Laplace transform and region of convergence
Defines the Laplace transform and its region of convergence, and uses the ROC, poles, partial fractions and the initial- and final-value theorems to analyse causality, stability and time response.
Drafted with Aria, reviewed by the AiCanCode.org team. Spotted an error? Use Give Feedback at the bottom of the page.
Why it matters
Sensors, signal-conditioning circuits and control loops are described by differential equations, and the Laplace transform turns those into algebra. It also handles signals and systems the Fourier transform cannot (growing exponentials, unstable plants), and the region of convergence (ROC) tells you at a glance whether a system is causal and stable.
Key ideas
Definition. The bilateral Laplace transform weights the signal by e^(−st) with complex s = σ + jω. The real part σ adds an exponential damping e^(−σt), which makes the integral converge for many signals whose Fourier transform does not exist. On the line σ = 0, X(s) reduces to the CTFT X(jω) — provided that line is inside the ROC.
Region of convergence. The ROC is the set of s for which the integral converges. It depends only on σ = Re(s), so it is always a vertical strip, a half-plane or the whole plane, and it never contains a pole. The same expression X(s) can belong to different signals; the ROC is part of the answer.
- Right-sided signal (zero before some time, e.g. causal): ROC is to the right of the rightmost pole,
Re(s) > σmax. - Left-sided signal (zero after some time): ROC is to the left of the leftmost pole.
- Two-sided signal: ROC is a strip between two poles (or does not exist).
- Finite-duration, absolutely integrable signal: ROC is the entire s-plane.
Causality and stability from the ROC (for an LTI system with transfer function H(s))
- Causal ⇔ ROC is a right half-plane (for rational
H(s)). - BIBO stable ⇔ ROC includes the jω-axis.
- Causal and stable ⇔ all poles have negative real parts (ROC is right of all poles and includes the jω-axis).
- A system can be stable but non-causal: poles at
−1and+2with the strip−1 < Re(s) < 2.
Unilateral transform. For causal signals and initial-value problems, the one-sided transform (lower limit 0⁻) is used. Its differentiation property brings in initial conditions automatically, which is why it is the tool for circuits and control systems.
Inverse transform. For rational X(s), use partial fractions, then match each term with a standard pair, choosing right- or left-sided according to the ROC. Repeated poles give terms like t·e^(−at); complex poles give damped sinusoids.
Poles and time behaviour (causal signals): a pole at s = −a gives e^(−at); a pair at −σ ± jωd gives e^(−σt) cos(ωd t + φ); poles on the jω-axis give undamped oscillation or a step; right-half-plane poles give growth.
Formulas
X(s) = ∫ from −∞ to ∞ of x(t) e^(−st) dt— bilateral transform;s = σ + jω.X(s) = ∫ from 0⁻ to ∞ of x(t) e^(−st) dt— unilateral transform.e^(−at)u(t) ↔ 1/(s + a), ROCRe(s) > −a.−e^(−at)u(−t) ↔ 1/(s + a), ROCRe(s) < −a.δ(t) ↔ 1(all s);u(t) ↔ 1/s, ROCRe(s) > 0;t·u(t) ↔ 1/s².tⁿ e^(−at) u(t) ↔ n!/(s + a)^(n+1).e^(−at) cos(ω0t) u(t) ↔ (s + a)/((s + a)² + ω0²),e^(−at) sin(ω0t) u(t) ↔ ω0/((s + a)² + ω0²).x(t − t0) ↔ e^(−st0) X(s);e^(−at) x(t) ↔ X(s + a);x * h ↔ X(s)H(s).dx/dt ↔ sX(s) − x(0⁻)(unilateral);∫ from 0⁻ to t of x ↔ X(s)/s.x(0⁺) = lim(s→∞) sX(s)— initial value theorem (X(s) strictly proper).x(∞) = lim(s→0) sX(s)— final value theorem, valid only if all poles ofsX(s)are in the left half-plane.
Symbols: s complex frequency (s⁻¹); σ neper frequency (s⁻¹); ω angular frequency (rad/s); a decay rate (s⁻¹); t0 delay (s); X(s) has the unit of x·s.
Worked examples
Example 1 (standard). Find the causal x(t) whose transform is X(s) = (s + 3)/(s² + 3s + 2), and check it with the initial- and final-value theorems.
- Factor:
s² + 3s + 2 = (s + 1)(s + 2). WriteX(s) = A/(s + 1) + B/(s + 2). A = (s + 3)/(s + 2)ats = −1=2/1 = 2.B = (s + 3)/(s + 1)ats = −2=1/(−1) = −1.- Causal ⇒ ROC
Re(s) > −1:x(t) = (2e^(−t) − e^(−2t)) u(t). - Initial value:
lim(s→∞) s(s + 3)/(s² + 3s + 2) = 1, andx(0⁺) = 2 − 1 = 1✓. - Final value: poles of
sX(s)are at −1, −2 (left half-plane), sox(∞) = lim(s→0) sX(s) = 0✓. - x(t) = (2e^(−t) − e^(−2t)) u(t), with
x(0⁺) = 1andx(∞) = 0.
Example 2 (GATE level). H(s) = 1/((s + 1)(s − 2)). List the possible ROCs, say which is stable, and find h(t) for the stable one.
- Poles at
s = −1ands = +2. Possible ROCs:Re(s) > 2(causal, unstable),Re(s) < −1(anti-causal, unstable),−1 < Re(s) < 2(two-sided). - Only the strip
−1 < Re(s) < 2contains the jω-axis ⇒ stable but non-causal. - Partial fractions:
H(s) = (−1/3)/(s + 1) + (1/3)/(s − 2)(check: ats = −1,1/(−1 − 2) = −1/3; ats = 2,1/(2 + 1) = 1/3). - Pole at −1 is to the left of the strip ⇒ right-sided term:
(−1/3) e^(−t) u(t). - Pole at +2 is to the right of the strip ⇒ left-sided term:
(1/3) × [−e^(2t) u(−t)]. - h(t) = −(1/3) e^(−t) u(t) − (1/3) e^(2t) u(−t) — both terms decay away from t = 0, so
∫|h| = 1/3 + 1/6 = 0.5is finite.
Common mistakes
- Giving
X(s)without its ROC, or assuming every ROC isRe(s) > something. - Using the final value theorem on a signal with poles on the jω-axis or in the right half-plane (e.g.
x(t) = sin t), which gives a meaningless number. - Forgetting the minus sign in the left-sided pair:
−e^(−at)u(−t) ↔ 1/(s + a). - Applying the initial value theorem to an improper
X(s)(degree of numerator ≥ denominator) — split off the impulse part first. - Equating "poles in the left half-plane" with stability for a non-causal system; the real test is whether the ROC contains the jω-axis.
- Dropping the initial condition in
sX(s) − x(0⁻)when solving a differential equation.
For GATE IN
Questions ask you to pick the correct ROC for a given signal, decide causality and stability from pole locations and an ROC, find an inverse transform by partial fractions, or use the initial- and final-value theorems (including recognising when the final value theorem does not apply). Practise two-sided signals and the three-ROC case for two real poles.
Quick check
- What is the ROC of
x(t) = e^(−2t)u(t) + e^(−5t)u(t)? - What is the Laplace transform of
e^(3t)u(−t), with ROC? X(s) = 10/(s(s + 5))— final value ofx(t)?- Can the ROC of a finite-duration pulse contain a pole?
- A causal system has a pole at
s = 0.5. Is it stable? Answers: 1.Re(s) > −2. 2.−1/(s − 3),Re(s) < 3. 3.lim s·10/(s(s + 5)) = 2. 4. No — its ROC is the whole s-plane, and its transform has no finite poles. 5. No — causal with a right-half-plane pole is unstable.
Interview questions
All Signals and Systems interview questionsTry answering each one aloud before you open it.
1.What is the Laplace transform and why is it used in signals and systems?Concept
The Laplace transform is a mathematical technique used to transform a time-domain function into a complex frequency-domain representation. It is used in signals and systems to simplify the analysis of linear time-invariant systems, especially for solving differential equations and analyzing system stability.
2.Explain the concept of the region of convergence (ROC) in the context of the Laplace transform.Concept
The region of convergence (ROC) is the range of complex frequencies for which the Laplace transform of a signal converges to a finite value. It is crucial for determining the stability and causality of a system. The ROC depends on the poles of the system's transfer function.
3.How does the Laplace transform differ from the Fourier transform?Concept
The Laplace transform is a generalization of the Fourier transform. While the Fourier transform is used for analyzing signals in the frequency domain, the Laplace transform includes a real part (σ) in addition to the imaginary part (jω), allowing it to handle a broader class of signals, including those that grow exponentially.
4.Why is the Laplace transform preferred over the Fourier transform for analyzing control systems?Application
The Laplace transform is preferred for control systems because it can handle initial conditions and transient states, which are common in control systems. It also provides a more comprehensive analysis of system stability and dynamics through its complex frequency domain representation.
5.What happens if the ROC does not include the imaginary axis in the Laplace transform?Application
Then the LTI system is not BIBO stable, because the ROC contains the jω-axis exactly when h(t) is absolutely integrable. It also means the Fourier transform of h(t) does not exist as an ordinary function, so the system has no meaningful frequency response: H(jω) cannot be obtained by putting s = jω. A causal system with a pole in the right half-plane is the usual example.
6.Explain how the poles of a transfer function relate to the ROC in the Laplace transform.Concept
The poles of a transfer function are the values of s (complex frequency) that make the denominator of the transfer function zero. The ROC is determined by the location of these poles and is typically a region in the s-plane that does not include any poles. The ROC must be a connected region that includes the imaginary axis for the system to be stable.
7.What is the significance of the ROC being a right half-plane in the Laplace transform?Application
A right-half-plane ROC, Re(s) > σmax, means the signal is right-sided; for a rational transfer function it means the system is causal. It does not by itself mean stability: the system is also BIBO stable only if that half-plane includes the jω-axis, i.e. σmax < 0, so all poles have negative real parts. For example, 1/(s − 2) with ROC Re(s) > 2 is causal but unstable.
8.Calculate the Laplace transform of the function f(t) = e^(-2t)u(t), where u(t) is the unit step function.Numerical
The Laplace transform of f(t) = e^(-2t)u(t) is F(s) = 1 / (s + 2), with the ROC being Re(s) > -2.
9.Determine the Laplace transform and ROC of the signal x(t) = e^(3t)u(−t).Numerical
X(s) = ∫ from −∞ to 0 of e^(3t)e^(−st) dt = [e^((3−s)t)/(3 − s)] from −∞ to 0 = 1/(3 − s) = −1/(s − 3). The integral converges only when Re(s) < 3, because the lower limit needs e^((3−σ)t) → 0 as t → −∞. So X(s) = −1/(s − 3) with ROC Re(s) < 3 — a left half-plane, as expected for a left-sided signal; note the minus sign compared with the causal pair e^(3t)u(t) ↔ 1/(s − 3).
10.Why is it important to consider the ROC when using the Laplace transform in system analysis?Application
Considering the ROC is important because it determines the conditions under which the Laplace transform converges. It affects the stability and causality of the system, and without a proper ROC, the analysis may lead to incorrect conclusions about the system's behavior.
Finished this topic? Mark it so your progress, study plan and readiness keep up.
Stuck on something here?