Laplace and Z-Transform Methods

Introduction

Transform methods convert differential and difference equations into algebraic equations. This conversion is especially valuable in electronics and communication engineering, where circuits, filters, control loops, transmission systems and sampled-data processors are commonly modeled as linear time-invariant (LTI) systems.

The Laplace transform is primarily used for continuous-time systems. It incorporates exponential behavior, initial conditions, transfer functions and continuous-time convolution. The Z-transform provides corresponding tools for discrete-time signals and systems. In both domains, poles, zeros and regions of convergence determine important properties such as causality, stability and transient behavior.

Transform methods do not eliminate the need for physical reasoning. A mathematically valid result must still be checked for correct initial conditions, units, stability and consistency with the underlying engineering system.

Learning Objectives

After studying this chapter, students should be able to:

• compute Laplace and Z-transforms and their inverses;
• apply transform properties, including time shifting and differentiation;
• solve differential equations with nonzero initial conditions;
• solve difference equations and separate zero-input and zero-state responses;
• evaluate continuous- and discrete-time convolution;
• construct and interpret transfer functions;
• identify poles, zeros, regions of convergence, causality and stability;
• apply initial- and final-value theorems under valid conditions; and
• relate continuous-time poles to sampled discrete-time poles.

1. Continuous-Time System Representation

A continuous-time LTI system may be described by a constant-coefficient differential equation:

Here, x(t) is the input, y(t) is the output, and a_k and b_m are constant coefficients. Their units must make all terms dimensionally consistent.

The response is conventionally separated as

where y_zi(t) is the zero-input response caused by stored initial energy and y_zs(t) is the zero-state response caused by the applied input when all initial conditions are zero.

This distinction is essential: a transfer function describes only the zero-state input-output relationship.

2. Laplace Transform

2.1 Definitions and region of convergence

The bilateral Laplace transform of x(t) is

where s has units of inverse seconds, σ controls exponential weighting and ω is angular frequency in rad/s. The set of values of s for which the integral converges is the region of convergence (ROC).

For causal engineering problems, the unilateral transform is particularly useful:

The lower limit 0^- permits impulses at the origin and preserves initial conditions immediately before switching.

For example,

The algebraic expression alone is not sufficient to identify a bilateral transform; its ROC is also required.

2.2 Common Laplace-transform pairs

The unit step u(t) is dimensionless, while δ(t) has units of inverse seconds so that its integral is dimensionless.

2.3 Important properties

For constants a and b, linearity gives

For a delay t₀>0,

Frequency translation gives

For unilateral transforms, derivatives retain initial conditions:

More generally,

These formulas assume the required derivatives exist in the ordinary or generalized-function sense and satisfy suitable exponential-order conditions.

3. Inverse Laplace Transform

The inverse transform reconstructs a time-domain signal from X(s). Common engineering methods include transform tables, completing the square, partial fractions and convolution.

For a rational function with distinct poles,

partial-fraction expansion gives

where, for a simple pole,

Repeated poles generate terms such as 1/(s−p)^m, whose causal inverse is

For bilateral transforms, the ROC determines whether the inverse is right-sided, left-sided or two-sided.

4. Solving Differential Equations with Nonzero Initial Conditions

Worked Example 1: Second-order system with stored initial energy

Consider

with initial conditions

Taking unilateral Laplace transforms,

Therefore,

and

The first term is the zero-input contribution:

so

For the zero-state contribution,

Thus,

The total response is

Verification gives y(0^+)=1 and y'(0^+)=−1, consistent with the prescribed initial conditions. The steady value is 1/2, which also follows by setting derivatives to zero in the original equation.

5. Convolution and Continuous-Time Transfer Functions

For an LTI system initially at rest, the output is the convolution of the input and impulse response:

For causal signals, the integral reduces to

because both signals vanish for negative time.

The convolution theorem states

Consequently, under zero initial conditions,

is the transfer function. Its physical units are output units divided by input units.

Worked Example 2: Convolution response

Let

Then

and hence

Transform-domain multiplication gives the same result because

6. Poles, Zeros and Continuous-Time Stability

For a rational transfer function,

z_i are zeros, p_k are poles and K is a gain constant. Zeros suppress particular modal or frequency components, whereas poles generate the system's natural modes.

For a causal rational continuous-time LTI system:

• BIBO stability requires every pole to lie strictly in the open left-half plane, Re(p_k)<0.
• A pole on the imaginary axis generally prevents BIBO stability.
• Right-half-plane poles produce exponentially growing modes.
• Repeated poles increase polynomial factors multiplying the exponential response.

Pole-zero cancellation should not automatically be assumed to guarantee internal stability. An unstable mode may remain hidden internally because component uncertainty prevents exact cancellation.

7. Initial- and Final-Value Theorems for Laplace Transforms

The initial-value theorem is

provided the limit exists and x(t) does not contain an impulse at the origin that invalidates the ordinary finite initial value.

The final-value theorem is

It is valid only when all poles of sX(s) lie strictly in the open left-half plane. Equivalently, X(s) may contain at most a simple pole at the origin while all other poles are strictly stable.

For the response in Worked Example 1,

which agrees with the time-domain final value. Applying the theorem to an undamped sinusoid would be invalid because the required pole condition is not satisfied.

8. The Z-Transform

8.1 Bilateral and unilateral definitions

The bilateral Z-transform is

where z is dimensionless. The ROC is the annular region in the complex z-plane where the sum converges.

The unilateral Z-transform is

It is particularly useful for causal difference equations with specified pre-initial values.

8.2 Common Z-transform pairs

The transform of δ[n] is the constant one and therefore has ROC over the entire z-plane. For other sequences, the ROC is an essential part of the transform specification.

8.3 Properties

Linearity gives

For the bilateral transform, a shift by n₀ samples gives

Time multiplication satisfies

Discrete convolution is

and the convolution theorem gives

9. Inverse Z-Transform

The inverse Z-transform may be obtained using transform pairs, partial fractions, power-series expansion or contour integration. The same rational expression can represent different sequences when the ROCs differ.

For example,

represents aⁿ u[n] when |z|>|a|, but represents −aⁿ u[−n−1] when |z|<|a|.

For a causal rational sequence, partial fractions are commonly written in terms of z/(z−a), because this form maps directly to aⁿ u[n]. Polynomial division can also reveal the sequence coefficients as powers of z⁻¹.

10. Difference Equations and Discrete-Time Transfer Functions

A general constant-coefficient difference equation is

Under initial rest, its transfer function is

For a causal rational system, BIBO stability requires every pole to be strictly inside the unit circle. This condition follows from the requirement that the ROC include the unit circle.

Worked Example 3: Correct separation of discrete responses

Consider

with y[−1]=2 and x[n]=δ[n]. Using unilateral transforms,

Therefore,

so

The zero-input response is obtained by setting x[n]=0:

Thus,

The zero-state response is obtained by setting y[−1]=0 while retaining the impulse input:

and hence

The total response is therefore

Direct substitution verifies y[0]=2 because y[0]−0.5y[−1]=2−1=1=δ[0]. The example demonstrates why the initial-condition term must be transformed carefully.

11. Initial- and Final-Value Theorems for Z-Transforms

For a right-sided sequence,

provided the limit exists.

The discrete final-value theorem is

It is valid when all poles of (1−z⁻¹)X(z) lie strictly inside the unit circle. Equivalently, X(z) may have a simple pole at z=1, while its other relevant poles must be strictly inside the unit circle.

For x[n]=aⁿ u[n] with |a|<1,

which agrees with lim_n→∞aⁿ=0. The theorem is not valid for a persistent oscillation such as (−1)ⁿ, because a pole lies on the unit circle at z=−1.

12. Relationship Between Laplace and Z Domains

If a continuous-time mode e^st is sampled every T seconds, then

Consequently:

• the continuous-time left-half plane maps inside the unit circle;
• the imaginary axis maps onto the unit circle; and
• the right-half plane maps outside the unit circle.

For s=σ+jω,

The modulo operation explains frequency aliasing: continuous frequencies separated by integer multiples of the sampling angular frequency 2π/T map to the same discrete-time angle.

The relation z=e^sT maps individual sampled modes exactly. It should not be confused with approximate filter-design substitutions such as the bilinear transform.

13. Applied Case: Sampled Telemetry Low-Pass Channel

Background

Consider an illustrative telemetry receiver that uses a first-order analog low-pass stage to suppress high-frequency noise before digital processing. The stage has time constant τ=1 ms and unity steady-state gain. Its input changes from 0 V to 2 V at t=0. Samples are taken every T=0.5 ms.

Problem

The engineer must determine the analog response, obtain an exact sampled recurrence under a zero-order-held input, assess stability and estimate the 2% settling time.

Analysis

The continuous-time transfer function is

The pole is at s=−1000 s⁻¹, so the analog circuit is stable. Its 2 V step response is

The sampled pole is

With a zero-order-held input and y[0]=0, the exact sampled model is

or numerically,

The corresponding transfer function is

Its pole, 0.60653, lies inside the unit circle. For the 2 V held input,

The normalized error is aⁿ. For 2% settling,

Since n≥ln(0.02)/ln(0.60653)≈7.824, the first integer meeting the requirement is n=8. The sampled settling time is therefore 4 ms. The continuous estimate is −τln(0.02)≈3.912 ms.

Outcome

The transform analysis confirms that both the analog and sampled systems are stable and that the digital model accurately reproduces the analog samples under the stated zero-order-hold assumption.

Learning Points

• Pole mapping provides a direct connection between analog and digital stability.
• Sampling and hold assumptions must be stated before constructing a discrete model.
• Settling requirements must be converted to an integer number of samples.
• Coefficient rounding may alter pole locations in fixed-point implementations.

14. Limitations, Implementation and Professional Practice

Transform methods are strongest for LTI systems. Nonlinear, time-varying or strongly uncertain systems may require linearization, state-space models, numerical simulation or nonlinear analysis. Other practical limitations include:

• pole-zero cancellation sensitivity;
• inaccurate initial-condition measurements;
• sampling aliasing and unmodeled hold behavior;
• finite-word-length overflow, quantization and coefficient errors;
• improper use of final-value theorems for unstable or oscillatory systems; and
• numerical ill-conditioning in high-order polynomial factorizations.

Modern engineering software can compute symbolic transforms, residues, pole-zero maps and simulated responses. Such tools should support rather than replace analytical verification. Best practice is to check initial values, final values, units, stability and selected time-domain samples independently.

In communication, medical, automotive and safety-critical systems, undocumented model assumptions can create governance and safety risks. Engineers should preserve sampling rates, coefficient precision, transform conventions and validation evidence. AI-assisted design tools may suggest models or filters, but their results require human verification against physical constraints and regulatory requirements.

Chapter Summary

The Laplace and Z-transforms convert continuous- and discrete-time system equations into algebraic forms. Unilateral transforms incorporate initial conditions, while bilateral transforms emphasize signal structure and regions of convergence. Convolution becomes multiplication in the transform domain, enabling efficient determination of LTI responses.

Transfer functions describe zero-state behavior. Their poles, zeros and ROCs reveal natural modes, causality and stability. A causal continuous-time rational system is BIBO stable when its poles lie in the open left-half plane; a causal discrete-time rational system is stable when its poles lie inside the unit circle.

Initial- and final-value theorems provide useful checks but require strict applicability conditions. Nonzero initial conditions must be handled explicitly to distinguish zero-input, zero-state and total responses. Continuous and sampled modes are connected through z=e^sT, allowing analog stability and transient behavior to be interpreted in the discrete domain.

Key Terms

Bilateral Laplace transform: Integral transform defined over all continuous time.

Unilateral Laplace transform: Transform over nonnegative time used to incorporate initial conditions.

Z-transform: Transform representing a discrete sequence as a function of a complex variable.

Region of convergence: Set of complex-variable values for which a transform exists.

Inverse transform: Recovery of a time-domain signal or sequence from its transform.

Convolution: Operation that combines an input with an impulse response to obtain the output.

Impulse response: Output of an initially relaxed LTI system when driven by a unit impulse.

Transfer function: Zero-state output-to-input transform ratio.

Pole: Value of the transform variable at which a rational transfer function becomes unbounded.

Zero: Value of the transform variable at which a transfer function becomes zero.

Zero-input response: Response caused only by initial stored energy.

Zero-state response: Response caused only by the input with zero initial conditions.

BIBO stability: Property that every bounded input produces a bounded output.

Causality: Property whereby output does not depend on future input values.

Initial-value theorem: Rule for finding a signal's initial value from its transform.

Final-value theorem: Rule for finding a stable signal's limiting value from its transform.

Difference equation: Discrete-time relation involving shifted input and output samples.

Pole mapping: Conversion of a continuous-time pole to a sampled pole using z=e^sT.

Review Questions

Short Answer Questions

1. Distinguish between unilateral and bilateral Laplace transforms.
2. Why must the ROC accompany a bilateral Z-transform?
3. Define zero-input and zero-state responses.
4. State the continuous-time convolution integral for causal signals.
5. State the pole condition for BIBO stability of a causal rational discrete-time system.

Long Answer Questions

1. Derive the unilateral Laplace transform of the nth derivative, showing how initial conditions enter.
2. Explain partial-fraction inversion for distinct and repeated Laplace-domain poles.
3. Discuss how poles, zeros, causality and the ROC jointly determine a system's behavior.
4. Derive the transfer function of a general constant-coefficient difference equation under initial rest.
5. Compare the continuous- and discrete-time initial- and final-value theorems, including their applicability conditions.

Analytical / Application Questions

1. Solve y'(t)+4y(t)=2u(t) for y(0^-)=3, separating the zero-input and zero-state responses.
2. For y[n]−0.25y[n−1]=x[n], y[−1]=4 and x[n]=δ[n], determine the zero-input, zero-state and total responses.
3. A continuous pole at s=−200+j500 rad/s is sampled at T=1 ms. Find the corresponding Z-plane pole and assess discrete-time stability.

Multiple Choice Questions with Answers

Question 1

For the unilateral Laplace transform, what is L{y''(t)} when y(0^-)=y₀ and y'(0^-)=v₀?

A. s²Y(s)−sy₀−v₀

B. s²Y(s)−y₀−sv₀

C. sY(s)−y₀−v₀

D. s²Y(s)+sy₀+v₀

Correct Answer: A. s²Y(s)−sy₀−v₀

Question 2

For y''+3y'+2y=u(t), y(0^-)=1 and y'(0^-)=−1, what is the total response for t≥0?

A. e^−t

B. 1/2+(1/2)e^−2t

C. 1−e^−2t

D. 1/2−e^−t+(1/2)e^−2t

Correct Answer: B. 1/2+(1/2)e^−2t

Question 3

If x(t)=u(t) and h(t)=e^−2tu(t), what is their convolution?

A. e^−2tu(t)

B. (1+e^−2t)u(t)/2

C. (1−e^−2t)u(t)/2

D. 2(1−e^−t)u(t)

Correct Answer: C. (1−e^−2t)u(t)/2

Question 4

Which pole set makes a causal rational continuous-time system BIBO stable?

A. All poles satisfy Re(p)<0

B. All poles satisfy Re(p)≤0

C. At least one pole lies at the origin

D. All poles lie inside the unit circle

Correct Answer: A. All poles satisfy Re(p)<0

Question 5

What is the Z-transform and ROC of δ[n]?

A. z, with |z|>1

B. 1, with the entire z-plane as ROC

C. 1/(1−z⁻¹), with |z|>1

D. z⁻¹, with |z|<1

Correct Answer: B. 1, with the entire z-plane as ROC

Question 6

What is the causal inverse Z-transform of z/(z−0.4)?

A. (0.4)ⁿ u[n

B. −(0.4)ⁿ u[−n−1

C. (2.5)ⁿ u[n

D. δ[n−1

Correct Answer: A. (0.4)ⁿ u[n

Question 7

For y[n]−0.5y[n−1]=δ[n] with y[−1]=2, what is the zero-input response?

A. 2(0.5)ⁿ u[n

B. (0.5)ⁿ u[n

C. (0.5)ⁿ⁺¹u[n

D. 2ⁿ u[n

Correct Answer: B. (0.5)ⁿ u[n

Question 8

When is the discrete final-value theorem lim_n→∞x[n]=lim_z→1(1−z⁻¹)X(z) valid?

A. When every pole of X(z) is outside the unit circle

B. When all poles of (1−z⁻¹)X(z) are strictly inside the unit circle

C. Whenever X(1) exists

D. Only for finite-duration sequences

Correct Answer: B. When all poles of (1−z⁻¹)X(z) are strictly inside the unit circle

Question 9

A discrete transfer function has poles at 0.6 and −0.8 and is causal. What is its BIBO stability classification?

A. Unstable because one pole is negative

B. Marginally stable because one pole has magnitude 0.8

C. Stable because both pole magnitudes are less than one

D. Unstable because the poles are unequal

Correct Answer: C. Stable because both pole magnitudes are less than one

Question 10

A continuous-time pole is s=−1000 s⁻¹ and the sampling period is T=0.5 ms. What is the corresponding sampled pole?

A. e^0.5≈1.6487

B. e⁻²≈0.1353

C. e^−0.5≈0.6065

D. −0.5

Correct Answer: C. e^−0.5≈0.6065