Difference between revisions of "Aufgaben:Exercise 2.3: Sinusoidal Characteristic"

From LNTwww
 
(9 intermediate revisions by 2 users not shown)
Line 4: Line 4:
  
 
[[File:P_ID894__LZI_A_2_3.png|right|frame|Sinusoidal characteristic curve]]
 
[[File:P_ID894__LZI_A_2_3.png|right|frame|Sinusoidal characteristic curve]]
We consider a system with input  $x(t)$  and output  $y(t)$. For simplicity of description, the signals are considered to be dimensionless.
+
We consider a system with input  $x(t)$  and output  $y(t)$.  For simplicity of description, the signals are considered to be dimensionless.
  
The relationship between the input signal  $x(t)$  and the output signal  $y(t)$  is given in the range between  $-\pi/2$  and  $+\pi/2$  by the following characteristic curve.
+
The relationship between the input signal  $x(t)$  and the output signal  $y(t)$  is given by the following characteristic curve in the range between  $-\pi/2$  and  $+\pi/2$:
 
:$$g(x) =  \sin(x) = x - \frac{x^3}{3!} + \frac{x^5}{5!} -
 
:$$g(x) =  \sin(x) = x - \frac{x^3}{3!} + \frac{x^5}{5!} -
 
  \hspace{0.05cm}\text{...}$$
 
  \hspace{0.05cm}\text{...}$$
Line 17: Line 17:
 
:$$g_5(x) = x- x^3\hspace{-0.1cm}/{6}+x^5\hspace{-0.1cm}/{120}\hspace{0.05cm}.$$
 
:$$g_5(x) = x- x^3\hspace{-0.1cm}/{6}+x^5\hspace{-0.1cm}/{120}\hspace{0.05cm}.$$
  
The input signal  $x(t) = A \cdot \cos(\omega_0 \cdot t)$  is always assumed wherat the values  $A = 0.5$,  $A = 1.0$  and  $A = 1.5$  are to be considered for the (dimensionless) signal amplitude.
+
*The input signal  $x(t) = A \cdot \cos(\omega_0 \cdot t)$  is always assumed. 
 +
*The values  $A = 0.5$,  $A = 1.0$  and  $A = 1.5$  are to be considered for the (dimensionless) signal amplitude.
  
  
Line 27: Line 28:
  
 
''Please note:''  
 
''Please note:''  
*The task belongs to the chapter   [[Linear_and_Time_Invariant_Systems/Nonlinear_Distortion|Nonlinear Distortion]].
+
*The task belongs to the chapter   [[Linear_and_Time_Invariant_Systems/Nonlinear_Distortion|Nonlinear Distortions]].
 
*The resulting signal curves for  $x(t)$  and  $y(t)$  are shown graphically on the page  [[Linear_and_Time_Invariant_Systems/Nonlinear_Distortion#Description_of_nonlinear_systems|Description of nonlinear systems]] .
 
*The resulting signal curves for  $x(t)$  and  $y(t)$  are shown graphically on the page  [[Linear_and_Time_Invariant_Systems/Nonlinear_Distortion#Description_of_nonlinear_systems|Description of nonlinear systems]] .
 
 
*All powers required here refer to the resistance  $R = 1 \ \rm \Omega$  and thus have the unit  ${\rm V}^2$.
 
*All powers required here refer to the resistance  $R = 1 \ \rm \Omega$  and thus have the unit  ${\rm V}^2$.
 
*The following trigonometric relations are assumed to be known:
 
*The following trigonometric relations are assumed to be known:
Line 58: Line 58:
  
  
{Which of the following statements are true? Here,&nbsp; $K$&nbsp; denotes the distortion factor of the sine function&nbsp; $g(x)$. <br>$K_{\rm g3}$&nbsp; and&nbsp; $K_{\rm g5}$&nbsp; are based on the approximations&nbsp; $g_3(x)$&nbsp; and&nbsp; $g_5(x)$, respectively.
+
{Which of the following statements are true?&nbsp; Here,&nbsp; $K$&nbsp; denotes the distortion factor of the sine function&nbsp; $g(x)$. <br>$K_{\rm g3}$&nbsp; and&nbsp; $K_{\rm g5}$&nbsp; are based on the approximations&nbsp; $g_3(x)$&nbsp; and&nbsp; $g_5(x)$, respectively.
 
|type="[]"}
 
|type="[]"}
 
+ $K_{\rm g5}$&nbsp; generally represents a better approximation for&nbsp; $K$&nbsp; than&nbsp; $K_{\rm g3}$.
 
+ $K_{\rm g5}$&nbsp; generally represents a better approximation for&nbsp; $K$&nbsp; than&nbsp; $K_{\rm g3}$.
- $K_{\rm g3} < K_{\rm g5}$ holds for&nbsp; $A = 1.0$ &nbsp;.
+
- $K_{\rm g3} < K_{\rm g5}$ holds for&nbsp; $A = 1.0$.
+ $K_{\rm g3} \approx K_{\rm g5}$ will hold for&nbsp; $A = 0.5$ &nbsp;.
+
+ $K_{\rm g3} \approx K_{\rm g5}$ will hold for&nbsp; $A = 0.5$.
  
  
Line 74: Line 74:
  
  
'''(2)'''&nbsp; The analytical spectrum (positive frequencies only) of the input signal is:
+
'''(2)'''&nbsp; The analytical spectrum&nbsp; (positive frequencies only)&nbsp; of the input signal is:
 
:$$X_+(f) = A  \cdot {\rm \delta}(f- f_0) .$$
 
:$$X_+(f) = A  \cdot {\rm \delta}(f- f_0) .$$
  
Line 88: Line 88:
 
:$$A_1 = A  - {A^3}\hspace{-0.1cm}/{8},  \hspace{0.5cm}A_3 =  - {A^3}\hspace{-0.1cm}/{24}.$$
 
:$$A_1 = A  - {A^3}\hspace{-0.1cm}/{8},  \hspace{0.5cm}A_3 =  - {A^3}\hspace{-0.1cm}/{24}.$$
  
*Using $A = 0.5$ the following is obtained: $A_1 \approx 0.484$ and&nbsp; $A_3 \approx 0.005$. Thus, the distortion factor is:
+
*Using&nbsp; $A = 0.5$&nbsp; the following is obtained:&nbsp; $A_1 \approx 0.484$&nbsp; and&nbsp; $A_3 \approx 0.005$.&nbsp; Thus, the distortion factor is:
 
:$$K = K_3 ={|A_3|}/{A_1}= {0.005}/{0.484} \hspace{0.15cm}\underline{ =  1.08\%}.$$
 
:$$K = K_3 ={|A_3|}/{A_1}= {0.005}/{0.484} \hspace{0.15cm}\underline{ =  1.08\%}.$$
  
 
:Note that for the approximation&nbsp; $g_3(x)$&nbsp; only the cubic part&nbsp; $K_3$&nbsp; of the distortion factor is effective.  
 
:Note that for the approximation&nbsp; $g_3(x)$&nbsp; only the cubic part&nbsp; $K_3$&nbsp; of the distortion factor is effective.  
  
*For&nbsp; $A = 1.0$&nbsp; and&nbsp; $A = 1.5$&nbsp; the following numerical values arise as aresult:
+
*For&nbsp; $A = 1.0$&nbsp; and&nbsp; $A = 1.5$&nbsp; the following numerical values:
 
:$$A = 1.0: A_1 \approx 0.875, \hspace{0.2cm} A_3 \approx
 
:$$A = 1.0: A_1 \approx 0.875, \hspace{0.2cm} A_3 \approx
 
-0.041\hspace{0.3cm} \Rightarrow \hspace{0.3cm} \hspace{0.15cm}\underline{K \approx 4.76\%}\; \; \Rightarrow \; \; K_{g3},$$
 
-0.041\hspace{0.3cm} \Rightarrow \hspace{0.3cm} \hspace{0.15cm}\underline{K \approx 4.76\%}\; \; \Rightarrow \; \; K_{g3},$$
Line 101: Line 101:
  
  
'''(3)'''&nbsp; Similarly as in subtask&nbsp; '''(2)'''&nbsp;  
+
'''(3)'''&nbsp; Similarly as in subtask&nbsp; '''(2)''',&nbsp;  
 
:$$y(t) = A_1 \cdot {\rm cos}(\omega_0  t ) + A_3 \cdot {\rm
 
:$$y(t) = A_1 \cdot {\rm cos}(\omega_0  t ) + A_3 \cdot {\rm
cos}(3\omega_0  t )+ A_5 \cdot {\rm cos}(5\omega_0  t )$$
+
cos}(3\omega_0  t )+ A_5 \cdot {\rm cos}(5\omega_0  t )$$  
  
:with the following coefficients:
+
:holds with the following coefficients:
 
:$$A_1 = A  - {A^3}\hspace{-0.1cm}/{8} + {A^5}\hspace{-0.1cm}/{192},\hspace{0.3cm}
 
:$$A_1 = A  - {A^3}\hspace{-0.1cm}/{8} + {A^5}\hspace{-0.1cm}/{192},\hspace{0.3cm}
 
A_3 =  - {A^3}\hspace{-0.1cm}/{24} +  {A^5}\hspace{-0.1cm}/{384},\hspace{0.3cm}
 
A_3 =  - {A^3}\hspace{-0.1cm}/{24} +  {A^5}\hspace{-0.1cm}/{384},\hspace{0.3cm}
 
A_5 =  {A^5}\hspace{-0.1cm}/{1920}.$$
 
A_5 =  {A^5}\hspace{-0.1cm}/{1920}.$$
  
*Daraus ergeben sich mit&nbsp; $A=1$&nbsp; die Zahlenwerte:
+
*From this, the following numerical values arise a result with&nbsp; $A=1$&nbsp;:
 
:$$A_1 \approx 1 -0.125 +0.005 = 0.880,\hspace{0.3cm}
 
:$$A_1 \approx 1 -0.125 +0.005 = 0.880,\hspace{0.3cm}
 
A_3 \approx  -0.042 +0.003 = -0.039,\hspace{0.3cm}
 
A_3 \approx  -0.042 +0.003 = -0.039,\hspace{0.3cm}
Line 119: Line 119:
  
  
'''(4)'''&nbsp; Richtig sind die <u>Lösungsvorschläge 1 und 3</u>:
+
'''(4)'''&nbsp; <u>Approaches 1 and 3</u>&nbsp; are correct:
*Der Ansatz&nbsp; $g_5(x)$&nbsp; ist im gesamten Bereich eine bessere Näherung für die Sinusfunktion&nbsp; $g(x)$&nbsp; als die Näherung&nbsp; $g_3(x)$.  
+
*The approach&nbsp; $g_5(x)$&nbsp; is a better approximation for the sine function&nbsp; $g(x)$&nbsp; than the approximation&nbsp; $g_3(x)$ in the entire domain.  
*Deshalb ist der in der Teilaufgabe&nbsp; '''(3)'''&nbsp; berechnete Wert&nbsp; $K_{g5}$&nbsp; eine bessere Näherung für den tatsächlichen Klirrfaktor als&nbsp; $K_{g3}$. <br>die erste Aussage ist somit richtig.
+
*Thus, the value&nbsp; $K_{g5}$&nbsp; computed in the subtask&nbsp; '''(3)'''&nbsp; is a better approximation for the actual distortion factor than&nbsp; $K_{g3}$. <br>Therefore, the first statement is correct.
*Die zweite Aussage ist falsch, wie schon die Berechnung für&nbsp; $A=1$&nbsp; gezeigt hat: &nbsp; $K_{g3} \approx 4.76 \%$&nbsp; ist größer als&nbsp; $K_{g5} \approx 4.45 \%$.  
+
*The second statement is false as already shown by the computation for&nbsp; $A=1$&nbsp;: &nbsp; $K_{g3} \approx 4.76 \%$&nbsp; is greater than&nbsp; $K_{g5} \approx 4.45 \%$.  
*Der Grund hierfür ist, dass $g_3(x)$ unterhalb von $g_5(x)$ liegt und damit auch eine größere Abweichung vom linearen Verlauf vorliegt.
+
*The reason for this is that&nbsp; $g_3(x)$&nbsp; is below&nbsp; $g_5(x)$&nbsp; and thus there is also a greater deviation from the linear curve.
*Für&nbsp; $A=0.5$&nbsp; wird&nbsp; $K_{g5} \approx K_{g3} = 1.08 \%$&nbsp; gelten.  
+
*For&nbsp; $A=0.5$&nbsp;,&nbsp; $K_{g5} \approx K_{g3} = 1.08 \%$&nbsp; will hold.  
*Die Kennlinie auf der Angabenseite zeigt, dass  für&nbsp; $|x| \le 0.5$&nbsp; die Funktionen&nbsp; $g_3(x)$&nbsp; und&nbsp; $g_5(x)$&nbsp; innerhalb der Zeichengenauigkeit nicht zu unterscheiden sind.  
+
*The characteristic curve on the information page shows that for&nbsp; $|x| \le 0.5$&nbsp; the functions&nbsp; $g_3(x)$&nbsp; and&nbsp; $g_5(x)$&nbsp; are indistinguishable within the accuracy of drawing.  
*Damit ergeben sich auch gleiche Klirrfaktoren.  
+
*This also results in&nbsp; (nearly)&nbsp; the same distortion factors.  
 
{{ML-Fuß}}
 
{{ML-Fuß}}
  
  
  
[[Category:Linear and Time-Invariant Systems: Exercises|^2.2 Nichtlineare Verzerrungen^]]
+
[[Category:Linear and Time-Invariant Systems: Exercises|^2.2 Nonlinear Distortions^]]

Latest revision as of 15:12, 29 September 2021

Sinusoidal characteristic curve

We consider a system with input  $x(t)$  and output  $y(t)$.  For simplicity of description, the signals are considered to be dimensionless.

The relationship between the input signal  $x(t)$  and the output signal  $y(t)$  is given by the following characteristic curve in the range between  $-\pi/2$  and  $+\pi/2$:

$$g(x) = \sin(x) = x - \frac{x^3}{3!} + \frac{x^5}{5!} - \hspace{0.05cm}\text{...}$$

The second part of this equation describes the series expansion of the sine function.

As approximations for the nonlinear characteristic curve the following is used in this task:

$$g_1(x) = x\hspace{0.05cm},$$
$$ g_3(x) = x- x^{3}\hspace{-0.1cm}/6\hspace{0.05cm},$$
$$g_5(x) = x- x^3\hspace{-0.1cm}/{6}+x^5\hspace{-0.1cm}/{120}\hspace{0.05cm}.$$
  • The input signal  $x(t) = A \cdot \cos(\omega_0 \cdot t)$  is always assumed. 
  • The values  $A = 0.5$,  $A = 1.0$  and  $A = 1.5$  are to be considered for the (dimensionless) signal amplitude.





Please note:

  • The task belongs to the chapter  Nonlinear Distortions.
  • The resulting signal curves for  $x(t)$  and  $y(t)$  are shown graphically on the page  Description of nonlinear systems .
  • All powers required here refer to the resistance  $R = 1 \ \rm \Omega$  and thus have the unit  ${\rm V}^2$.
  • The following trigonometric relations are assumed to be known:
$$\cos^3(\alpha) = {3}/{4} \cdot \cos(\alpha) + {1}/{4} \cdot \cos(3\alpha) \hspace{0.05cm}, $$
$$ \cos^5(\alpha) = {10}/{16} \cdot \cos(\alpha) + {5}/{16} \cdot \cos(3\alpha) + {1}/{16} \cdot \cos(5\alpha)\hspace{0.05cm}.$$


Questions

1

What distortion factor  $K$  is obtained with the approximation  $\underline{g_1(x)}$  of the characteristic curve independent of the amplitude  $A$  of the input signal?

$K \ = \ $

$\ \%$

2

Compute the distortion factor  $K$  for the input signal  $x(t) = A \cdot \cos(\omega_0 \cdot t)$  and the approximation  $\underline{g_3(x)}$.
What values arise as a result for  $A = 0.5$  and  $A = 1.0$?

$A = 0.5\hspace{-0.08cm}:\ \ K \ = \ $

$\ \%$
$A = 1.0\hspace{-0.08cm}:\ \ K \ = \ $

$\ \%$

3

What is the distortion factor for  $\underline{A = 1.0}$  considering the approximation  $\underline{g_5(x)}$?

$K \ = \ $

$\ \%$

4

Which of the following statements are true?  Here,  $K$  denotes the distortion factor of the sine function  $g(x)$.
$K_{\rm g3}$  and  $K_{\rm g5}$  are based on the approximations  $g_3(x)$  and  $g_5(x)$, respectively.

$K_{\rm g5}$  generally represents a better approximation for  $K$  than  $K_{\rm g3}$.
$K_{\rm g3} < K_{\rm g5}$ holds for  $A = 1.0$.
$K_{\rm g3} \approx K_{\rm g5}$ will hold for  $A = 0.5$.


Solution

(1)  The very inaccurate approximation  $g_1(x) = x$  is linear in  $x$  and therefore does not result in nonlinear distortions. Hence, the distortion factor is $\underline{K = 0}$.


(2)  The analytical spectrum  (positive frequencies only)  of the input signal is:

$$X_+(f) = A \cdot {\rm \delta}(f- f_0) .$$
  • Then, the following signal is applied to the output of the nonlinear characteristic curve  $g_3(x)$ :
$$y(t) = A \cdot {\rm cos}(\omega_0 t ) - \frac{A^3}{6} \cdot {\rm cos}^3(\omega_0 t )= A \cdot {\rm cos}(\omega_0 t ) - \frac{3}{4} \cdot \frac{A^3}{6} \cdot {\rm cos}(\omega_0 t )- \frac{1}{4} \cdot \frac{A^3}{6} \cdot {\rm cos}(3\omega_0 t ) = A_1 \cdot {\rm cos}(\omega_0 t ) + A_3 \cdot {\rm cos}(3\omega_0 t ).$$
  • For the coefficients  $A_1$  and $A_3$ the following is obtained by comparison of coefficients:
$$A_1 = A - {A^3}\hspace{-0.1cm}/{8}, \hspace{0.5cm}A_3 = - {A^3}\hspace{-0.1cm}/{24}.$$
  • Using  $A = 0.5$  the following is obtained:  $A_1 \approx 0.484$  and  $A_3 \approx 0.005$.  Thus, the distortion factor is:
$$K = K_3 ={|A_3|}/{A_1}= {0.005}/{0.484} \hspace{0.15cm}\underline{ = 1.08\%}.$$
Note that for the approximation  $g_3(x)$  only the cubic part  $K_3$  of the distortion factor is effective.
  • For  $A = 1.0$  and  $A = 1.5$  the following numerical values:
$$A = 1.0: A_1 \approx 0.875, \hspace{0.2cm} A_3 \approx -0.041\hspace{0.3cm} \Rightarrow \hspace{0.3cm} \hspace{0.15cm}\underline{K \approx 4.76\%}\; \; \Rightarrow \; \; K_{g3},$$
$$A = 1.5: A_1 \approx 1.078, \hspace{0.2cm} A_3 \approx -0.140\hspace{0.3cm} \Rightarrow \hspace{0.3cm} \hspace{0.15cm}{K \approx 13 \%}.$$


(3)  Similarly as in subtask  (2)

$$y(t) = A_1 \cdot {\rm cos}(\omega_0 t ) + A_3 \cdot {\rm cos}(3\omega_0 t )+ A_5 \cdot {\rm cos}(5\omega_0 t )$$
holds with the following coefficients:
$$A_1 = A - {A^3}\hspace{-0.1cm}/{8} + {A^5}\hspace{-0.1cm}/{192},\hspace{0.3cm} A_3 = - {A^3}\hspace{-0.1cm}/{24} + {A^5}\hspace{-0.1cm}/{384},\hspace{0.3cm} A_5 = {A^5}\hspace{-0.1cm}/{1920}.$$
  • From this, the following numerical values arise a result with  $A=1$ :
$$A_1 \approx 1 -0.125 +0.005 = 0.880,\hspace{0.3cm} A_3 \approx -0.042 +0.003 = -0.039,\hspace{0.3cm} A_5 \approx 0.0005$$
$$\Rightarrow \hspace{0.3cm}K_3 = {|A_3|}/{A_1}= 0.0443,\hspace{0.3cm}K_5 = {|A_5|}/{A_1}= 0.0006 \hspace{0.3cm} \Rightarrow \hspace{0.3cm} K = \sqrt{K_3^2 + K_5^2} \hspace{0.15cm}\underline{\approx 4.45\%} \; \; \Rightarrow \; \; K_{g5}.$$


(4)  Approaches 1 and 3  are correct:

  • The approach  $g_5(x)$  is a better approximation for the sine function  $g(x)$  than the approximation  $g_3(x)$ in the entire domain.
  • Thus, the value  $K_{g5}$  computed in the subtask  (3)  is a better approximation for the actual distortion factor than  $K_{g3}$.
    Therefore, the first statement is correct.
  • The second statement is false as already shown by the computation for  $A=1$ :   $K_{g3} \approx 4.76 \%$  is greater than  $K_{g5} \approx 4.45 \%$.
  • The reason for this is that  $g_3(x)$  is below  $g_5(x)$  and thus there is also a greater deviation from the linear curve.
  • For  $A=0.5$ ,  $K_{g5} \approx K_{g3} = 1.08 \%$  will hold.
  • The characteristic curve on the information page shows that for  $|x| \le 0.5$  the functions  $g_3(x)$  and  $g_5(x)$  are indistinguishable within the accuracy of drawing.
  • This also results in  (nearly)  the same distortion factors.