The frequency sensitivity is 40Hz/V. The carrier frequency is 100kHz. Calculate (a) the f d i ti d (b) th d l ti i dfrequency deviation, and (b) the modulation index
Solution:Solution:Frequency deviation HzAkf mf 200540 =×==Δ
Modulation index 2.01000200
==Δ
=mffβ
2010/2011 Meixia Tao @ SJTU 20
Spectrum Analysis of Sinusoidal FM WaveSpectrum Analysis of Sinusoidal FM WaveThe FM wave for sinusoidal modulation is
[ ][ ] [ ] )2sin()2sin(sin)2cos()2sin(cos
)2sin(2cos)(tftfAtftfA
tftfAts mcc
ππβππβπβπ
−=+=
[ ] [ ] )2sin()2sin(sin)2cos()2sin(cos tftfAtftfA cmccmc ππβππβ
[ ])2i ()( fA β
In-phase component Quadrature-phase component
Hence, the complex envelop of FM wave is
[ ])2sin(cos)( tfAts mcI πβ= [ ])2sin(sin)( tfAts mcQ πβ=
, p p)2sin()()()(~ tfj
cQImeAtjststs πβ=+=
retains complete information about s(t))(~ ts[ ]{ } [ ]tfjtftfj
ccmc etseAts ππβπ 2)2sin(2 )(~ReRe)( == +
2010/2011 Meixia Tao @ SJTU 21
is periodic can be expanded in Fourier series as)(~ ts is periodic, can be expanded in Fourier series as
where
)(ts
∑∞
−∞=
=n
tnfjn
mects π2)(~ )2sin()(~ tfjc
meAts πβ=
[ ]
dtetsfc
m
m
m
m
f tftfj
f
f
tnfjmn
∫
∫−−=
)2/(1 2)2i (
)2/(1
)2/(1
2)(~
β
π
Let x = 2πfmt
[ ]dteAf m
m
mmf
f
tnftfjcm ∫−
−=)(
)2/(1
2)2sin( ππβ
A ( )exp sin2
cn
Ac j x nx dxπ
πβ
π −= −⎡ ⎤⎣ ⎦∫
n-th order Bessel function of the first kind Jn(β) is defined as
( )1( ) exp sin2nJ j x nx dx
π
πβ β
π= −⎡ ⎤⎣ ⎦∫
Hence,
( )2 ππ − ⎣ ⎦∫
)(βncn JAc =
2010/2011 Meixia Tao @ SJTU 22
ncn
Substituting c into )(~ ts )()( βJAtc =Substituting cn into )(ts
( )∑∞
∞=
=n
mnc tnfjJAts πβ 2exp)()(~
)()( βnc JAtc =
Hence, FM wave in time domain can be represented by−∞=n
[ ]∞⎧ ⎫
⎨ ⎬∑ [ ]
[ ]
( ) Re ( )exp 2 ( )
( ) 2 ( )
c n c mn
s t A J j f nf t
A J f f
β π
β
=−∞
∞
⎧ ⎫= +⎨ ⎬⎩ ⎭∑
∑
In frequency domain we have
[ ]( ) cos 2 ( )c n c mn
A J f nf tβ π=−∞
= +∑
In frequency-domain, we have
[ ])()()(2
)( mcmcnc nfffnfffJAfS +++−−= ∑
∞
δδβ
2010/2011 Meixia Tao @ SJTU 23
[ ])()()(2
)( mcmcn
nc
fffffff ∑−∞=
β
Property 1:Narrowband FM (for small β≤0.3 )– Approximations
2)(1)(0
≈≈
JJ
βββ ? In what ways do a
conventinal AM wave and b d FM
– Substituting above into s(t)
1,0)(2)(1
>≈≈
nJJ
n βββ a narrow band FM wave
differ from each other
g ( )
[ ]tffAtfAts mcc
cc )(2cos2
)2cos()( ++≈ πβπ
[ ]tffAmc
c )(2cos2
−− πβ
ββ 0)(J ∞→→ ββ as0)(nJ
⎩⎨⎧
=− dd)(even,)(
)(J
nJJ n
n ββ
β
2010/2011 Meixia Tao @ SJTU10/4/2004 24
⎩⎨− odd,)(
)(nJn
n ββ
Property 2: Wideband FM (for large β>1 )Property 2: Wideband FM (for large β 1 )In theory, s(t) contains a carrier and an infinite number of side-frequency components, with no approximations made
Property 3: Constant average powerThe envelop of FM wave is constant, so the average power is also constantalso constant,
The average power is also given by
2/2cAP =
[ ])(2cos)()( nffJAts += ∑∞
πβg p g y
2)(
2
22
2c
nc AJAP == ∑ β
[ ])(2cos)()( mcn
nc nffJAts +∑−∞=
πβ
22 n∑
1)(2 =∑∞
nJ β
2010/2011 Meixia Tao @ SJTU 25
)(∑−∞=n
n β
ExampleGoal: to investigate how the amplitude A and frequency fGoal: to investigate how the amplitude Am, and frequency fm, of a sinusoidal modulating wave affect the spectrum of FM wave Fixed fm and varying Am ⇒ frequency deviation Δf = kfAm and modulation index β =Δf/fm are varied
5β
1.01.0
5=β
f
1=β
cf
Increasing Am increases the number of harmonics in the fΔ2
cffΔ2
cf
2010/2011 Meixia Tao @ SJTU 26
bandwidth
Fixed Am and varying fm ⇒ Δf is fixed, but β is varied
1.01.0
1=β
1.0
5=β
fΔ2cf
fΔ2cf
Increasing fm decreases the number of harmonics but t th ti i th i b t th
f
at the same time increases the spacing between the harmonics.
2010/2011 Meixia Tao @ SJTU 27
Effective Bandwidth of FW WavesEffective Bandwidth of FW Waves
Theoretically FM bandwidth = infiniteTheoretically, FM bandwidth = infinite
In practice for a single tone FM wave when β is large B isIn practice, for a single tone FM wave, when β is large, B is only slightly greater than the total frequency excursion 2Δf. when β is small, the spectrum is effectively limited to [fc - fm, fc + fm]
C ’ R l i ti f i l t d l tiCarson’s Rule approximation for single-tone modulating wave of frequency fm
2 2 2(1 )m mB f f fβ≈ Δ + = +
2010/2011 Meixia Tao @ SJTU 28
99% bandwidth approximation99% bandwidth approximationThe separation between the two frequencies beyond which none of the side-frequencies is greater than 1% of the gunmodulated carrier amplitudei.e where nmax is the max n that satisfies mfnB max2≈
01.0)( >βnJ
β 0.1 0.3 0.5 1.0 2.0 5.0 10 20 30
2nmax 2 4 4 6 8 16 28 50 70
2010/2011 Meixia Tao @ SJTU 29
A universal curve for evaluating the 99% bandwidth As β increases, the bandwidth occupied by the significant side-frequencies drops toward that over which the carrier frequencyfrequencies drops toward that over which the carrier frequency actually deviates, i.e. B become less affected by β
20
2
2010/2011 Meixia Tao @ SJTU 30
0.2 2
FM by an Arbitrary MessageFM by an Arbitrary MessageConsider an arbitrary m(t) with highest freq component WConsider an arbitrary m(t) with highest freq component W
Define deviation ratio D = Δf / W, where max ( )ff k m tΔ =
D ⇔ β and W ⇔ fm
Carson’s rule applies as
( )2 2 2 1B f W W D≈ Δ + = +
Carson’s rule somewhat underestimate the FM bandwidth requirement, while universal curve yields a somewhat conservative result
Assess FM bandwidth between the bounds given by
2010/2011 Meixia Tao @ SJTU 31
Carson’s rule and the universal curve
E lExampleI th A i th i l f fIn north America, the maximum value of frequency deviation Δf is fixed at 75kHz for commercial FM broadcasting by ratio. If we take the modulation frequency W = 15kHz, which is typically the maximum audio frequency of interest in FM transmission the corresponding value of the deviation ratiotransmission, the corresponding value of the deviation ratio is D = 75/15 = 5Using Carson’s rule the approximate value of theUsing Carson s rule, the approximate value of the transmission bandwidth of the FM wave is
B = 2 (75+15) = 180kHz( )Using universal curve,
B = 3 2 Δf = 3 2 x 75 = 240kHz2010/2011 Meixia Tao @ SJTU 32
B = 3.2 Δf = 3.2 x 75 = 240kHz
E iExercise( )4Assuming that , determine the
transmission bandwidth of an FM modulated signal with
( )4( ) 10sinc 10m t t=
4000k =with 4000fk =
2010/2011 Meixia Tao @ SJTU 33
G ti f FMGeneration of FM wavesDirect approach
Design an oscillator whose frequency changes with the input voltage => voltage controlled oscillator (VCO)input voltage => voltage-controlled oscillator (VCO)
Indirect approachFirst generate a narrowband FM signal and thenFirst generate a narrowband FM signal and then change it to a wideband signalDue to the similarity of conventional AM signals, the y g ,generation of a narrowband FM signal is straightforward.
2010/2011 Meixia Tao @ SJTU 34
Generation of Narrow-band FMGeneration of Narrow-band FM
Consider a narrow band FM waveConsider a narrow band FM wave
where
[ ])(2cos)( 1111 ttfAts φπ +=
∫t f = carrier frequencywhere
Gi en φ (t) <<1 ith β ≤ 0 3 e ma se
∫=t
dmkt011 )(2)( ττπφ f1 = carrier frequency
k1 = frequency sensitivity
Given φ1(t) <<1 with β ≤ 0.3, we may use[ ][ ]⎩
⎨⎧
≈≈
)()(sin1)(cos 1
ttt
φφφ
Correspondingly, we may approximate s1(t) as
[ ]⎩ ≈ )()(sin 11 tt φφ
( ) ( ) ttfAtfAt )(2i2)( φ( ) ( )( ) ( )∫−=
−=t
dmtfAktfA
ttfAtfAts
011111
111111
)(2sin22cos
)(2sin2cos)(
ττπππ
φππ
2010/2011 Meixia Tao @ SJTU 35
Narrow-band FW wave
Narrow-band frequency modulator
integratorm(t) Product
Modulator +-
+Narrow-bandFM wave s (t)
q y
)2sin( 11 tfA π
-900 phaseshifter )2cos( 11 tfA π
Carrier wave
+ FM wave s1(t)
Next, pass s1(t) through a frequency multiplier, which consists of a non-linear device and a bandpass filter.
Memorylessnonlinear device
Narrow-bandFM wave
Band-passfilter
Wideband FMWave
– The input-output relationship of the non-linear device is modeled as
o ea de ce e
)()()()( 1212112 tsatsatsats n
n+++= Kmodeled as
– The BPF is used to Pass the FM wave centred at nf1 and with
2010/2011 Meixia Tao @ SJTU 36
1deviation nΔf1 and suppress all other FM spectra
Example: frequency multiplier with n = 2a p e eque cy u p e
Problem: Consider a square law device based frequency multiplierProblem: Consider a square-law device based frequency multiplier
with)()()( 2
12112 tsatsats +=
⎟⎞⎜⎛ ∫twith
Specify the midband freq. and bandwidth of BPF used in the freq. f f
⎟⎠⎞⎜
⎝⎛ += ∫
tdmktfAts
01111 )(22cos)( ττππ
multiplier for the resulting freq. deviation to be twice that at the input of the nonlinear deviceSolution:Solution:
⎞⎛⎞⎛
⎟⎠⎞⎜
⎝⎛ ++⎟
⎠⎞⎜
⎝⎛ += ∫∫
tt
tt
AaAa
dmktfAadmktfAats
22
01122
12011112 )(22cos)(22cos)( ττππττππ
⎟⎠⎞⎜
⎝⎛ +++⎟
⎠⎞⎜
⎝⎛ += ∫∫
ttdmktfAaAadmktfAa
0111212
01111 )(44cos22
)(22cos ττππττππ
fc=2f1
2010/2011 Meixia Tao @ SJTU 37
Removed by BPF with BW > 2Δf = 4Δf1
Thus connecting the narrow-band frequency modulatorThus, connecting the narrow-band frequency modulator and the frequency multiplier, we may build the wideband frequency modulator
⎞⎛q y
Message Wid b d
⎟⎠⎞⎜
⎝⎛ += ∫
tdmktfAts
01111 )(22cos)( ττππ
Integrator
Messagesignal Narrow-band
phasemodulator
WidebandFM signalFrequency
multiplier
(2 )A f tCrystal-controlled
oscillator
1cos(2 )cA f tπ
⎥⎦⎤
⎢⎣⎡ += ∫
t
fcc dmktfAts0
)(22cos)( ττππ 1
1
nkknff
f
c
==
2010/2011 Meixia Tao @ SJTU 38
⎥⎦⎢⎣ ∫01fnf Δ=Δ
MiMixerf f may not be the desired carrier frequency. The modulator performs an up/down conversion to shift the modulated signal to the desired center freq
1cf nf=
the modulated signal to the desired center freq. This consists of a mixer and a BPF
(t)s(t) v2(t)v1(t) Band-pass filter
)2cos( tflπ
2010/2011 Meixia Tao @ SJTU 39
Exercise: A typical FM transmitterExercise: A typical FM transmitter
Problem: Gi th i lifi d bl k di f t i l FMProblem: Given the simplified block diagram of a typical FM transmitter used to transmit audio signals containing frequencies in the range 100Hz to 15kHz. Desired FM wave: fc = 100MHz, Δf = 75kHz. Set β1 = 0.2 in the narrowband phase modulation to limit harmonic distortiondistortion. Specify the two-stage frequency multiplier factors n1 and n2
0.1MHz 9.5MHz
2010/2011 Meixia Tao @ SJTU 40
Demodulation of FMBalanced Frequency DiscriminatorBalanced Frequency Discriminator
Given FM wave ⎟⎠⎞⎜
⎝⎛ += ∫
t
fcc dmktfAts0
)(22cos)( ττππ⎠⎝ ∫fcc f
0)()(
0( ) 2 2 ( ) sin 2 2 ( )
t
c c f c fd s t A f k m t f t k m ddt
π π π π τ τ⎡ ⎤⎡ ⎤= − + +⎣ ⎦ ⎢ ⎥⎣ ⎦∫Hybrid modulated wave with AM and FM
Differentiator + envelop detector = FM demodulatorFrequency discriminator: a “freq to amplitude” transform device
Hybrid-modulated wave with AM and FM
Frequency discriminator: a freq to amplitude transform device
FMwave
Basebandsignal
Envelopdetector
Slope circuitH1(f)
∑+
wave
Slope circuitH2(f)
Envelopdetector
∑-
( )( )
⎪
⎪⎨
⎧+−≤≤−−−++≤≤−+−
= 2/2/,2/22/2/ ,2/2
)(1 BffBfBffajBffBfBffaj
fH ccc
ccc
ππ
)()( 12 fHfH −=
2010/2011 Meixia Tao @ SJTU 41
⎪⎩ elsewhere ,0
Thi kThink …Compared with amplitude modulation, angle modulation requires a higher implementation complexity and a higher bandwidth occupancycomplexity and a higher bandwidth occupancy.
What is the usefulness of angle modulation systems?What is the usefulness of angle modulation systems?
2010/2011 Meixia Tao @ SJTU 43
Application: FM Radio broadcastingApplication: FM Radio broadcastingAs with standard AM radio most FM radio receivers areAs with standard AM radio, most FM radio receivers are of super-heterodyne type
Typical freq parameters
Limiter
Typical freq parameters– RF carrier range = 88~108
MHzLimiter
discriminator
– Midband of IF = 10.7MHz– IF bandwidth = 200kHz
Baseband low-pass filter
l d k
– Peak freq. deviation = 75KHz
Audio amplifier with de-emphasis
loudspeaker
2010/2011 Meixia Tao @ SJTU 44
FM Radio Stereo M ltiple ingFM Radio Stereo MultiplexingStereo multiplexing is a form of FDM ∑
+ml(t)Stereo multiplexing is a form of FDM designed to transmit two separate signals via the same carrier. Widely used in FM broadcasting to
∑
∑ ∑
++
-
mr(t)+
+ m(t)
Widely used in FM broadcasting to send two different elements of a program (e.g. vocalist and accompanist in an orchestra) so as to
Frequencydoubler
K
accompanist in an orchestra) so as to give a spatial dimension to its perception by a listener at the receiving end
)2cos( tf cπ
[ ][ ]
)()()( tmtmtm rl +=receiving end [ ]
)2cos()4cos()()(
tfKtftmtm
c
crl
ππ
+−+
The sum signal is left unprocessed in its baseband formTh diff i l d 38 kH The difference signal and a 38-kHz subcarrier produce a DSBSC waveThe 19-kHz pilot is included as a
f f h t d t ti
fc = 19kHz
2010/2011 Meixia Tao @ SJTU10/4/2004 45
reference for coherent detection
FM Stereo ReceiverFM-Stereo Receiver
Baseband LPF
m(t)
∑+
++
2ml(t)ml(t)+mr(t)
To two loudspeakersBPF centered at 2fc=38kHz
Frequency
m(t)∑
+
- 2mr(t)
Baseband LPF
ml(t)-mr(t)
Narrow-band filter tuned to
f 19kH
Frequency doubler
fc=19kHz
2010/2011 Meixia Tao @ SJTU10/4/2004 46
ff f3.4 Effect of Noise on Angle Modulation
Block diagram of an angle demodulator
( ) ( )s t n t+( ) ( ) ( )r t s t n t= + ( )y t S
N⎛ ⎞⎜ ⎟⎝ ⎠BPF
BW=Bc Demod LPFBW=W
( ) ( )ws t n t+ oN⎝ ⎠
Input to the demodulator is( ) ( )s t n t+
[ ]cos ( ) ( )cos ( )sinA t t n t t n t tω φ ω ω= + + −
( )r t =
[ ]cos ( ) ( )cos ( )sinc c c c s cA t t n t t n t tω φ ω ω= + + −
( )cos( ( ))n c nR t w t tθ= +
2010/2011 Meixia Tao @ SJTU 47
( ) ( ( ))n c n
Assume that the signal is much larger than the noise( )( ) ( )cos ( ) ( )c n nr t A R t t tθ φ≈ + − ⋅⎡ ⎤⎣ ⎦
( )( )
1 ( )sin ( ) ( )cos ( )
( )cos ( ) ( )n n
cc n n
R t t tw t t tg
A R t t tθ φ
φθ φ
−⎛ ⎞−+ +⎜ ⎟⎜ ⎟+ −⎝ ⎠
( )nR t( )c n n⎝ ⎠
The phase term can be further approximated as
( )sn t( )
etθ ( ) ( )
nt
tθ
φ−
( )yR
tcA( )( )( ) ( ) sin ( ) ( )n
r nc
R tt t t tA
θ φ θ φ= + −
pp
( )n t
( )nR t( )tφ ( )y tθ ( )n tθ
2010/2011 Meixia Tao @ SJTU 48
( )cn t2 2 ( )cA E n t⎡ ⎤⎣ ⎦
Th f h f h d d l iTherefore, the output of the demodulator is
( )( )( ) ( ) ( ) sin ( ) ( )nr f n
R td dy t t k m t t tθ θ φ= = + −( )( ) ( ) ( ) ( ) ( )2 2
( ) ( )2
r f nc
f n
ydt dt A
dk m t Y tdt
φπ π
= +2f ndtπ
Desired signal Noise
The noise component is inversely proportional to the signal amplitude Ac. (This is not the case for AM system)
( ) 1R t ( ) [ ]( ) 1( ) sin ( ) ( ) ( )cos ( ) ( )sin ( )nn n s c
c c
R tY t t t n t t n t tA A
θ φ φ φ= − = −
[ ]1 ( )cos ( )sinn t n tφ φ= − (Since is slowly varying)( )tφ
2010/2011 Meixia Tao @ SJTU 49
[ ]( )cos ( )sins cc
n t n tA
φ φ= (Since is slowly varying)( )tφ
1 dThe power spectral density of is 1 ( )2 n
d Y tdtπ
2 22 24 cos sinfπ φ φ⎡ ⎤⎛ ⎞ ⎛ ⎞⎢ ⎥2
2
2
4 cos sin( ) ( ) ( )4 n s cY n n
c c
f S f f S f S fA A
f
π φ φπ
⎡ ⎤⎛ ⎞ ⎛ ⎞⎢ ⎥= +⎜ ⎟ ⎜ ⎟⎢ ⎥⎝ ⎠ ⎝ ⎠⎣ ⎦
⎧202
2
2( )
0 otherwisec
Cn c
c
f N f Bf S f AA
⎧≤⎪= = ⎨
⎪⎩⎩
( )nfS f( )noS f
At the output of LPF, the noise is limited to the freqnoise is limited to the freq. range [-W, W]
2010/2011 Meixia Tao @ SJTU 50xf− xf2TB− 2TB fW-W
Now we can determine the output SNR in FMFirst, the output signal power is
The output noise power is2
os f mP k P=
320 0
2 2
23o
W
n Wc c
N N WP f dfA A−
= =∫Then, the output SNR is
0
2 23s f c mP k A PS⎛ ⎞
⎜ ⎟23 f mP Sβ ⎛ ⎞
⎜ ⎟0
202
o
f m
o nN P W N W= =⎜ ⎟
⎝ ⎠ ( )2max ( )
f
bNm t= ⎜ ⎟
⎝ ⎠
2010/2011 Meixia Tao @ SJTU 51
Ob tiObservationsIncreasing the modulation index increases the output SNR, in contrast to AMI i th b d idth i th t t SNR
β
Increasing the bandwidth increases the output SNR.Therefore, angle modulation provides a way to trade off bandwidth for transmitted poweroff bandwidth for transmitted powerIncreasing the transmitted power increases output SNR in both FM and AM systems, but the y ,mechanisms are totally different (explain!)Increasing up to a certain value improves the βg p pperformance, but cannot continue indefinitely due to the threshold effect
β
2010/2011 Meixia Tao @ SJTU 52
Th h ld Eff tThreshold EffectThere exists a specific SNR at the input of the demodulator below which the signal is not distinguishable from the noisedistinguishable from the noise
FMS FM
DSB
0
0
NS
DSB
i
i
NS
2010/2011 Meixia Tao @ SJTU 53
0 a
C i f A l M d l tiComparison of Analog-ModulationBandwidth efficiencyBandwidth efficiency
SSB is the most bandwidth efficient, but cannot effectively transmit DCVSB is a good compromisePM/FM are the least favorable systems
Power efficiencyPower efficiencyFM provides high noise immunityConventional AM is the least power efficient
Ease of implementation (transmitter and receiver)The simplest receiver structure is conventional AMFM receivers are also easy to implementFM receivers are also easy to implementDSB-SC and SSB-SC requires coherent detector and hence is much more complicated.
2010/2011 Meixia Tao @ SJTU 54