Tools › Electrical and Controls › Control Systems › Transfer Function Analyzer › Example

Position servo K / (s(s + 1)(s + 2)) in a unity-feedback loop: stable range of K, margins, step response and root locus

ToolTransfer Function Analyzer
SourceOriginal case (the classic type-1 loop with three real open-loop poles), worked with closed forms and checked against numpy and scipy.signal in test_transfer_function_analyzer.php.
LevelUndergraduate

Problem

The open loop of a unity-feedback position servo is K G(s) = K / (s(s + 1)(s + 2)). For K = 3 find the closed-loop poles, the gain and phase margins and the step-response measures, and find the range of K for which the loop is stable.

Inputs to Enter

  • Numerator 1; denominator 1 3 2 0 (s³ + 3s² + 2s); gain K = 3
  • G(s) is the open loop, closed with unity feedback; input signal: unit step
Open the Transfer Function Analyzer tool (loaded with this example)

Answers

  • Characteristic polynomial s³ + 3s² + 2s + 3; Routh first column 1, 3, 1, 3: no sign change, stable
  • With K as a symbol the s¹ entry is (6 − K) / 3: stable for 0 < K < 6; at K = 6 the auxiliary polynomial 3s² + 6 gives s = ± j 1.414, an oscillation at √2 = 1.414 rad/s
  • Closed-loop poles −0.1642 ± j 1.047 (damping ratio 0.1549) and −2.672
  • Step response: rise time 1.207 s, overshoot 56.46 % (peak 1.565 at 3.379 s), settling time 22.2 s (2 %), final value 1
  • Gain margin 6 / 3 = 2 (6.021 dB) at 1.414 rad/s; phase margin 20.04° at 0.9693 rad/s; the loop could take 0.3608 s of time delay
  • Type 1: zero error to a step, K_v = 1.5 and a ramp error of 0.6667
  • Root locus: asymptotes at 60°, 180°, 300° from σ = −1; break point s = −1 + 1/√3 = −0.4226 at K = 2/(3√3) = 0.3849; crosses the imaginary axis at K = 6

Notes

The tool opens on this case; press Submit, or open it in the tool: /analysis-tools/electrical/controls/transfer-function-analyzer/?try=servo-feedback. The critical gain, its frequency, the break point and the asymptotes are closed forms; the step measures are checked against a 400,001-point scipy.signal response and the margins against a brute-force search.

More Examples for This Tool