A transfer function is the Laplace-domain rule that connects the input of a linear time-invariant system to its output. With zero initial conditions, it is defined as

H(s)=Y(s)X(s)H(s) = \frac{Y(s)}{X(s)}

where X(s)X(s) is the transformed input and Y(s)Y(s) is the transformed output. In plain language, it tells you how strongly the system responds to different inputs without solving the full differential equation from scratch every time.

This is not just "output divided by input" in any situation. The definition only works under specific conditions, and those conditions matter.

What A Transfer Function Tells You

The transfer function packages a system's behavior into one expression. Once you know H(s)H(s), you can often read off whether the system amplifies, reduces, delays, or filters parts of the input.

For sinusoidal steady-state questions, you evaluate it on the imaginary axis as H(iω)H(i\omega). That gives two practical pieces of information:

  • the magnitude, which tells you how much a sinusoidal input at angular frequency ω\omega is amplified or reduced
  • the phase, which tells you how much the output is shifted relative to the input

That is why transfer functions show up in circuits, vibrations, filtering, and control.

When H(s)=Y(s)/X(s)H(s) = Y(s)/X(s) Is Valid

The usual formula assumes the system is linear and time-invariant. If linearity fails, inputs do not combine in the usual superposition way. If time invariance fails, the system can behave differently at different times, so one fixed transfer function is not enough.

Zero initial conditions matter too. Stored energy in a capacitor, inductor, or mechanical oscillator changes the actual output, but that extra contribution is not part of the transfer function itself. The transfer function describes the system's built-in input-output rule under the standard zero-initial-condition setup.

Worked Example: RC Low-Pass Filter

Take a resistor RR in series with a capacitor CC, and measure the output across the capacitor. In the Laplace domain, the capacitor impedance is 1/(sC)1/(sC), so the voltage-divider rule gives

H(s)=Vout(s)Vin(s)=1sCR+1sC=11+sRCH(s) = \frac{V_{out}(s)}{V_{in}(s)} = \frac{\frac{1}{sC}}{R + \frac{1}{sC}} = \frac{1}{1 + sRC}

This is a low-pass transfer function. Low frequencies pass more easily than high frequencies, which is why the output looks like a smoothed version of the input.

Choose one concrete case:

R=1000 Ω,C=1 μFR = 1000\ \Omega, \qquad C = 1\ \mu\mathrm{F}

Then

RC=103 sRC = 10^{-3}\ \mathrm{s}

so the transfer function becomes

H(s)=11+0.001sH(s) = \frac{1}{1 + 0.001s}

The cutoff angular frequency is

ωc=1RC=1000 rad/s\omega_c = \frac{1}{RC} = 1000\ \mathrm{rad/s}

which corresponds to

fc=ωc2π159 Hzf_c = \frac{\omega_c}{2\pi} \approx 159\ \mathrm{Hz}

At the cutoff,

H(iωc)=120.707\left|H(i\omega_c)\right| = \frac{1}{\sqrt{2}} \approx 0.707

So the output amplitude is about 70.7%70.7\% of the input amplitude at that frequency. That one number already tells you something useful: the circuit is starting to noticeably attenuate signals around 159 Hz159\ \mathrm{Hz} and above.

For a quick check of the intuition, if ω1000 rad/s\omega \ll 1000\ \mathrm{rad/s}, then H(iω)|H(i\omega)| is close to 11, so the output is almost the same size as the input. If ω1000 rad/s\omega \gg 1000\ \mathrm{rad/s}, the magnitude becomes small, so rapid oscillations are strongly reduced.

Common Transfer Function Mistakes

  • Using the term for systems that are not being modeled as linear and time-invariant.
  • Forgetting to define which variable is the input and which is the output.
  • Treating the transfer function as if it already includes arbitrary initial conditions.
  • Mixing up the general Laplace-domain transfer function H(s)H(s) with the frequency response H(iω)H(i\omega).
  • Reading only the magnitude and ignoring the phase shift when phase matters physically.

Where Transfer Functions Are Used

Transfer functions are useful whenever a system can be modeled by linear differential equations and you care about how inputs propagate to outputs. Common examples include RC and RLC circuits, damped mechanical oscillators, feedback systems, and simple sensor models.

In physics, they are especially useful when the main question is not the full time history, but how the system responds to driving, filtering, or oscillation across frequencies.

Try A Similar Transfer Function

Try the same RC circuit, but measure the output across the resistor instead of the capacitor. You will get a high-pass transfer function, and that comparison makes one key idea stick: changing the output changes the transfer function.

Need help with a problem?

Upload your question and get a verified, step-by-step solution in seconds.

Open GPAI Solver →