Difference between revisions of "Aufgaben:Exercise 2.3Z: DSB-AM due to Nonlinearity"

From LNTwww
m
 
(7 intermediate revisions by 2 users not shown)
Line 1: Line 1:
  
{{quiz-Header|Buchseite=Modulationsverfahren/Zweiseitenband-Amplitudenmodulation
+
{{quiz-Header|Buchseite=Modulation_Methods/Double-Sideband_Amplitude_Modulation
 
}}
 
}}
  
 
[[File:P_ID999__Mod_Z_2_3.png|right|frame|DSB–AM with nonlinearity]]
 
[[File:P_ID999__Mod_Z_2_3.png|right|frame|DSB–AM with nonlinearity]]
In this task, we consider the realization of a double-sideband amplitude modulation using the nonlinear characteristic curve:
+
In this task, we consider the realization of a double-sideband amplitude modulation  $\text{DSB-AM}$  using the nonlinear characteristic curve:
$$y  = g(x)  =  c_1 \cdot x + c_2 \cdot x^2+ c_3 \cdot x^3\hspace{0.05cm}$$
+
:$$y  = g(x)  =  c_1 \cdot x + c_2 \cdot x^2+ c_3 \cdot x^3\hspace{0.05cm}$$
:$$ \Rightarrow   c_1 = 2,\hspace{0.2cm}c_2 = 0.25/{\rm V},\hspace{0.2cm}c_3 = 0 \hspace{0.1cm}{\rm bzw.}\hspace{0.1cm}c_3 = 0.01/{\rm V^2}\hspace{0.05cm}.$$
+
:$$ \Rightarrow \hspace{0.5cm} c_1 = 2,\hspace{0.5cm}c_2 = 0.25/{\rm V},\hspace{0.5cm}c_3 = 0 \hspace{0.5cm}{\rm or}\hspace{0.5cm}c_3 = 0.01/{\rm V^2}\hspace{0.05cm}.$$
The sum of the carrier signal and source signal is present at the input of this characteristic curve:
+
The sum of the carrier signal  $ z(t)$  and the source signal  $ q(t)$  is present at the input of this characteristic curve:
$$ x(t) = z(t) + q(t) = A_{\rm T} \cdot \cos(\omega_{\rm T} t)+ q(t),\hspace{0.2cm} A_{\rm T} = 4\,{\rm V}\hspace{0.05cm}.$$
+
:$$ x(t) = z(t) + q(t) = A_{\rm T} \cdot \cos(\omega_{\rm T} t)+ q(t),\hspace{0.2cm} A_{\rm T} = 4\,{\rm V}\hspace{0.05cm}.$$
*It is known that the source signal  $q(t)$  contains spectral components between  $1 \ \rm kHz$  and  $9 \ \rm kHz$  (up to and including these limits).
+
*It is known that the source signal  $q(t)$  contains spectral components between  $1 \ \rm kHz$  and  $9 \ \rm kHz$   (up to and including these limits).
 
*From subtask   '''(5)'''  onwards, the following source signal should be assumed:
 
*From subtask   '''(5)'''  onwards, the following source signal should be assumed:
 
:$$q(t) = A_{\rm 1} \cdot \cos(\omega_{\rm 1} t)+A_{\rm 9} \cdot \cos(\omega_{\rm 9} t) \hspace{0.05cm}.$$
 
:$$q(t) = A_{\rm 1} \cdot \cos(\omega_{\rm 1} t)+A_{\rm 9} \cdot \cos(\omega_{\rm 9} t) \hspace{0.05cm}.$$
Line 20: Line 20:
 
:$$ y_2(t)  =  c_2 \cdot[z(t) + q(t)]^2,$$  
 
:$$ y_2(t)  =  c_2 \cdot[z(t) + q(t)]^2,$$  
 
:$$y_3(t)  =  c_3 \cdot [z(t) + q(t)]^3 \hspace{0.05cm}.$$
 
:$$y_3(t)  =  c_3 \cdot [z(t) + q(t)]^3 \hspace{0.05cm}.$$
Thus, the transmitted signals  $s(t)$  i.e.,  $s_1(t)$,  $s_2(t)$  and  $s_3(t)$ , result from band-limiting to the range from  $90 \ \rm kHz$  to  $110 \ \rm kHz$.
+
Thus,  the transmitted signals  $s(t)$  i.e.,  $s_1(t)$,  $s_2(t)$  and  $s_3(t)$,  result from band-limiting to the range from  $90 \ \rm kHz$  to  $110 \ \rm kHz$.
  
  
Line 26: Line 26:
  
  
 
+
Hints:  
 
 
 
 
 
 
''Hints:''
 
 
*This exercise belongs to the chapter  [[Modulation_Methods/Double-Sideband_Amplitude_Modulation|Double-Sideband Amplitude Modulation]].
 
*This exercise belongs to the chapter  [[Modulation_Methods/Double-Sideband_Amplitude_Modulation|Double-Sideband Amplitude Modulation]].
 
*Particular reference is made to the page   [[Modulation_Methods/Double-Sideband_Amplitude_Modulation#Amplitude_modulation_with_a_quadratic_characteristic_curve|Amplitude modulation with a quadratic characteristic curve]].
 
*Particular reference is made to the page   [[Modulation_Methods/Double-Sideband_Amplitude_Modulation#Amplitude_modulation_with_a_quadratic_characteristic_curve|Amplitude modulation with a quadratic characteristic curve]].
Line 47: Line 43:
 
$f_{\rm T} \ = \ $  { 100 3%  } $\ \text{kHz}$
 
$f_{\rm T} \ = \ $  { 100 3%  } $\ \text{kHz}$
  
{What signal components does  $s_1(t)$ include?
+
{What signal components does  $s_1(t)$  include?
 
|type="[]"}
 
|type="[]"}
+ the term  $c_1 \cdot z(t)$,
+
+ The term  $c_1 \cdot z(t)$,
 
- the term  $c_1 \cdot q(t)$.
 
- the term  $c_1 \cdot q(t)$.
  
{What signal components does  $s_2(t)$ include?
+
{What signal components does  $s_2(t)$  include?
 
|type="[]"}
 
|type="[]"}
- the term  $c_2 · z^2(t)$,
+
- The term  $c_2 · z^2(t)$,
 
- the term  $c_2 · q^2(t)$,
 
- the term  $c_2 · q^2(t)$,
 
+ the term  $2c_2 · z(t) · q(t)$.
 
+ the term  $2c_2 · z(t) · q(t)$.
  
{What signal components does  $s_3(t)$  (at least partially) include?
+
{What signal components does  $s_3(t)$  (at least partially)  include?
 
|type="[]"}
 
|type="[]"}
+ the term  $c_3 · z^3(t)$,
+
+ The term  $c_3 · z^3(t)$,
- the term $3 · c_3 · z^2(t) · q(t)$,
+
- the term  $3 · c_3 · z^2(t) · q(t)$,
 
+ the term  $3 · c_3 · z(t) · q^2(t)$,
 
+ the term  $3 · c_3 · z(t) · q^2(t)$,
 
- the term  $c_3 · q^3(t)$.
 
- the term  $c_3 · q^3(t)$.
  
{Calculate &nbsp;$s(t)$, when &nbsp;$c_3 = 0$&nbsp; and the source signal &nbsp;$q(t)$&nbsp; is composed of two cosine oscillations.<br>What is the modulation depth &nbsp; $m$?
+
{Calculate &nbsp;$s(t)$,&nbsp; when &nbsp;$c_3 = 0$&nbsp; and the source signal &nbsp;$q(t)$&nbsp; is composed of two cosine oscillations.&nbsp; What is the modulation depth&nbsp; $m$?
 
|type="{}"}
 
|type="{}"}
 
$m \ = \ $ { 0.75 3% }  
 
$m \ = \ $ { 0.75 3% }  
  
{Now calculate the transmitted signal &nbsp;$s(t)$&nbsp; given&nbsp;$c_3 = \rm 0.01/V^{2}$.&nbsp; Which of the following statements is true?
+
{Now calculate the transmitted signal &nbsp;$s(t)$&nbsp; given&nbsp; $c_3 = \rm 0.01/V^{2}$.&nbsp; Which of the following statements are true?
 
|type="[]"}
 
|type="[]"}
+ Because &nbsp;$c_3 ≠ 0$&nbsp; the spectral line changes at &nbsp;$f_{\rm T}$&nbsp;.
+
+ Because &nbsp;$c_3 ≠ 0$&nbsp; the spectral line changes at &nbsp;$f_{\rm T}$.
- Because &nbsp;$c_3 ≠ 0$&nbsp; linear distortions arise, which can be compensated for.
+
- Because &nbsp;$c_3 ≠ 0$&nbsp; linear distortions arise,&nbsp; which can be compensated for.
+ Because &nbsp;$c_3 ≠ 0$&nbsp; nonlinear (i.e. irreversible) distortions arise.
+
+ Because &nbsp;$c_3 ≠ 0$&nbsp; nonlinear&nbsp; (i.e. irreversible) distortions arise.
  
 
</quiz>
 
</quiz>
Line 80: Line 76:
 
{{ML-Kopf}}
 
{{ML-Kopf}}
 
'''(1)'''&nbsp; The carrier frequency is most sensibly equal to the center frequency of the bandpass:&nbsp; $f_{\rm T}\hspace{0.15cm}\underline{ = 100\ \rm kHz}$.  
 
'''(1)'''&nbsp; The carrier frequency is most sensibly equal to the center frequency of the bandpass:&nbsp; $f_{\rm T}\hspace{0.15cm}\underline{ = 100\ \rm kHz}$.  
*If&nbsp;  $f_{\rm T}$&nbsp; does not deviate from this by more than&nbsp; $±1 \ \rm kHz$&nbsp; this also results in a "DSB-AM".
+
*If&nbsp;  $f_{\rm T}$&nbsp; does not deviate from this by more than&nbsp; $±1 \ \rm kHz$,&nbsp; this also results in a&nbsp; "DSB-AM".
  
  
  
'''(2)'''&nbsp; $s_1(t)$&nbsp; includes only the carrier&nbsp; $z(t)$ &nbsp; &rArr; &nbsp; <u>Answer 1</u>. The source signal&nbsp; $q(t)$&nbsp;  is removed by the bandpass.
+
'''(2)'''&nbsp; $s_1(t)$&nbsp; includes only the carrier&nbsp; $z(t)$ &nbsp; &rArr; &nbsp; <u>Answer 1</u>. The source signal&nbsp; $q(t)$&nbsp;  is removed by the band-pass.
  
  
  
'''(3)'''&nbsp; The quadratic term&nbsp; $z^2(t)$&nbsp;consists of a DC component &nbsp; $($at&nbsp; $f = 0)$&nbsp; and a component at &nbsp; $2f_{\rm T}$.  
+
'''(3)'''&nbsp; The quadratic term&nbsp; $z^2(t)$&nbsp; consists of a DC component &nbsp; $($at&nbsp; $f = 0)$&nbsp; and a component at &nbsp; $2f_{\rm T}$.  
*Additionally, all spectral components of &nbsp; $q^2(t)$&nbsp; are outside the bandpass.  
+
*Additionally,&nbsp; all spectral components of &nbsp; $q^2(t)$&nbsp; are outside the band-pass.  
*Therefore the <u>last answer</u> is correct.
+
*Therefore the&nbsp; <u>last answer</u>&nbsp; is correct.
  
  
  
'''(4)'''&nbsp;<u>Answers 1 and 3</u> are correct:
+
'''(4)'''&nbsp;<u>Answers 1 and 3</u>&nbsp; are correct:
 
*The term&nbsp; $\cos^3(ω_Tt)$&nbsp; has its largest signal component at&nbsp; $f = f_{\rm T}$.  
 
*The term&nbsp; $\cos^3(ω_Tt)$&nbsp; has its largest signal component at&nbsp; $f = f_{\rm T}$.  
 
*The third answer&nbsp; $(3 · c_3 · z(t) · q^2(t))$&nbsp; lies between&nbsp; $100\ \rm  kHz ± 18 \ \rm  kHz $.  
 
*The third answer&nbsp; $(3 · c_3 · z(t) · q^2(t))$&nbsp; lies between&nbsp; $100\ \rm  kHz ± 18 \ \rm  kHz $.  
*Some parts – namely the frequency components between &nbsp; $90\ \rm  kHz $ and $110 \ \rm  kHz$&nbsp; – are not removed by the bandpass and are thus also included in &nbsp; $s(t)$&nbsp;.  
+
*Some parts – namely the frequency components between &nbsp; $90\ \rm  kHz $ and $110 \ \rm  kHz$&nbsp; – are not removed by the band-pass and are thus also included in&nbsp; $s(t)$.  
  
  
  
[[File:P_ID993__Mod_Z_2_3_f.png|right|frame|Generated AM Spectrum]]
+
[[File:P_ID993__Mod_Z_2_3_f.png|right|frame|Generated DSB-AM spectrum]]
 
'''(5)'''&nbsp; The transmitted signal consists of a total of five frequencies:
 
'''(5)'''&nbsp; The transmitted signal consists of a total of five frequencies:
 
:$$s(t)  =  c_1 \cdot A_{\rm T} \cdot \cos(\omega_{\rm T} t)+ c_2 \cdot A_{\rm T} \cdot A_{\rm 1} \cdot \cos((\omega_{\rm T} \pm \omega_{\rm 1})t) + c_2 \cdot A_{\rm T} \cdot A_{\rm 2} \cdot \cos((\omega_{\rm T} \pm \omega_{\rm 2})t) \hspace{0.05cm}.$$
 
:$$s(t)  =  c_1 \cdot A_{\rm T} \cdot \cos(\omega_{\rm T} t)+ c_2 \cdot A_{\rm T} \cdot A_{\rm 1} \cdot \cos((\omega_{\rm T} \pm \omega_{\rm 1})t) + c_2 \cdot A_{\rm T} \cdot A_{\rm 2} \cdot \cos((\omega_{\rm T} \pm \omega_{\rm 2})t) \hspace{0.05cm}.$$
  
*Note that the second and third terms each contain two signal frequencies::  
+
*Note that the second term and the third term each contain two signal frequencies::  
 
**$\text{99 kHz}$&nbsp; and &nbsp;$\text{101 kHz}$ and  
 
**$\text{99 kHz}$&nbsp; and &nbsp;$\text{101 kHz}$ and  
 
**$\text{91 kHz}$&nbsp; and&nbsp; $\text{109 kHz}$, respectively.  
 
**$\text{91 kHz}$&nbsp; and&nbsp; $\text{109 kHz}$, respectively.  
  
  
*Given &nbsp; $A_{\rm T} = 4 \ \rm  V$,&nbsp; $A_1 = 1 V$,&nbsp; $A_9 = 2  \ \rm  V$,&nbsp; $c_1 = 1$&nbsp; and&nbsp; $c_2 = 1/A_{\rm T} = \rm 0.25/V$&nbsp;, it holds that:
+
*Given &nbsp; $A_{\rm T} = 4 \ \rm  V$,&nbsp; $A_1 = 1 V$,&nbsp; $A_9 = 2  \ \rm  V$,&nbsp; $c_1 = 1$&nbsp; and&nbsp; $c_2 = 1/A_{\rm T} = \rm 0.25/V$,&nbsp; it holds that:
 
:$$s(t) = 4\,{\rm V} \cdot \cos(\omega_{\rm T} t) + 1\,{\rm V} \cdot \cos((\omega_{\rm T} \pm \omega_{\rm 1})t) + 2\,{\rm V}\cdot \cos((\omega_{\rm T} \pm \omega_{\rm 2})t) \hspace{0.05cm}.$$
 
:$$s(t) = 4\,{\rm V} \cdot \cos(\omega_{\rm T} t) + 1\,{\rm V} \cdot \cos((\omega_{\rm T} \pm \omega_{\rm 1})t) + 2\,{\rm V}\cdot \cos((\omega_{\rm T} \pm \omega_{\rm 2})t) \hspace{0.05cm}.$$
  
Line 118: Line 114:
  
  
'''(6)'''&nbsp; The graph above shows the spectrum &nbsp; $S_+(f)$&nbsp; – that is, only positive frequencies – with&nbsp; $c_3 = 0$.&nbsp;  
+
'''(6)'''&nbsp; The graph above shows the spectrum &nbsp; $S_+(f)$&nbsp; – that is, only positive frequencies – where&nbsp; $c_3 = 0$.&nbsp;  
 
* When&nbsp; $c_3 ≠ 0$&nbsp; the following additional spectral components arise:
 
* When&nbsp; $c_3 ≠ 0$&nbsp; the following additional spectral components arise:
 
:$$c_3 \cdot z^3(t)= \frac{c_3 \cdot A_{\rm T}^3}{4} \cdot \left[ 3 \cdot \cos(\omega_{\rm T} t) + \cos(3\omega_{\rm T} t)\right] \hspace{0.05cm}.$$
 
:$$c_3 \cdot z^3(t)= \frac{c_3 \cdot A_{\rm T}^3}{4} \cdot \left[ 3 \cdot \cos(\omega_{\rm T} t) + \cos(3\omega_{\rm T} t)\right] \hspace{0.05cm}.$$
*The first component falls in the passband of the bandpass.  As a result, the Dirac weight at &nbsp; $f_{\rm T} = 100\ \rm kHz$&nbsp; is increased from the original&nbsp; $8 \ \rm V$&nbsp; to&nbsp; $\text{8 V + 0.75 · 0.01/V}^2 · 4^3 \text{ V}^3 = 8.48 \ \rm V$&nbsp;.
+
*The first component falls in the pass-band of the band-pass.&nbsp; As a result,&nbsp; the Dirac weight at &nbsp; $f_{\rm T} = 100\ \rm kHz$&nbsp; is increased from the original&nbsp; $8 \ \rm V$&nbsp; to&nbsp;  
 
+
:$$\text{8 V + 0.75 · 0.01/V}^2 · 4^3 \text{ V}^3 = 8.48 \ \rm V.$$
  
*Weiterhin liefert der dritte Spektralanteil von Teilaufgabe&nbsp; '''(4)'''&nbsp; einen unerwünschten Beitrag zu&nbsp; $S_+(f)$.&nbsp; Dabei gilt:
+
*Furthermore,&nbsp; the third spectral component of subtask&nbsp; '''(4)'''&nbsp; provides an unwanted contribution to&nbsp; $S_+(f)$.&nbsp; In this case:
 
:$$q^2(t)  =  \left[A_{\rm 1} \cdot \cos(\omega_{\rm 1} t)+A_{\rm 9} \cdot \cos(\omega_{\rm 9} t)\right]^2 = A_{\rm 1}^2 \cdot \cos^2(\omega_{\rm 1} t)+ A_{\rm 9}^2 \cdot \cos^2(\omega_{\rm 9}t) +  
 
:$$q^2(t)  =  \left[A_{\rm 1} \cdot \cos(\omega_{\rm 1} t)+A_{\rm 9} \cdot \cos(\omega_{\rm 9} t)\right]^2 = A_{\rm 1}^2 \cdot \cos^2(\omega_{\rm 1} t)+ A_{\rm 9}^2 \cdot \cos^2(\omega_{\rm 9}t) +  
 
   2 \cdot A_{\rm 1} \cdot A_{\rm 9} \cdot \cos(\omega_{\rm 1} t)\cdot \cos(\omega_{\rm 9} t)$$
 
   2 \cdot A_{\rm 1} \cdot A_{\rm 9} \cdot \cos(\omega_{\rm 1} t)\cdot \cos(\omega_{\rm 9} t)$$
 
:$$ \Rightarrow \hspace{0.2cm} q^2(t)  =  \frac{A_{\rm 1}^2}{2} +\frac{A_{\rm 1}^2}{2} \cdot \cos(\omega_{\rm 2} t)+ \frac{A_{\rm 9}^2}{2} + \frac{A_{\rm 9}^2}{2} \cdot \cos(\omega_{\rm 18} t) +  A_{\rm 1} \cdot A_{\rm 9} \cdot \cos(\omega_{\rm 8} t)+ A_{\rm 1} \cdot A_{\rm 9} \cdot \cos(\omega_{\rm 10} t).$$
 
:$$ \Rightarrow \hspace{0.2cm} q^2(t)  =  \frac{A_{\rm 1}^2}{2} +\frac{A_{\rm 1}^2}{2} \cdot \cos(\omega_{\rm 2} t)+ \frac{A_{\rm 9}^2}{2} + \frac{A_{\rm 9}^2}{2} \cdot \cos(\omega_{\rm 18} t) +  A_{\rm 1} \cdot A_{\rm 9} \cdot \cos(\omega_{\rm 8} t)+ A_{\rm 1} \cdot A_{\rm 9} \cdot \cos(\omega_{\rm 10} t).$$
*Nach der Multiplikation mit&nbsp; $z(t)$&nbsp; fallen alle diese Beiträge bis auf den vierten in den Bereich von&nbsp; $\text{90 kHz}$&nbsp; bis&nbsp; $\text{110 kHz}$.&nbsp; Das Gewicht bei&nbsp; $f_{\rm T} = 100\ \rm kHz$&nbsp;  wird um&nbsp; $3 · c_3 · A_{\rm T} · 0.5 (A_1^2 + A_9^2) = 0.6\ \rm  V$&nbsp; weiter erhöht und ist somit&nbsp; $9.08 \ \rm V$.  
+
*After multiplying by &nbsp; $z(t)$,&nbsp; all but the fourth of these contributions fall within the &nbsp; $\text{90 kHz}$&nbsp; to&nbsp; $\text{110 kHz}$ range.&nbsp; The weight at&nbsp; $f_{\rm T} = 100\ \rm kHz$&nbsp;  is further increased by&nbsp; $3 · c_3 · A_{\rm T} · 0.5 (A_1^2 + A_9^2) = 0.6\ \rm  V$&nbsp; and is thus &nbsp; $9.08 \ \rm V$.  
  
  
Weitere Anteile ergeben sich bei:
+
Further components are obtained at:
*$98 \ \rm kHz$&nbsp; und&nbsp; $102 \ \rm kHz$&nbsp; mit den Gewichten&nbsp; $c_3 · A_{\rm T}/2 · A_1^2/2 = 0.03\ \rm  V$,
+
*$98 \ \rm kHz$&nbsp; and&nbsp; $102 \ \rm kHz$&nbsp; with weights&nbsp; $c_3 · A_{\rm T}/2 · A_1^2/2 = 0.03\ \rm  V$,
* $92 \ \rm kHz$&nbsp; und&nbsp; $108 \ \rm kHz$&nbsp; mit den Gewichten&nbsp; $3c_3 · A_{\rm T}/2 · A_1 · A_9 = 0.12\ \rm  V$,
+
* $92 \ \rm kHz$&nbsp; and&nbsp; $108 \ \rm kHz$&nbsp; with weights&nbsp; $3c_3 · A_{\rm T}/2 · A_1 · A_9 = 0.12\ \rm  V$,
* $90 \ \rm kHz$&nbsp; und&nbsp; $110 \ \rm kHz$&nbsp; mit den Gewichten&nbsp; $3c_3 · A_{\rm T}/2 · A_1 · A_9 = 0.12\ \rm  V$.
+
* $90 \ \rm kHz$&nbsp; and&nbsp; $110 \ \rm kHz$&nbsp; with weights&nbsp; $3c_3 · A_{\rm T}/2 · A_1 · A_9 = 0.12\ \rm  V$.
  
  
Die untereSkizze in obiger  Grafik zeigt das Spektrum&nbsp; $S_+(f)$&nbsp; unter Berücksichtigung der kubischen Anteile.&nbsp; Man erkennt, dass neue Frequenzen entstanden sind, was auf nichtlineare Verzerrungen hindeutet.&nbsp; Richtig sind somit die <u>Lösungsvorschläge 1 und 3</u>.
+
The lower plot in the above graph shows the spectrum &nbsp; $S_+(f)$&nbsp; considering the cubic components.&nbsp; It can be seen that new frequencies have appeared, indicating nonlinear distortions.&nbsp; The correct solutions are therefore <u>Answers 1 and 3</u>.
  
 
{{ML-Fuß}}
 
{{ML-Fuß}}

Latest revision as of 17:02, 24 March 2022

DSB–AM with nonlinearity

In this task, we consider the realization of a double-sideband amplitude modulation  $\text{DSB-AM}$  using the nonlinear characteristic curve:

$$y = g(x) = c_1 \cdot x + c_2 \cdot x^2+ c_3 \cdot x^3\hspace{0.05cm}$$
$$ \Rightarrow \hspace{0.5cm} c_1 = 2,\hspace{0.5cm}c_2 = 0.25/{\rm V},\hspace{0.5cm}c_3 = 0 \hspace{0.5cm}{\rm or}\hspace{0.5cm}c_3 = 0.01/{\rm V^2}\hspace{0.05cm}.$$

The sum of the carrier signal  $ z(t)$  and the source signal  $ q(t)$  is present at the input of this characteristic curve:

$$ x(t) = z(t) + q(t) = A_{\rm T} \cdot \cos(\omega_{\rm T} t)+ q(t),\hspace{0.2cm} A_{\rm T} = 4\,{\rm V}\hspace{0.05cm}.$$
  • It is known that the source signal  $q(t)$  contains spectral components between  $1 \ \rm kHz$  and  $9 \ \rm kHz$   (up to and including these limits).
  • From subtask   (5)  onwards, the following source signal should be assumed:
$$q(t) = A_{\rm 1} \cdot \cos(\omega_{\rm 1} t)+A_{\rm 9} \cdot \cos(\omega_{\rm 9} t) \hspace{0.05cm}.$$
  • Let the angular frequencies be  $ω_1 = 2 π · 1 \ \rm kHz$  and  $ω_9 = 2 π · 9\ \rm kHz$.  The corresponding amplitudes are given as:  $A_1 = 1\ \rm V$  and  $A_9 = 2\ \rm V$.


The following abbreviations are used in the questions for this task:

$$ y(t) = y_1(t) + y_2(t)+y_3(t),$$
$$y_1(t) = c_1 \cdot [z(t) + q(t)],$$
$$ y_2(t) = c_2 \cdot[z(t) + q(t)]^2,$$
$$y_3(t) = c_3 \cdot [z(t) + q(t)]^3 \hspace{0.05cm}.$$

Thus,  the transmitted signals  $s(t)$  i.e.,  $s_1(t)$,  $s_2(t)$  and  $s_3(t)$,  result from band-limiting to the range from  $90 \ \rm kHz$  to  $110 \ \rm kHz$.



Hints:

  • The following trigonometric transformations are given:
$$ \cos^2(\alpha) = {1}/{2} \cdot \left[ 1 + \cos(2\alpha)\right] \hspace{0.05cm}, \hspace{0.5cm} \cos^3(\alpha) = {1}/{4} \cdot \left[ 3 \cdot \cos(\alpha) + \cos(3\alpha)\right] \hspace{0.05cm}.$$


Questions

1

What is a reasonable choice for the carrier frequency?

$f_{\rm T} \ = \ $

$\ \text{kHz}$

2

What signal components does  $s_1(t)$  include?

The term  $c_1 \cdot z(t)$,
the term  $c_1 \cdot q(t)$.

3

What signal components does  $s_2(t)$  include?

The term  $c_2 · z^2(t)$,
the term  $c_2 · q^2(t)$,
the term  $2c_2 · z(t) · q(t)$.

4

What signal components does  $s_3(t)$  (at least partially)  include?

The term  $c_3 · z^3(t)$,
the term  $3 · c_3 · z^2(t) · q(t)$,
the term  $3 · c_3 · z(t) · q^2(t)$,
the term  $c_3 · q^3(t)$.

5

Calculate  $s(t)$,  when  $c_3 = 0$  and the source signal  $q(t)$  is composed of two cosine oscillations.  What is the modulation depth  $m$?

$m \ = \ $

6

Now calculate the transmitted signal  $s(t)$  given  $c_3 = \rm 0.01/V^{2}$.  Which of the following statements are true?

Because  $c_3 ≠ 0$  the spectral line changes at  $f_{\rm T}$.
Because  $c_3 ≠ 0$  linear distortions arise,  which can be compensated for.
Because  $c_3 ≠ 0$  nonlinear  (i.e. irreversible) distortions arise.


Solution

(1)  The carrier frequency is most sensibly equal to the center frequency of the bandpass:  $f_{\rm T}\hspace{0.15cm}\underline{ = 100\ \rm kHz}$.

  • If  $f_{\rm T}$  does not deviate from this by more than  $±1 \ \rm kHz$,  this also results in a  "DSB-AM".


(2)  $s_1(t)$  includes only the carrier  $z(t)$   ⇒   Answer 1. The source signal  $q(t)$  is removed by the band-pass.


(3)  The quadratic term  $z^2(t)$  consists of a DC component   $($at  $f = 0)$  and a component at   $2f_{\rm T}$.

  • Additionally,  all spectral components of   $q^2(t)$  are outside the band-pass.
  • Therefore the  last answer  is correct.


(4) Answers 1 and 3  are correct:

  • The term  $\cos^3(ω_Tt)$  has its largest signal component at  $f = f_{\rm T}$.
  • The third answer  $(3 · c_3 · z(t) · q^2(t))$  lies between  $100\ \rm kHz ± 18 \ \rm kHz $.
  • Some parts – namely the frequency components between   $90\ \rm kHz $ and $110 \ \rm kHz$  – are not removed by the band-pass and are thus also included in  $s(t)$.


Generated DSB-AM spectrum

(5)  The transmitted signal consists of a total of five frequencies:

$$s(t) = c_1 \cdot A_{\rm T} \cdot \cos(\omega_{\rm T} t)+ c_2 \cdot A_{\rm T} \cdot A_{\rm 1} \cdot \cos((\omega_{\rm T} \pm \omega_{\rm 1})t) + c_2 \cdot A_{\rm T} \cdot A_{\rm 2} \cdot \cos((\omega_{\rm T} \pm \omega_{\rm 2})t) \hspace{0.05cm}.$$
  • Note that the second term and the third term each contain two signal frequencies::
    • $\text{99 kHz}$  and  $\text{101 kHz}$ and
    • $\text{91 kHz}$  and  $\text{109 kHz}$, respectively.


  • Given   $A_{\rm T} = 4 \ \rm V$,  $A_1 = 1 V$,  $A_9 = 2 \ \rm V$,  $c_1 = 1$  and  $c_2 = 1/A_{\rm T} = \rm 0.25/V$,  it holds that:
$$s(t) = 4\,{\rm V} \cdot \cos(\omega_{\rm T} t) + 1\,{\rm V} \cdot \cos((\omega_{\rm T} \pm \omega_{\rm 1})t) + 2\,{\rm V}\cdot \cos((\omega_{\rm T} \pm \omega_{\rm 2})t) \hspace{0.05cm}.$$
  • From this it can be seen that for the modulation depth:
$$m =\frac{A_1 + A_9}{A_{\rm T}} = \rm \frac{1\ V + 2 \ V}{4 \ V} \hspace{0.15cm}\underline{=0.75}.$$


(6)  The graph above shows the spectrum   $S_+(f)$  – that is, only positive frequencies – where  $c_3 = 0$. 

  • When  $c_3 ≠ 0$  the following additional spectral components arise:
$$c_3 \cdot z^3(t)= \frac{c_3 \cdot A_{\rm T}^3}{4} \cdot \left[ 3 \cdot \cos(\omega_{\rm T} t) + \cos(3\omega_{\rm T} t)\right] \hspace{0.05cm}.$$
  • The first component falls in the pass-band of the band-pass.  As a result,  the Dirac weight at   $f_{\rm T} = 100\ \rm kHz$  is increased from the original  $8 \ \rm V$  to 
$$\text{8 V + 0.75 · 0.01/V}^2 · 4^3 \text{ V}^3 = 8.48 \ \rm V.$$
  • Furthermore,  the third spectral component of subtask  (4)  provides an unwanted contribution to  $S_+(f)$.  In this case:
$$q^2(t) = \left[A_{\rm 1} \cdot \cos(\omega_{\rm 1} t)+A_{\rm 9} \cdot \cos(\omega_{\rm 9} t)\right]^2 = A_{\rm 1}^2 \cdot \cos^2(\omega_{\rm 1} t)+ A_{\rm 9}^2 \cdot \cos^2(\omega_{\rm 9}t) + 2 \cdot A_{\rm 1} \cdot A_{\rm 9} \cdot \cos(\omega_{\rm 1} t)\cdot \cos(\omega_{\rm 9} t)$$
$$ \Rightarrow \hspace{0.2cm} q^2(t) = \frac{A_{\rm 1}^2}{2} +\frac{A_{\rm 1}^2}{2} \cdot \cos(\omega_{\rm 2} t)+ \frac{A_{\rm 9}^2}{2} + \frac{A_{\rm 9}^2}{2} \cdot \cos(\omega_{\rm 18} t) + A_{\rm 1} \cdot A_{\rm 9} \cdot \cos(\omega_{\rm 8} t)+ A_{\rm 1} \cdot A_{\rm 9} \cdot \cos(\omega_{\rm 10} t).$$
  • After multiplying by   $z(t)$,  all but the fourth of these contributions fall within the   $\text{90 kHz}$  to  $\text{110 kHz}$ range.  The weight at  $f_{\rm T} = 100\ \rm kHz$  is further increased by  $3 · c_3 · A_{\rm T} · 0.5 (A_1^2 + A_9^2) = 0.6\ \rm V$  and is thus   $9.08 \ \rm V$.


Further components are obtained at:

  • $98 \ \rm kHz$  and  $102 \ \rm kHz$  with weights  $c_3 · A_{\rm T}/2 · A_1^2/2 = 0.03\ \rm V$,
  • $92 \ \rm kHz$  and  $108 \ \rm kHz$  with weights  $3c_3 · A_{\rm T}/2 · A_1 · A_9 = 0.12\ \rm V$,
  • $90 \ \rm kHz$  and  $110 \ \rm kHz$  with weights  $3c_3 · A_{\rm T}/2 · A_1 · A_9 = 0.12\ \rm V$.


The lower plot in the above graph shows the spectrum   $S_+(f)$  considering the cubic components.  It can be seen that new frequencies have appeared, indicating nonlinear distortions.  The correct solutions are therefore Answers 1 and 3.