Lesson 15 of 19 · Module 3: Stability analysis

Lead, lag and lead-lag compensation

Control System Engineering · EEE2510 · Sessions 42–45 · CO5

By the end of this lesson you can

  • Describe lead, lag and lead-lag networks
  • Choose a compensator to meet a specification
  • Outline the Bode design procedure

Notes

$$G_c(s) = K_c\frac{s + z}{s + p}$$

Lead ($z < p$)Lag ($z > p$)
PhasePositive (adds phase)Negative (small lag)
Main effectMore phase margin, faster response, wider bandwidthHigher low-frequency gain, smaller steady-state error
Similar toPDPI
DrawbackAmplifies high-frequency noiseSlower response

Maximum lead phase $\phi_m = \sin^{-1}\dfrac{1 - \alpha}{1 + \alpha}$ (with $\alpha = z/p < 1$) at $\omega_m = \sqrt{zp}$.

Lead design on the Bode plot

  1. Choose $K_c$ to meet the steady-state error requirement.
  2. Find the existing phase margin.
  3. Required lead $\phi_m$ = desired PM − existing PM + 5° to 12° margin.
  4. Find $\alpha$, place $\omega_m$ at the new gain crossover (where the uncompensated gain is $-10\log(1/\alpha)$ dB).
  5. Check the result.

Lead-lag combines both: the lag part fixes steady-state error and the lead part fixes transient response.

Design in MATLAB with margin() and bode(), then compare the step responses before and after compensation.

Lecture

🎬
Lecture video
Coming soon. The recorded lecture for this lesson will appear here.

Exercises

  1. A system needs 25° more phase margin. With a 5° safety margin, find $\alpha$ for a lead compensator.
    Show solution
    $\phi_m = 30°$: $\sin30° = 0.5 = (1-\alpha)/(1+\alpha) \Rightarrow \alpha = 1/3$.
  2. A system has a good transient response but too much steady-state error. Which compensator?
    Show solution
    Lag compensator: raises low-frequency gain while barely changing the phase near crossover.
  3. Why place the lag compensator's zero and pole well below the gain crossover frequency?
    Show solution
    So its phase lag does not reduce the phase margin; near crossover it then acts almost like a unity gain.