Control a Process Manually

The Core Idea

In manual control, you choose the controller output yourself and watch how the process responds. The target guides your decisions, but an automatic controller is not adjusting the output for you.

What you will learn

• Use manual output to move a process toward its target.

• Allow time for a command to affect the measurement before adjusting again.

• Explain why a manual setting may need changing when the load changes.

Why it matters

Operating the process by hand makes its response and waiting time tangible before you ask an automatic controller to do the same job.

Explained like you're five

Adjust a shower tap and wait to feel the temperature change. If you keep turning it before the new water reaches you, you can go past the temperature you wanted. You have to look, make an adjustment, and give it time.

Here, you do that job for a heater. You choose the output yourself; the target is there to guide you.

The problem

Keep the heater near 50 °C. It starts at 20 °C, and you command controller output (CV) from 0% to 100%. SP is the target and PV is the measured temperature, just as in the feedback-loop lesson.

This experiment is in manual mode: you set CV directly, and no automatic controller corrects your choices. At 30 simulated seconds, an additional cooling load starts and stays on. The target remains 50 °C.

See it: hold one output

Set Simulator speed to 1× to give yourself time to operate. Set Manual output (CV %) to 30%, then press Run. Keep that output unchanged for the full 60 simulated seconds, including after the load starts.

Watch the first few seconds, the period around 30 s, and the end. Record SP, PV, and CV at those points. Does one fixed command meet the target under both loads? Press Save this trace when the run finishes.

Take over the adjustment

Use Restart in the same experiment. Restart keeps your current output setting, so set it back to 30% before running. The saved PV trace stays on the chart as your baseline.

This time, respond when the cooling load arrives. Pause if you need time to choose an output, change CV, then resume and watch what happens before making another change. Record when you changed CV, its old and new percentages, when PV first responded, and whether PV rose or fell.

Try a separate run starting at 60% for 10 s, then reduce the command to 30%. Does PV stop rising as soon as you reduce CV? Compare with your baseline, then restart for a more measured attempt.

What happened and why

Your fixed 30% run moves toward the target before the extra load, then falls away from it afterward. CV stays where you left it. A manual output does not change just because error has changed.

The heater also takes time to respond. There is an initial interval with no measured response, followed by a gradual temperature change. Several quick adjustments can be waiting to affect the heater at once, which is why reacting to every momentary gap can carry PV past SP.

An automatic controller repeats this observation and adjustment for you. Manual operation makes the reason for feedback tangible: the same output need not hold the same temperature when conditions change.

Check your run

Use the same experiment to complete a full 60 s trial. PV must come within 0.6 °C of SP at least once. The hands-on check also asks for an average absolute error no greater than 1.5 °C over the final 12 s, with overshoot no greater than 3 °C. These last two limits are 5% and 10% of the 30 °C start-to-target change; the check allows many output histories.

The result appears after the run finishes. It grades the measured response, so also keep your record of one deliberate output change and explain the delay you observed. You can restart and retry as often as you like.

Hint 1 — Which way should you move?

After the load arrives, compare PV with the unchanged 50 °C target. If this heater is too cool, should it receive more commanded power or less?

Hint 2 — Give each change time

Make one output adjustment, then wait through the initial 2 s delay and watch the trend. A falling PV just after you increase CV does not prove the increase was wrong: the new command has not taken full effect yet.

Hint 3 — A useful manual schedule

Start at 30%. After the cooling load starts at 30 s, try increasing to about 40%, then leave it there to observe the result. Changing at 35 s also meets the check in this model. This is one workable schedule, not a required sequence of clicks.

Compare your observations

With CV held at 30%, PV is about 48.2 °C just before the load and finishes near 40.4 °C. Raising CV to 40% at 35 s lets PV recover to about 49.4 °C by 60 s. The output change has a delayed, gradual effect; the target did not move.

Starting at 60% and waiting 10 s before reducing output produces a peak near 58 °C. Backing off after the process has already heated strongly leaves you dealing with the consequences of earlier commands.

Common mistakes

In the next lesson, you will separate the size, speed, and initial delay of that response so you can describe what kind of process you are controlling.

Enable JavaScript to run the embedded simulator and complete the interactive exercises.