How Commissioning Engineers Measure Rise Time and Damping Ratios with a PLC Oscilloscope

Commissioning engineers use a PLC oscilloscope to measure step-response behavior, not just to watch tags move. In OLLA Lab, the embedded oscilloscope supports visual analysis of rise time, overshoot, settling, and damping ratio so loop behavior can be diagnosed and revised before logic reaches a live process.

A changing number in a watch window is not the same thing as a measured response. For PID commissioning, numerical tag observation alone cannot reliably show overshoot shape, settling behavior, oscillation decay, or valve-related lag. Time-series context is required.

A recent internal Ampergon Vallis benchmark found that users completing simulated pump-loop tuning tasks with the embedded OLLA oscilloscope reached a bounded “stable tuning” target faster than users relying on the variables panel alone. Ampergon Vallis metric: 62% faster median time to first stable tuning outcome. Methodology: n=500 simulated pump commissioning scenarios; task definition = achieve bounded stable response within scenario acceptance criteria after a 10% setpoint step; baseline comparator = variables-panel-only observation without oscilloscope trace; time window = January–March 2026. This supports the claim that visual waveform access improves diagnostic speed inside the simulated task. It does not support broader claims about field productivity, operator competence, or employability.

“Simulation-Ready,” in this context, means an engineer can prove, observe, diagnose, and harden control logic against realistic process behavior before it reaches a live process. That is a higher bar than knowing ladder syntax.

A visual oscilloscope is critical because PID tuning is a time-domain problem. Rise time, overshoot, settling, and oscillation decay are defined by waveform behavior over time, not by isolated values in a tag table.

What fails when engineers rely only on numerical monitoring?

Numerical monitoring is useful for state inspection, but weak for dynamic diagnosis. The failure modes are predictable:

- No visible time axis: Without a time base, settling time cannot be measured defensibly. - Poor visibility of overshoot shape: A changing integer may show that the PV crossed the SP, but not how sharply, how often, or with what decay pattern. - Aliasing at the human level: Even if tags update quickly, a human reading changing values cannot reconstruct a waveform accurately. - No direct comparison of signals: PID diagnosis often requires SP, PV, and CV on the same trace. - Weak fault discrimination: A flat PV with changing CV may indicate stiction, deadband, or process lag. A number alone does not volunteer that diagnosis.

A watch window answers “what is the value now?” Commissioning usually needs “what did the system just do, and why?” Those are different questions.

What does “guessing” mean in PID tuning?

In this article, guessing means heuristic trial-and-error tuning based mainly on changing numerical tags, without graphical measurement of the step response.

That does not mean heuristics are useless. Field engineers use them constantly. It means heuristics become weak when the response must be quantified, repeated, compared, or defended.

What does “engineering” mean in PID tuning?

In this article, engineering means measuring the system’s step response on a time-scaled visual trace and using that trace to calculate or estimate tuning-relevant quantities such as:

• rise time \(T_r\)
• peak overshoot \(M_p\)
• settling time \(T_s\)
• damping behavior
• decay ratio between successive peaks

The distinction is simple: tag watching is observation; waveform measurement is analysis.

Rise time is measured by applying a known step change, capturing the PV response, and timing how long the PV takes to move from 10% to 90% of its final value. That is the standard practical definition used in control engineering texts such as Ogata.

OLLA Lab is useful here as a bounded rehearsal environment. It allows engineers to induce step changes, observe SP/PV/CV behavior, pause simulation, and inspect the consequences without stressing live equipment. It is a validation environment, not an auto-tuner.

### Step-by-step: measuring rise time in the OLLA oscilloscope

Plot at least:

• Setpoint (SP)
• Process Variable (PV)

If the PV moved from \(PV_0\) to \(PV_f\), then:

• 10% level = \(PV_0 + 0.1(PV_f - PV_0)\)
• 90% level = \(PV_0 + 0.9(PV_f - PV_0)\)

1. Establish a stable baseline. Run the simulated process until the PV is steady near the initial setpoint.
2. Apply a defined step change. Use the variables panel to change the setpoint by a known amount, commonly 5% to 10% of span.
3. Display the relevant traces. In many cases, add Control Variable (CV) as well.
4. Let the response develop. Observe the PV as it begins moving toward the new steady-state value.
5. Pause or freeze the simulation if needed. OLLA Lab’s simulation controls are operationally useful here because they let the user inspect the waveform without the usual “blink and miss it” problem.
6. Determine the final value. Estimate the new steady-state PV after the transient settles.
7. Mark the 10% and 90% levels.
8. Measure elapsed time between those crossings. The time from the 10% crossing to the 90% crossing is the practical rise time \(T_r\).

Why does rise time matter during commissioning?

Rise time matters because it shows how aggressively the loop responds to a setpoint change or disturbance. A loop that is too slow may fail process objectives. A loop that is too fast may overshoot, hunt, or excite mechanical problems.

Fast is not always good. “Responsive” and “well-behaved” are not synonyms.

Peak overshoot is calculated from the first peak above the final steady-state value. Damping ratio is then inferred from the overshoot magnitude or from the decay between successive peaks, depending on the method used.

For a standard underdamped second-order approximation, peak overshoot is:

\(M_p = \frac{C(t_p) - C(\infty)}{C(\infty)} \times 100\%\)

Where:

• \(C(t_p)\) = value of the first peak
• \(C(\infty)\) = final steady-state value

This formula is only meaningful when the response is interpreted carefully. Real industrial loops are often higher-order, nonlinear, filtered, saturated, or valve-limited. The waveform still tells the truth, but the math must be applied with judgment.

How do engineers interpret damping visually?

The damping pattern can often be classified directly from the trace before any detailed calculation:

| Response Type | Damping Condition | What the OLLA Trace Looks Like | Practical Meaning | |---|---|---|---| | Underdamped | \(\zeta < 1\) | PV crosses SP, overshoots, and oscillates with decaying peaks | Fast but oscillatory response | | Critically damped | \(\zeta = 1\) | PV approaches final value quickly without oscillation | Fastest non-oscillatory response | | Overdamped | \(\zeta > 1\) | PV approaches final value slowly with no overshoot | Stable but sluggish response |

This classification is a practical approximation, not a declaration that the plant is a neat textbook second-order system.

How do you estimate damping ratio from overshoot?

For a second-order underdamped approximation, damping ratio \(\zeta\) can be estimated from the fractional overshoot \(M_p\) using:

\(\zeta = \frac{-\ln(M_p)}{\sqrt{\pi^2 + (\ln(M_p))^2}}\)

Where \(M_p\) is expressed as a fraction, not a percentage. For example, 20% overshoot means \(M_p = 0.20\).

This is useful when the waveform has a clear first peak and a credible final value. It becomes less reliable when the loop is heavily nonlinear, clipped by output limits, or disturbed by noise and deadband.

The quarter-decay ratio method evaluates how much successive oscillation peaks shrink. A classic target is that each peak is roughly one-quarter of the previous peak’s amplitude relative to the final value.

This method is historically associated with practical tuning rules such as Ziegler–Nichols. It is not sacred, and it is not universally optimal. It is a tuning heuristic anchored in measured response shape.

How is quarter-decay ratio measured on the scope?

4. Compute the ratio:

1. Apply a step change and capture an underdamped response.
2. Identify the first peak amplitude above the final value.
3. Identify the second peak amplitude above the final value.

\(\text{Decay Ratio} = \frac{\text{Second peak amplitude}}{\text{First peak amplitude}}\)

1. Compare the result to 0.25.

If the ratio is near 0.25, the response is near quarter-decay behavior.

What does the quarter-decay ratio tell you?

It tells you whether oscillations are dying out at a rate consistent with a classic aggressive tuning target.

- Ratio greater than 0.25: damping is weak; oscillations are dying out too slowly. - Ratio near 0.25: classic quarter-decay behavior. - Ratio much less than 0.25: response is more heavily damped.

This is useful for comparison, not worship. Many process loops should be tuned more conservatively than quarter-decay, especially where valve wear, thermal lag, water hammer, or interaction with upstream/downstream units matters.

Valve hysteresis or stiction can be diagnosed by comparing the control output trace against the process response trace. If the CV moves while the PV remains flat and then the PV suddenly jumps, the likely problem is mechanical or process-side nonlinearity rather than a ladder-logic error.

That distinction matters during commissioning. Otherwise, engineers start “fixing” logic that was innocent to begin with.

What waveform pattern suggests hysteresis or stiction?

A typical pattern includes:

• the CV changes smoothly
• the PV remains nearly unchanged
• after a threshold is reached, the PV moves abruptly
• the pattern may repeat differently on increasing versus decreasing output

This indicates deadband, stiction, backlash, or hysteresis in the final control element or process path.

Why is the oscilloscope better than a tag list for this diagnosis?

The oscilloscope shows temporal causality. It reveals that the controller commanded movement before the process responded. A numerical panel can show both values changing, but often hides the delay pattern that distinguishes mechanical resistance from poor tuning.

In OLLA Lab, the value is bounded but real: the engineer can rehearse the diagnostic sequence safely, compare ladder state to simulated equipment state, and revise logic or assumptions before touching a live valve.

Useful waveform measurement depends on sampling discipline. If the trace is too coarse, the engineer measures the display artifact instead of the process behavior.

What sampling practices improve measurement quality?

Faster loops need shorter sample intervals.

• Match sample time to loop dynamics.

Sparse traces can hide overshoot peaks and distort rise time.

• Avoid excessive downsampling.

Single-signal plots are often insufficient for diagnosis.

• Trend SP, PV, and CV together.

A compressed trace hides detail; an over-zoomed trace hides context.

• Keep scaling readable.

Comparison across tuning revisions requires consistent excitation.

• Use repeatable step sizes.

A trace is only as honest as the sampling behind it. Oscilloscopes are not magic; they are merely less forgiving than intuition.

Example configuration block

[Language: Structured Text] PID_Pump.Ts := 0.05; // 50 ms sample time PID_Pump.Kp := 2.5; // Proportional gain PID_Pump.Tn := 1.2; // Integral time

This example does not prescribe correct tuning values for a real plant. It shows the principle that controller update timing and waveform visibility should be aligned when analyzing response behavior.

“Simulation-Ready” means the engineer can produce evidence that control logic behaves correctly under normal, transitional, and faulted conditions before deployment. It is an operational standard, not a flattering adjective.

For oscilloscope-based debugging, that means the engineer can:

• define what “correct” response looks like
• induce a controlled disturbance or setpoint step
• capture SP, PV, and CV traces
• identify overshoot, lag, oscillation, or deadband
• revise logic or tuning based on measured behavior
• retest under the same conditions

This is where OLLA Lab becomes operationally useful. It supports rehearsal of high-risk commissioning tasks that are expensive, disruptive, or unsafe to learn for the first time on live equipment.

What engineering evidence should a learner or junior engineer build?

Do not build a screenshot gallery. Build a compact body of engineering evidence:

1. System Description Define the process, loop purpose, and control objective.
2. Operational definition of “correct” State measurable acceptance criteria such as allowable overshoot, rise time range, settling time, or fault response.
3. Ladder logic and simulated equipment state Show the logic and the associated simulated machine or process behavior.
4. The injected fault case Document the abnormal condition introduced, such as sensor lag, stuck valve behavior, noisy analog input, or failed permissive.
5. The revision made Record the tuning change, interlock revision, filter addition, or sequence correction.
6. Lessons learned State what the waveform proved, what the original assumption missed, and what changed after revision.

That structure is more credible than “here is a rung and it seems fine.”

Oscilloscope-based diagnosis in a simulator is valuable, but bounded. A simulator can reproduce control logic behavior, process approximations, and fault patterns, yet it does not erase the gap between simulated validation and field deployment.

What OLLA Lab supports credibly

OLLA Lab supports:

• ladder logic development in a browser-based environment
• simulation of logic execution and I/O behavior
• observation of variables and analog behavior
• scenario-based rehearsal of process sequences and faults
• digital-twin-style validation against realistic machine models
• guided learning and AI-assisted support through GeniAI

In this article’s context, the key value is narrower: it provides a safe environment to observe and measure the consequences of control logic and tuning changes before physical deployment.

What OLLA Lab does not claim to replace

OLLA Lab does not replace:

• site acceptance testing
• instrument calibration
• valve signature testing
• SIL verification
• formal functional safety assessment
• operator training on the exact live plant
• field competence earned under actual site conditions

A simulated loop can save wear, time, and embarrassment. It cannot sign the turnover package.

Oscilloscope evidence should drive specific, testable revisions. The point is not to admire the waveform. The point is to change the loop intelligently.

Common waveform observations and likely actions

Likely action: reduce aggressiveness, review proportional gain, integral action, and process dead time assumptions.

• High overshoot with repeated oscillation

Likely action: increase responsiveness if process constraints allow.

• Very slow rise with no overshoot

Likely action: investigate stiction, hysteresis, deadband, or output scaling.

• CV movement with delayed PV jump

Likely action: review filtering, sensor quality, and derivative sensitivity if used.

• Noisy PV causing unstable control action

Likely action: inspect integral behavior, interaction effects, or actuator saturation.

• Long settling despite acceptable rise time

The revision cycle should be explicit: measure, infer, revise, retest.

A PLC oscilloscope matters because commissioning is a measurement problem before it becomes a tuning problem. Rise time, overshoot, settling, and damping ratio are observable properties of how a loop behaves after a change.

OLLA Lab’s embedded oscilloscope is best understood as a bounded diagnostic environment for that work. It does not tune loops automatically, certify competence, or replace field commissioning. It does allow engineers to induce step changes, compare SP/PV/CV behavior, pause the simulation, inspect abnormal response patterns, and revise logic before the process is real and expensive.

That is the practical shift from syntax to deployability.

- IEC 61508 Functional Safety overview - exida Functional Safety and Automation resources - Åström & Hägglund, *Advanced PID Control* (ISA) - Skogestad (2003), PID tuning rules00062-8) - Tao et al. (2019), Digital Twin in Industry