1
\$\begingroup\$

I have the simple transfer function of an RC filter:

\$ H(s) = \frac{sRC}{1 + sRC} \$

In order to find the magnitude, square the previous equation and take the square root:

\$ (H(s))^2 = \frac{(sRC)^2}{(sRC + 1)^2} ; s = j\omega \$

\$ (H(s))^2 = \frac{(j\omega R C)^2}{(j\omega R C + 1)^2} \qquad(1) \$

This is where I get stuck. I know that the correct magnitude should be of the form:

\$ |H(s)| = \sqrt{\frac{(\omega R C)^2}{(\omega R C)^2 + 1}} \qquad (2) \$

However, when I actually try to calculate this out on paper I don't know how to deal with the \$j\$. I don't understand how to get from equation I get from equation (1) to equation (2). The \$-1\$ from \$j^2\$ causes the form to look like the following:

\$ (H(s))^2 = \frac{-top}{1 - bottom} \$

If I were to take the square root of this I do not get equation (2).

What am I doing wrong?

\$\endgroup\$
  • \$\begingroup\$ You didn't take the square of the denominator properly: it should be (1+sRC)^2. \$\endgroup\$ – Barry Nov 19 '17 at 23:30
  • \$\begingroup\$ You're correct, I did not do that correctly. However, doing it again I get that: H(s)^2 = [(j^2)(w^2)(R^2)(C^2)] / [1 + 2jwRC + (j^2)(w^2)(R^2)(C^2)] ; This leaves me.. more confused than before? \$\endgroup\$ – Zearia Nov 19 '17 at 23:43
1
\$\begingroup\$

Given what you already developed:

$$\begin{align*} H\left(s\right)\quad &=\quad \frac{s R C}{1 + s R C}\\\\ \mid\: H\left(s\right)\: \mid^{\:2}\quad &=\quad H\left(s\right)\cdot H\left(s^*\right)\\\\ &=\quad\frac{s R C}{1 + s R C}\cdot\frac{s^* R C}{1 + s^* R C}\\\\ &=\quad\frac{\left(\sigma^2+\omega^2\right) R^2 C^2}{1 + 2\sigma R C+\sigma^2 R^2 C^2+\omega^2 R^2 C^2} \end{align*}$$

Laplace uses \$s=\sigma + j \omega\$. But this collapses to Fourier along the imaginary axis portion, or the \$j \omega\$ part of that. Since you want the periodic and non-decaying part of the system response -- the frequency response -- you set \$\sigma=0\$ in \$s\$.

So the above devolves to:

$$\begin{align*} \mid\:H\left(j\omega\right)\:\mid\quad &=\sqrt{ \frac{\omega^2 R^2 C^2}{1 + \omega^2 R^2 C^2}}\\\\ &=\quad\frac{\omega R C}{\sqrt{1+\omega^2 R^2 C^2}} \end{align*}$$

I hope that helps.

\$\endgroup\$
  • \$\begingroup\$ What I seem to be confusing about this whole process is that when you square the s, would you not get s^2 = (sigma^2) + [(j^2)(w^2)]? If sigma was set to 0 would that not still yield that s^2 = -(w^2)? \$\endgroup\$ – Zearia Nov 20 '17 at 1:27
  • 1
    \$\begingroup\$ @Zearia By definition, the magnitude squared of a complex number is \$\mathscr{C}\cdot\mathscr{C^*}\$. See: Absolute value and argument. And no, we are talking about \$s=\sigma+j \omega\$. The magnitude squared is \$\mid s\mid ^2=\left( \sigma+j \omega\right )\cdot \left( \sigma-j \omega\right)\$. \$\endgroup\$ – jonk Nov 20 '17 at 2:03
  • 1
    \$\begingroup\$ @Zearia If you don't see why, just plot a point on the complex plane and work out how you'd compute the hypotenuse. It should be very clear, then, I think. \$\endgroup\$ – jonk Nov 20 '17 at 2:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.