Notes โบ ENGR 2341: Signals and Systems Lecture 8
Fourier Series
352 words 3 min Modified
Table of Contents
Harmonic Sum of Sinusoids
- Harmonic frequencies are integer multiples of a fundamental frequency
- Signals that are a sum of sinusoids are periodic as multiples of the fundamental period
- $T_0 = \frac{2\pi}{\omega_0}= \frac{1}{f_0}$
- General form:
- The fundamental frequency is the greatest common denominator of the frequency components of the signal
- $f_0 = \gcd(f_1, f_2, \dots, f_N)$ (measured in Hz, without the $2\pi$)
- $\omega_0 = \gcd(\omega_1, \omega_2, \dots, \omega_N)$ (measured in rad/s, with the $2\pi$)
Fourier Series
-
Allows representation of non-sinusoids as harmonic sums of sinusoids
- Fourier analysis: finding spectrum from signal
- Fourier synthesis: finding signal from spectrum
-
What signals have a Fourier series representation?
- Must be periodic
- Must meet Dirichlet conditions
-
Fourier series representations:
Fourier Analysis
- To get the Fourier series representation of a signal:
- Determine $T_0$ and $\omega_0 = \frac{2\pi}{T_0}$
- Determine the coefficient of the DC (constant) term
- Determine the coefficients of the remaining harmonics
- For the Sum of complex exponentials Fourier series:
- $k$ is the harmonic number
- $k\omega_0$ is the frequency of the harmonic
- $a_k$ is the complex amplitude of the harmonic
- $(k\omega_0, a_k)$ pairs correspond to the two-sided spectrum
- Complex amplitude formula
- The DC amplitude lacks an exponential term inside the integrand
Fourier Synthesis
- To synthesize a signal from its spectrum, perform a running loop of each term in the Fourier series
- Should already have complex amplitude and frequency arrays
- $a_k$ should be halved if summing over complex exponentials
- Make sure to initialize the following:
- $t$ array over the proper time interval
- $x(t)$ array with all zeros
sumvariable for running sum of the Fourier series
Complex exponentials
- Loop over $k \in [-N, N]$
- Compute each harmonic $x_k(t) = a_k e^{j\omega_0 kt}$
- Add each computed harmonic to the running sum $x(t)$
Sum of cosines
- Loop over $k \in [0, N]$
- Compute each harmonic $x_k(t) = A_{k} \cos(\omega_{0}kt)$
- Add each computed harmonic to the running sum $x(t)$
References
- Fourier series (course handout)
Sources
- Fourier series