Skip to content
Tutorials Center
My learning

Topics

Lesson 12 of 37 ยท Time-domain Analysis

Time-Domain Specifications

Define rise time, peak time, overshoot and settling time, translate them into pole regions, and design a gain to meet specifications, verified in Python.

A customer never says "I want poles at -3 plus or minus 4j". They say "reach the target quickly, do not overshoot much, and settle within two seconds". Time-domain specifications state such wishes in numbers measured from the step response. This lesson defines them, shows how they map to places in the s-plane, and then designs a controller gain to meet a specification.

The standard step-response measures

For a unit step with final value y_ss:

  • Rise time t_r: time to go from 10% to 90% of y_ss (sometimes 0 to 100% for underdamped systems; always say which).
  • Peak time t_p: time of the first peak.
  • Percent overshoot Mp: (y_peak - y_ss) / y_ss x 100%.
  • Settling time t_s: time after which the response stays within 2% (or 5%) of y_ss.
  • Steady-state error e_ss: difference between reference and output as time goes to infinity (see Steady-State Error).
  • Delay time t_d: time to reach 50% of the final value.
flowchart LR
    A["Customer wants"] --> B["Mp, ts, tr, ess"]
    B --> C["Allowed pole region"]
    C --> D["Choose gain or controller"]
    D --> E["Simulate and check"]

From wishes to a design: translate, choose, then verify by simulation.

From specifications to pole regions

For the standard second-order system (and for any system with a dominant pole pair), the formulas of the previous lesson give limits on pole position:

  • Mp <= M0 means zeta >= zeta_min, where zeta_min = -ln(M0/100) / sqrt(pi^2 + ln(M0/100)^2). Geometrically, poles must lie in a wedge around the negative real axis.
  • t_s <= T0 means zeta wn >= 4 / T0, a vertical line: poles must lie to the left of Re(s) = -4/T0.
  • t_r small means wn large: poles must lie outside a circle of some radius.

The intersection of these regions is the acceptable region for the dominant poles.

Specifications become regions

Overshoot A wedge; the angle limits zeta.
Settling time Left of a vertical line.
Rise time Outside a circle of radius wn.

Poles in the shared region meet all three.

1 / 2

Computing the pole region

Python Minimum damping ratio from overshoot

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

An overshoot of 16.3% corresponds to zeta = 0.5, a 60 degree wedge. This is the same formula as before, solved for zeta.

A design example: choosing a gain

A position servo has the plant G(s) = 1 / (s (s + 2)) and unity feedback with proportional gain K. The closed loop is

Text
T(s) = K / (s^2 + 2 s + K)

Comparing with the standard form: wn = sqrt(K) and 2 zeta wn = 2, so zeta = 1 / sqrt(K). Requirement: at most 10% overshoot and 2% settling time under 4 seconds.

  • Overshoot: zeta >= 0.591, so K <= 1 / 0.591^2 = 2.86.
  • Settling: ts = 4 / (zeta wn) = 4 / 1 = 4 s for every K here, because zeta wn = 1 always. The real part of the poles is fixed at -1.

So the best we can do with this gain alone is ts = 4 s and the settling spec is just barely met, for any K <= 2.86. To get faster settling we would need a different controller. This kind of observation, that one parameter cannot meet every spec, motivates the compensators of Chapter 6.

Python Sweeping the gain K

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

You will see overshoot rising with K while the rise time falls, which is the standard speed-versus-overshoot trade-off. For K < 1 the response is overdamped and slow (K = 1 is critically damped). For the underdamped rows the settling time stays between about 3.5 s and 4.2 s, close to the 4 s predicted by ts = 4/(zeta wn).

Measuring specs from any response

Real responses rarely match the formulas exactly, so always measure. The function below turns a simulated response into a table you can compare with the requirements.

Python A reusable step_info function

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The system is third order and its true steady state is 20/26 = 0.769 (not 1). The function measures overshoot relative to the actual final value, so it still works.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. Which specification limits the damping ratio from below?

    Overshoot depends only on zeta, so a maximum overshoot gives a minimum damping ratio.
  2. 2. The requirement ts <= 2 s (2% band) means the dominant poles must satisfy

    ts = 4/(zeta wn) <= 2 gives zeta wn >= 2, so the poles lie to the left of Re(s) = -2.
  3. 3. For T(s) = K/(s^2 + 2s + K), what is zeta in terms of K?

    wn^2 = K and 2 zeta wn = 2, so zeta = 1/wn = 1/sqrt(K). Larger K means less damping.
  4. 4. Increasing K in that loop typically

    More gain means higher natural frequency (faster) but lower damping (more overshoot).

Summary

  • Key measures: rise time, peak time, percent overshoot, settling time and steady-state error.
  • For a dominant second-order pair, overshoot gives a wedge, settling time a vertical line, and rise time a circle in the s-plane.
  • A single gain may not be able to meet all specifications; compensators fix that.
  • Always measure the actual simulated response, especially for higher-order systems.

Next: Steady-State Error and System Type