EIGENROOMLAB WORKSHEET 22ANSWER KEY · FOR INSTRUCTORSDC motor · Dead time and limits
Tuning against dead time
Goal
Answer the experiment's question with numbers read from the lab: On a motor with 200 ms of dead time, will the Ziegler–Nichols PI gains or half of them settle sooner?
Setup
Lab eigenroom.com/lab?lesson=detune-for-delay · short link eigenroom.com/detune-for-delay
Open it on a phone or a computer; there is nothing to log in to or install. The link loads run A's settings:
- Target speed
- 30 rad/s
- Loop
- Closed loop
- Controller type
- PID
- Proportional Kp
- 0.335
- Integral Ki
- 0.41
- Derivative Kd
- 0
- Anti-windup
- On
- Load torque
- 0 N·m
- Sensor noise
- 0
- Controller period
- 20 ms
- Disturbance events
- None
- Recording
- 12 s
- APreset, exactly as the link opens it.
- BColumn B: write down the one setting you changed.
1. Predict
On a motor with 200 ms of dead time, will the Ziegler–Nichols PI gains or half of them settle sooner?
Ziegler and Nichols aimed at quarter-amplitude decay. With 200 ms of dead time their PI gains leave a 19.1° phase margin: overshoot 63.13% and settling at 5.82 s. Halving both gains keeps the integral time but lowers the crossover from 4.33 to 2.33 rad/s, where the dead time takes 26.7° of phase instead of 49.6°. The margin rises to 47.4° and overshoot falls to 22.66%; the rise slows from 0.20 s to 0.38 s, yet the speed settles at 1.84 s. A tuning rule is a starting point: check it against the delay you measured.
2. Observe
| Quantity | Where to read it | A | B your change |
|---|---|---|---|
| Rise time, 10–90% of target (s) | Measured response | 0.20 s | — |
| Overshoot vs target (%) | Measured response | 63.13% | — |
| Settling, ±2% of target (s) | Measured response | 5.82 s | — |
| Final error (rad/s) | Measured response | -0.01 rad/s | — |
| Control effort (V·s) | Measured response | 30.46 V·s | — |
Depends on the change the student made.
Values computed by DC motor and cart-pole engine 1.5.0 at the settings above. A replay readout can differ in the last digit if the slider is not exactly at the stated time; values read from a plot are approximate.
3. Explain
Follow the lab's instruction: “Pin the Z–N run, then set P to 0.17 and I to 0.21, about half of each.” Record your changed run in column B. Which quantity changed the most, and in which direction?
Column A holds the preset values from the engine. Column B depends on the change each student made; check that only one setting changed between A and B.
Explain the change you measured. What you were asked to watch: “The preset is the Z–N PI suggestion from the reaction curve, Kp 0.336 and Ki 0.412, rounded to the sliders as 0.335 and 0.41. Compare overshoot, rise and settling with the halved gains, and the phase margin in the frequency workbench.”
The lab's explanation: Ziegler and Nichols aimed at quarter-amplitude decay. With 200 ms of dead time their PI gains leave a 19.1° phase margin: overshoot 63.13% and settling at 5.82 s. Halving both gains keeps the integral time but lowers the crossover from 4.33 to 2.33 rad/s, where the dead time takes 26.7° of phase instead of 49.6°. The margin rises to 47.4° and overshoot falls to 22.66%; the rise slows from 0.20 s to 0.38 s, yet the speed settles at 1.84 s. A tuning rule is a starting point: check it against the delay you measured.
Extension (optional)
Repeat the change with two more values of the same setting and sketch the measured quantity against the setting. Is the relationship linear?
Answers depend on the values chosen. Saturation, limits and nonlinear plant models usually make the relationship curve.