0

I am trying to evaluate the trigonometric Fourier series expansions of three quasi-square waves, which are essentially the same quasi-square waveform, phase-shifted with respect to each other by multiples of $\frac{2\pi}{3}$. The three waveforms are as shown:

Phase-shifted quasi-square waveforms

$$ S_a(x) = \begin{cases} 0 & : 0 ≤ x < \frac{\pi}{6} \\ 1 & : \frac{\pi}{6} ≤ x < \frac{5\pi}{6} \\ 0 & : \frac{5\pi}{6} ≤ x < \frac{7\pi}{6} \\ -1 & : \frac{7\pi}{6} ≤ x < \frac{11\pi}{6} \\ 0 & : \frac{11\pi}{6} ≤ x < {2\pi} \\ S_a((x \mod 2\pi)) & : \text{otherwise} \end{cases} $$

$$S_b(x)=S_a(x-\frac{2\pi}{3})$$ $$S_c(x)=S_a(x-\frac{4\pi}{3})$$

I have evaluated the Fourier series of the first waveform (in red) and the Fourier coefficients are as follows:

$$a_{0}^r = 0$$ $$a_{n}^r = 0$$ $$b_{n}^r = \begin{cases} \frac{4}{n\pi}\cos(\frac{n\pi}{6}) & : n\ \text{odd except triplen} \\ 0 & : \text{otherwise} \end{cases} $$

I am trying to approximate the Fourier coefficients for the other two waveforms from these coefficients. I know that for a complex Fourier series expansion, the time-shifting property of the Fourier series gives:

$$\operatorname{FourierSeriesCoeff}(f(x-x_0))=C_n'=C_{n}e^{-jnx_0}$$ where, $\operatorname{FourierSeriesCoeff}(f(x))=C_{n}$

I also know the relationship between the complex $C_n$ and trigonometric Fourier coefficients $a_n$ and $b_n$ as: $$C_{n}=\frac{1}{2}(a_n-jb_n)$$ $$C_{-n}=C_{n}^*=\frac{1}{2}(a_n+jb_n)$$

Combining the expressions above gives the following:

$$C_n'=\frac{1}{2}(a_n'-jb_n')=\frac{1}{2}(a_n-jb_n)e^{-jnx_0}$$ $$C_{-n}'=\frac{1}{2}(a_n'+jb_n')=\frac{1}{2}(a_n+jb_n)e^{-jnx_0}$$

Using Euler's formula $e^{-jnp}=cos(np)-jsin(np)$ and solving for $a_n'$ and $b_n'$, I get:

$$a_n'= a_n\cos(nx_0)-b_n\sin(nx_0)$$ $$b_n'= a_n\sin(nx_0)+b_n\cos(nx_0)$$

Using these for my initial problem, I should get for the blue waveform:

$$a_n^b= a_n^r\cos(\frac{n2\pi}{3})-b_n^r\sin(\frac{n2\pi}{3})$$ $$b_n^b= a_n^r\sin(\frac{n2\pi}{3})+b_n^r\cos(\frac{n2\pi}{3})$$

which resolves to:

$$a_n^b= \begin{cases} -b_n^r\sin(\frac{2n\pi}{3}) & : n\ \text{odd except triplen} \\ 0 & : \text{otherwise} \end{cases} $$

and

$$b_n^b= \begin{cases} b_n^r\cos(\frac{2n\pi}{3}) & : n\ \text{odd except triplen} \\ 0 & : \text{otherwise} \end{cases} $$

However, after trying to evaluate the trigonometric coefficients for the blue waveform manually, I have the following expression for $a_n^b$:

$$a_n^b=\frac{1}{n\pi}[-3sin(\frac{2n\pi}{3})cos(\frac{n\pi}{6})+cos(\frac{2n\pi}{3})sin(\frac{n\pi}{6})-sin(\frac{n\pi}{6})]$$

What am I doing wrong? I am attaching my attempt at a solution here: My Attempt

Edit 1: I was unable to simplify the expression for the manual calculation. However, after verifying with Mathematica, the coefficients follow the time-shifting property correctly, which is what I was trying to verify. The Fourier coefficients for the the three waveforms end up being:

$$a_{n}^r = 0$$ $$b_{n}^r = \begin{cases} \frac{4}{n\pi}\cos(\frac{n\pi}{6}) & : n\ \text{odd except triplen} \\ 0 & : \text{otherwise} \end{cases} $$

$$a_{n}^b= \begin{cases} -b_n^r\sin(\frac{2n\pi}{3}) & : n\ \text{odd except triplen} \\ 0 & : \text{otherwise} \end{cases} $$ $$b_n^b= \begin{cases} b_n^r\cos(\frac{2n\pi}{3}) & : n\ \text{odd except triplen} \\ 0 & : \text{otherwise} \end{cases} $$

$$a_{n}^o = -a_{n}^b= \begin{cases} b_n^r\sin(\frac{2n\pi}{3}) & : n\ \text{odd except triplen} \\ 0 & : \text{otherwise} \end{cases} $$ $$b_n^o= b_{n}^b = \begin{cases} b_n^r\cos(\frac{2n\pi}{3}) & : n\ \text{odd except triplen} \\ 0 & : \text{otherwise} \end{cases} $$

Here's the result from Mathematica

MNairA
  • 1
  • 2
  • A more detailed attempt: https://drive.google.com/file/d/1oP5GQmYV4XkdbUC-u4h9FJBhRrfzrG6I/view?usp=sharing – MNairA Feb 27 '24 at 19:40
  • Mathematica indicates your last formula is equivalent to $$a_n^b=-\frac{3 \sin \left(\frac{\pi n}{3}\right)}{\pi n}$$ using the assumption $n\in\mathbb{Z}$. – Steven Clark Feb 27 '24 at 21:07
  • Thank you for responding Steven. I have arrived at the same result myself through hand calc, but it doesn't seem to match the value from the time-shifting property. I have edited the question to reflect both values now. – MNairA Feb 27 '24 at 22:03
  • Could you please clarify if $$S_a(x)=\left{\begin{array}{cc} 0 & 0\leq x<\frac{\pi }{3} \ 1 & \frac{\pi }{3}\leq x<\frac{2 \pi }{3} \ 0 & \frac{2 \pi }{3}\leq x<\frac{4 \pi }{3} \ -1 & \frac{4 \pi }{3}\leq x<\frac{5 \pi }{3} \ 0 & \frac{5 \pi }{3}\leq x<\frac{6 \pi }{3} \ S_a((x \bmod (2 \pi ))) & \text{otherwise} \ \end{array}\right.$$ is the correct definition of $S_a(x)$, and the value of the constants $x_b$ and $x_c$ in $S_b(x)=S_a(x-x_b)$ and $S_c(x)=S_a(x-x_c)$? – Steven Clark Feb 28 '24 at 00:12
  • Hello Steven, I am terribly sorry for wasting your time. It seems I should have provided an equation for the waveform at the start. I have updated the problem statement to reflect the correct equation for the waveform. I am redoing the solution since I seem to have written my last solution using the wrong angle values. – MNairA Feb 28 '24 at 01:41
  • You can get properly sized parentheses (and other paired delimiters) that adjust to the size of their content by preceding them with \left and \right. You can get the proper font and spacing for $\cos$ and $\sin$ using \cos and \sin. For operators that don't have a command of their own, you can use \operatorname{name}. – joriki Feb 28 '24 at 02:28

0 Answers0