Feedback Control, Transfer Functions and Stability

Scope: feedback structure; block algebra; closed-loop poles; frequency margins. Cross-cutting case: temperature-loop performance in a continuous process unit.

What this unit builds

Closed-loop behavior depends on the entire measurement-to-actuation path, rather than on the process alone. A regulator can be mathematically stable and still have unacceptable control quality because the actuator moves too often, the sensor is biased, the sample interval is irregular or the achievable process output is constrained. Performance reporting therefore includes both controlled-variable error and the physical effort used to obtain it.

Lesson sequence

1. [Negative feedback and closed-loop sensitivity](../lessons/unit-03-lesson-01.html) — Negative feedback compares setpoint with measured output and uses an actuator to reduce error. The correct feedback sign depends on actuator and process direction; a valve increasing cooling may have a negative effect on temperature. Governing relation: `Y/R = CG/(1+CG)`.

2. [Block diagrams and transfer functions](../lessons/unit-03-lesson-02.html) — Block diagrams separate the controller, physical process, sensor, disturbance path and reference signal. Rearranging blocks algebraically does not remove the signal measurement noise or imply a static gain where a dynamic transfer is required. Governing relation: `G_cl(s)=G_c G_p/(1+G_c G_p H)`.

3. [Characteristic equations and pole locations](../lessons/unit-03-lesson-03.html) — Closed-loop stability is a property of pole locations, not a claim about the apparent smoothness of one plotted response. A pole with positive real part creates exponential growth in a linear continuous-time model. Governing relation: `1+C(s)G(s)H(s)=0`.

4. [Bode response and stability margins](../lessons/unit-03-lesson-04.html) — Gain and phase margins quantify proximity to the −1 point in a nominal loop. They do not form a universal stability guarantee under arbitrary model mismatch, and margins must use the correct crossover frequencies. Governing relation: `L(jw)=C(jw)G(jw)H(jw)`.

Model connections and physical interpretation

Negative feedback and closed-loop sensitivity

Negative feedback compares setpoint with measured output and uses an actuator to reduce error. The correct feedback sign depends on actuator and process direction; a valve increasing cooling may have a negative effect on temperature. The simple fraction assumes stable linear elements and unity measurement dynamics; actuator saturation breaks superposition. The calculation that anchors this lesson is deliberately synthetic: For a static positive plant gain G=2 and proportional controller C=3, setpoint closed-loop gain is 6/7=0.857 and steady setpoint error is 1/7 of the commanded change. The reasoning should be checked in the next topic, because a dynamic approximation useful for one type of experiment may be inappropriate when the controlled input, disturbance or measurement location changes.

Block diagrams and transfer functions

Block diagrams separate the controller, physical process, sensor, disturbance path and reference signal. Rearranging blocks algebraically does not remove the signal measurement noise or imply a static gain where a dynamic transfer is required. This nominal response assumes no transport delay and unbounded actuator authority; fast mathematical poles are not proof of fast equipment. The calculation that anchors this lesson is deliberately synthetic: For Gp=2/(10s+1), C=4 and H=1, Gcl=8/(10s+9). The closed-loop time constant is 10/9=1.11 s, while the DC setpoint gain is 8/9. The reasoning should be checked in the next topic, because a dynamic approximation useful for one type of experiment may be inappropriate when the controlled input, disturbance or measurement location changes.

Characteristic equations and pole locations

Closed-loop stability is a property of pole locations, not a claim about the apparent smoothness of one plotted response. A pole with positive real part creates exponential growth in a linear continuous-time model. Linear stability around an operating point does not guarantee safety in the presence of hard stops, delays or nonlinear runaway. The calculation that anchors this lesson is deliberately synthetic: The polynomial s²+4s+3 factors as (s+1)(s+3), so both modes decay. Replacing the constant term by −3 instead yields roots −2+sqrt(7)≈0.646 and −2−sqrt(7)≈−4.646, demonstrating one unstable mode. The reasoning should be checked in the next topic, because a dynamic approximation useful for one type of experiment may be inappropriate when the controlled input, disturbance or measurement location changes.

Bode response and stability margins

Gain and phase margins quantify proximity to the −1 point in a nominal loop. They do not form a universal stability guarantee under arbitrary model mismatch, and margins must use the correct crossover frequencies. A frequency-response model is valid only over the tested frequency region; sampled-data effects alter high-frequency behavior. The calculation that anchors this lesson is deliberately synthetic: A pure 5 s delay contributes −omega*5 radians of phase. At omega=0.20 rad/s this is −1 rad=−57.3°, although its magnitude remains exactly one. The reasoning should be checked in the next topic, because a dynamic approximation useful for one type of experiment may be inappropriate when the controlled input, disturbance or measurement location changes.

Unit synthesis: a reproducible experiment

Start from temperature-loop performance in a continuous process unit. Draw the boundary and place every relevant measurement on it. Give the operating point, consistent units and any assumptions about constant density, heat capacity or sensor behavior. Construct a hypothesis that could be tested with a small safe, synthetic input perturbation. Predict the output sign and a first approximation for the relevant time scale before computing anything.

Next compare a nominal model with one deliberately misspecified version. Change a gain, a lag, a delay or a sensor bias rather than modifying all uncertain features at once. Quantify the difference in output peak, settling time, residual steady error and manipulated-variable excursion. Record when the model ceases to be a good representation of the exercise. A successful classroom exercise is a transparent chain of assumptions and checks, not merely a curve.

Finally design a holdout test. In a process setting this might be a second flow rate, a different operating temperature or a load disturbance that was not included in model fitting. Compare model and measurement on the same time grid, keep a reproducible data table, and explain why any fitted coefficient cannot be assumed constant outside its observed range. Treat alarm and protective requirements as a separate design decision.

Unit knowledge check

1. For Negative feedback and closed-loop sensitivity, state the terms of Y/R = CG/(1+CG), choose a limiting or equilibrium case, and explain when it would be invalid.
2. For Block diagrams and transfer functions, state the terms of G_cl(s)=G_c G_p/(1+G_c G_p H), choose a limiting or equilibrium case, and explain when it would be invalid.
3. For Characteristic equations and pole locations, state the terms of 1+C(s)G(s)H(s)=0, choose a limiting or equilibrium case, and explain when it would be invalid.
4. For Bode response and stability margins, state the terms of L(jw)=C(jw)G(jw)H(jw), choose a limiting or equilibrium case, and explain when it would be invalid.

Continue to the [Unit 3 assessment](../assessments/unit-03.html) or [Unit 3 project](../projects/unit-03.html).

Unit 3, Lesson 1: Negative feedback and closed-loop sensitivity

Unit: Feedback, Transfer Functions, and Stability. Physical setting: temperature-loop performance in a continuous process unit. Reference model: Y/R = CG/(1+CG).

Learning outcomes

By the end of this lesson you should be able to (i) explain negative feedback and closed-loop sensitivity in physical terms, (ii) identify the variables and assumptions of Y/R = CG/(1+CG), (iii) reproduce and challenge a numerical benchmark, (iv) recognize how measurement, delay and actuator limits modify the model, and (v) design a specific experiment that can distinguish the current model from an alternative interpretation.

Physical reasoning, not just a formula

Negative feedback compares setpoint with measured output and uses an actuator to reduce error. The correct feedback sign depends on actuator and process direction; a valve increasing cooling may have a negative effect on temperature.

Consider the unit-wide equipment case: temperature-loop performance in a continuous process unit. In this setting the same numerical trend may have more than one physical cause. A change observed at the analyzer may reflect a real process response, a changed flow residence time or the instrument dynamics. The analyst must first specify which quantity is manipulated, which is observed and which variables are treated as external disturbances. This is particularly important for negative feedback and closed-loop sensitivity, where confusing a state with a measured output can give a mathematically solvable but physically incorrect model.

The chosen governing expression is Y/R = CG/(1+CG). It is a compact model, not a complete plant specification. Terms on opposite sides of an equality must have compatible units and consistent time bases. When a parameter is estimated from one transient, its result belongs to that operating range unless new data demonstrate transferability. The task is to explain the meaning of the relation before substituting numbers and then to compare its predictions with a checkable boundary condition or steady state.

Working the numerical reference case

Synthetic worked example. For a static positive plant gain G=2 and proportional controller C=3, setpoint closed-loop gain is 6/7=0.857 and steady setpoint error is 1/7 of the commanded change.

First transcribe every numerical input into a short table with its physical unit. Second, compute the baseline result without a software solver. Third, identify which output should change and which should remain unchanged when the manipulated input is zero. In this negative feedback and closed-loop sensitivity exercise, the reported figures are not fitted plant data: they are an internally stated teaching benchmark. Any extension to different equipment requires new coefficients and new validation.

A useful sensitivity check is to change the selected coefficient or input by +10% while retaining all other conditions. This is not equivalent to assuming the output must also rise by 10%; exponential responses, ratios, saturation and feedback can change the mapping. Write the original expression, substitute the perturbed value and compare the sign of the result against physical intuition. Next repeat with a −10% change. An asymmetric response is evidence of nonlinearity in the selected mapping, not necessarily a numerical error.

Derivation and model restrictions

The simple fraction assumes stable linear elements and unity measurement dynamics; actuator saturation breaks superposition. To turn a physical inventory into a differential state equation, write accumulation as incoming rate minus outgoing rate plus net production, if relevant. To turn an instrument or actuator into a dynamical element, specify its output, time constant and initial condition. To build a transfer function, take deviations about a steady operating point and verify that the assumed initial deviations are zero. Do not apply frequency-domain algebra blindly to a strongly nonlinear or switching regime.

For negative feedback and closed-loop sensitivity, the limiting checks should include a zero-change input, a long-time equilibrium where one exists and at least one infeasible input that the physical equipment cannot realize. If a parameter has dimensions of reciprocal time, use it only with a time expressed in matching units. If a parameter multiplies a temperature difference, its unit must convert the difference into a heat rate before the quantity is added to the energy balance. These simple checks detect mistakes that a graph renderer will not.

How to identify the model experimentally

Use the previously studied topic Negative feedback and closed-loop sensitivity as a source of prior knowledge and the next topic Block diagrams and transfer functions as an independent challenge. Define the reference operating point for negative feedback and closed-loop sensitivity in the process case and record a small safe input sequence, the actual actuator trajectory, the measurement times and instrument range. A step, a pulse and a slow ramp probe different properties, so do not collapse them into one claimed validation. If the fitted output differs from the recorded trace, inspect the timestamp alignment and sensor dynamics before increasing model order.

Where possible, obtain at least two input amplitudes. Normalize the measured output by the change in input and test whether the curves approximately collapse. Failure to collapse can indicate nonlinearity, saturation, temperature-dependent coefficients or changed initial conditions. A second holdout operating point is especially important when the main formula was derived by linearization. Report the residual error and uncertainty rather than relabeling an imperfect approximation as a universal physical law.

Uncertainty, troubleshooting, and design decisions

Failure mode to investigate: Changing from heating to cooling equipment without verifying control action can create reinforcing rather than correcting feedback. The first diagnostic question is which conservation law, measurement chain or actuator path the proposed explanation implicates. Reconstruct the actual manipulated-variable history and compare it with what the controller requested; a large difference is a plant-side constraint, not a parameter-estimation error that can be fixed by fitting a more elaborate process transfer function.

Decision from the evidence: Trace a positive sensor perturbation through controller, valve and plant and verify the predicted corrective direction. State what the experiment supports and what remains unidentifiable. The decision should include whether the next useful action is to add a sensor, run a second input condition, change an assumed model structure or restrict the admissible operating range. When model outputs could guide actual equipment, independent engineering review and safety analysis are necessary before any implementation.

A challenge to the simplest explanation

Compare the present explanation with Characteristic equations and pole locations, which appears elsewhere in this unit. Imagine obtaining a smooth measured response that both models could plausibly produce. Choose one additional measurement that would change differently under the two explanations. Define the expected sign and time scale under each. For negative feedback and closed-loop sensitivity, record not just peak error but also the total actuator travel, disturbances, measurement noise and how the reference signal was generated. This exercise forces you to separate apparent goodness of fit from the ability to predict new conditions.

Numerical implementation and independent checks

The companion [interactive model, Laboratory 06](../labs/lab-06.html) is Proportional feedback and offset. Change one slider at a time, export the series and independently reproduce at least one value from a governing equation. The browser tool is deterministic and offline. It represents a defined educational model, not an approved process simulator. Its chart scale may rescale as parameters change, so compare numerical coordinates rather than visual slope alone.

For a numerical transient, verify step-size sensitivity by recomputing with half the time increment. If a predicted instability disappears merely when the integration time step shrinks, it was likely numerical rather than physical. On the other hand, an unstable physical model may diverge more quickly when more accurately integrated. Preserve initial conditions, parameter units and exact software version in the report so another student can repeat the test.

Practice and oral-examination questions

1. Starting from Y/R = CG/(1+CG), identify its state(s), manipulated input(s) and measured output(s). State how you would verify all dimensions.
2. Recalculate the example after a 10% increase of the principal process parameter. What changes in the predicted steady value and in the transient, and why are these not necessarily the same?
3. Propose a sensor or actuator failure that would make an otherwise correct negative feedback and closed-loop sensitivity model appear wrong. What extra logged variable would distinguish that failure?
4. Explain how Characteristic equations and pole locations might confound the identification of this lesson’s mechanism and propose a discriminating test.
5. State one safety or operating limit that the simplified Y/R = CG/(1+CG) relation does not capture. How should an engineer document that omission?

Answer guide and reflection

Question 1 is assessed by correct units and by separating the actual physical inventory from its measured signal; a memorized symbolic expression without those definitions is insufficient. For Question 2, the correct method uses the original equation twice and recomputes the equilibrium or time constant where applicable; simple proportional scaling is not accepted without showing that the equation is linear in the varied coefficient. For Question 3, the measurement or actuator diagnosis must lead to a different predicted trace. For Question 4, a proposed experiment is credible only if it can distinguish two candidate mechanisms. For Question 5, a constraint must be named and its omission explicitly stated.

Takeaway for Feedback, Transfer Functions, and Stability. The lesson is complete only after the model can be challenged with a changed initial condition, a disturbance and a non-ideal measurement. A plotted agreement under one nominal set of synthetic conditions is the start of investigation, not its end. The next step is to test whether the assumptions retain their meaning when block diagrams and transfer functions is included in the model.

Unit 3, Lesson 2: Block diagrams and transfer functions

Unit: Feedback, Transfer Functions, and Stability. Physical setting: temperature-loop performance in a continuous process unit. Reference model: G_cl(s)=G_c G_p/(1+G_c G_p H).

Learning outcomes

By the end of this lesson you should be able to (i) explain block diagrams and transfer functions in physical terms, (ii) identify the variables and assumptions of G_cl(s)=G_c G_p/(1+G_c G_p H), (iii) reproduce and challenge a numerical benchmark, (iv) recognize how measurement, delay and actuator limits modify the model, and (v) design a specific experiment that can distinguish the current model from an alternative interpretation.

Physical reasoning, not just a formula

Block diagrams separate the controller, physical process, sensor, disturbance path and reference signal. Rearranging blocks algebraically does not remove the signal measurement noise or imply a static gain where a dynamic transfer is required.

Consider the unit-wide equipment case: temperature-loop performance in a continuous process unit. In this setting the same numerical trend may have more than one physical cause. A change observed at the analyzer may reflect a real process response, a changed flow residence time or the instrument dynamics. The analyst must first specify which quantity is manipulated, which is observed and which variables are treated as external disturbances. This is particularly important for block diagrams and transfer functions, where confusing a state with a measured output can give a mathematically solvable but physically incorrect model.

The chosen governing expression is G_cl(s)=G_c G_p/(1+G_c G_p H). It is a compact model, not a complete plant specification. Terms on opposite sides of an equality must have compatible units and consistent time bases. When a parameter is estimated from one transient, its result belongs to that operating range unless new data demonstrate transferability. The task is to explain the meaning of the relation before substituting numbers and then to compare its predictions with a checkable boundary condition or steady state.

Working the numerical reference case

Synthetic worked example. For Gp=2/(10s+1), C=4 and H=1, Gcl=8/(10s+9). The closed-loop time constant is 10/9=1.11 s, while the DC setpoint gain is 8/9.

First transcribe every numerical input into a short table with its physical unit. Second, compute the baseline result without a software solver. Third, identify which output should change and which should remain unchanged when the manipulated input is zero. In this block diagrams and transfer functions exercise, the reported figures are not fitted plant data: they are an internally stated teaching benchmark. Any extension to different equipment requires new coefficients and new validation.

A useful sensitivity check is to change the selected coefficient or input by +10% while retaining all other conditions. This is not equivalent to assuming the output must also rise by 10%; exponential responses, ratios, saturation and feedback can change the mapping. Write the original expression, substitute the perturbed value and compare the sign of the result against physical intuition. Next repeat with a −10% change. An asymmetric response is evidence of nonlinearity in the selected mapping, not necessarily a numerical error.

Derivation and model restrictions

This nominal response assumes no transport delay and unbounded actuator authority; fast mathematical poles are not proof of fast equipment. To turn a physical inventory into a differential state equation, write accumulation as incoming rate minus outgoing rate plus net production, if relevant. To turn an instrument or actuator into a dynamical element, specify its output, time constant and initial condition. To build a transfer function, take deviations about a steady operating point and verify that the assumed initial deviations are zero. Do not apply frequency-domain algebra blindly to a strongly nonlinear or switching regime.

For block diagrams and transfer functions, the limiting checks should include a zero-change input, a long-time equilibrium where one exists and at least one infeasible input that the physical equipment cannot realize. If a parameter has dimensions of reciprocal time, use it only with a time expressed in matching units. If a parameter multiplies a temperature difference, its unit must convert the difference into a heat rate before the quantity is added to the energy balance. These simple checks detect mistakes that a graph renderer will not.

How to identify the model experimentally

Use the previously studied topic Negative feedback and closed-loop sensitivity as a source of prior knowledge and the next topic Characteristic equations and pole locations as an independent challenge. Define the reference operating point for block diagrams and transfer functions in the process case and record a small safe input sequence, the actual actuator trajectory, the measurement times and instrument range. A step, a pulse and a slow ramp probe different properties, so do not collapse them into one claimed validation. If the fitted output differs from the recorded trace, inspect the timestamp alignment and sensor dynamics before increasing model order.

Where possible, obtain at least two input amplitudes. Normalize the measured output by the change in input and test whether the curves approximately collapse. Failure to collapse can indicate nonlinearity, saturation, temperature-dependent coefficients or changed initial conditions. A second holdout operating point is especially important when the main formula was derived by linearization. Report the residual error and uncertainty rather than relabeling an imperfect approximation as a universal physical law.

Uncertainty, troubleshooting, and design decisions

Failure mode to investigate: Mistaking a sensor lag as plant lag can lead to tuning that overreacts to noisy or delayed measurements. The first diagnostic question is which conservation law, measurement chain or actuator path the proposed explanation implicates. Reconstruct the actual manipulated-variable history and compare it with what the controller requested; a large difference is a plant-side constraint, not a parameter-estimation error that can be fixed by fitting a more elaborate process transfer function.

Decision from the evidence: Label units on each block; transform at least one flow and one temperature signal into dimensionally compatible variables. State what the experiment supports and what remains unidentifiable. The decision should include whether the next useful action is to add a sensor, run a second input condition, change an assumed model structure or restrict the admissible operating range. When model outputs could guide actual equipment, independent engineering review and safety analysis are necessary before any implementation.

A challenge to the simplest explanation

Compare the present explanation with Bode response and stability margins, which appears elsewhere in this unit. Imagine obtaining a smooth measured response that both models could plausibly produce. Choose one additional measurement that would change differently under the two explanations. Define the expected sign and time scale under each. For block diagrams and transfer functions, record not just peak error but also the total actuator travel, disturbances, measurement noise and how the reference signal was generated. This exercise forces you to separate apparent goodness of fit from the ability to predict new conditions.

Numerical implementation and independent checks

The companion [interactive model, Laboratory 06](../labs/lab-06.html) is Proportional feedback and offset. Change one slider at a time, export the series and independently reproduce at least one value from a governing equation. The browser tool is deterministic and offline. It represents a defined educational model, not an approved process simulator. Its chart scale may rescale as parameters change, so compare numerical coordinates rather than visual slope alone.

For a numerical transient, verify step-size sensitivity by recomputing with half the time increment. If a predicted instability disappears merely when the integration time step shrinks, it was likely numerical rather than physical. On the other hand, an unstable physical model may diverge more quickly when more accurately integrated. Preserve initial conditions, parameter units and exact software version in the report so another student can repeat the test.

Practice and oral-examination questions

1. Starting from G_cl(s)=G_c G_p/(1+G_c G_p H), identify its state(s), manipulated input(s) and measured output(s). State how you would verify all dimensions.
2. Recalculate the example after a 10% increase of the principal process parameter. What changes in the predicted steady value and in the transient, and why are these not necessarily the same?
3. Propose a sensor or actuator failure that would make an otherwise correct block diagrams and transfer functions model appear wrong. What extra logged variable would distinguish that failure?
4. Explain how Bode response and stability margins might confound the identification of this lesson’s mechanism and propose a discriminating test.
5. State one safety or operating limit that the simplified G_cl(s)=G_c G_p/(1+G_c G_p H) relation does not capture. How should an engineer document that omission?

Answer guide and reflection

Question 1 is assessed by correct units and by separating the actual physical inventory from its measured signal; a memorized symbolic expression without those definitions is insufficient. For Question 2, the correct method uses the original equation twice and recomputes the equilibrium or time constant where applicable; simple proportional scaling is not accepted without showing that the equation is linear in the varied coefficient. For Question 3, the measurement or actuator diagnosis must lead to a different predicted trace. For Question 4, a proposed experiment is credible only if it can distinguish two candidate mechanisms. For Question 5, a constraint must be named and its omission explicitly stated.

Takeaway for Feedback, Transfer Functions, and Stability. The lesson is complete only after the model can be challenged with a changed initial condition, a disturbance and a non-ideal measurement. A plotted agreement under one nominal set of synthetic conditions is the start of investigation, not its end. The next step is to test whether the assumptions retain their meaning when characteristic equations and pole locations is included in the model.

Unit 3, Lesson 3: Characteristic equations and pole locations

Unit: Feedback, Transfer Functions, and Stability. Physical setting: temperature-loop performance in a continuous process unit. Reference model: 1+C(s)G(s)H(s)=0.

Learning outcomes

By the end of this lesson you should be able to (i) explain characteristic equations and pole locations in physical terms, (ii) identify the variables and assumptions of 1+C(s)G(s)H(s)=0, (iii) reproduce and challenge a numerical benchmark, (iv) recognize how measurement, delay and actuator limits modify the model, and (v) design a specific experiment that can distinguish the current model from an alternative interpretation.

Physical reasoning, not just a formula

Closed-loop stability is a property of pole locations, not a claim about the apparent smoothness of one plotted response. A pole with positive real part creates exponential growth in a linear continuous-time model.

Consider the unit-wide equipment case: temperature-loop performance in a continuous process unit. In this setting the same numerical trend may have more than one physical cause. A change observed at the analyzer may reflect a real process response, a changed flow residence time or the instrument dynamics. The analyst must first specify which quantity is manipulated, which is observed and which variables are treated as external disturbances. This is particularly important for characteristic equations and pole locations, where confusing a state with a measured output can give a mathematically solvable but physically incorrect model.

The chosen governing expression is 1+C(s)G(s)H(s)=0. It is a compact model, not a complete plant specification. Terms on opposite sides of an equality must have compatible units and consistent time bases. When a parameter is estimated from one transient, its result belongs to that operating range unless new data demonstrate transferability. The task is to explain the meaning of the relation before substituting numbers and then to compare its predictions with a checkable boundary condition or steady state.

Working the numerical reference case

Synthetic worked example. The polynomial s²+4s+3 factors as (s+1)(s+3), so both modes decay. Replacing the constant term by −3 instead yields roots −2+sqrt(7)≈0.646 and −2−sqrt(7)≈−4.646, demonstrating one unstable mode.

First transcribe every numerical input into a short table with its physical unit. Second, compute the baseline result without a software solver. Third, identify which output should change and which should remain unchanged when the manipulated input is zero. In this characteristic equations and pole locations exercise, the reported figures are not fitted plant data: they are an internally stated teaching benchmark. Any extension to different equipment requires new coefficients and new validation.

A useful sensitivity check is to change the selected coefficient or input by +10% while retaining all other conditions. This is not equivalent to assuming the output must also rise by 10%; exponential responses, ratios, saturation and feedback can change the mapping. Write the original expression, substitute the perturbed value and compare the sign of the result against physical intuition. Next repeat with a −10% change. An asymmetric response is evidence of nonlinearity in the selected mapping, not necessarily a numerical error.

Derivation and model restrictions

Linear stability around an operating point does not guarantee safety in the presence of hard stops, delays or nonlinear runaway. To turn a physical inventory into a differential state equation, write accumulation as incoming rate minus outgoing rate plus net production, if relevant. To turn an instrument or actuator into a dynamical element, specify its output, time constant and initial condition. To build a transfer function, take deviations about a steady operating point and verify that the assumed initial deviations are zero. Do not apply frequency-domain algebra blindly to a strongly nonlinear or switching regime.

For characteristic equations and pole locations, the limiting checks should include a zero-change input, a long-time equilibrium where one exists and at least one infeasible input that the physical equipment cannot realize. If a parameter has dimensions of reciprocal time, use it only with a time expressed in matching units. If a parameter multiplies a temperature difference, its unit must convert the difference into a heat rate before the quantity is added to the energy balance. These simple checks detect mistakes that a graph renderer will not.

How to identify the model experimentally

Use the previously studied topic Block diagrams and transfer functions as a source of prior knowledge and the next topic Bode response and stability margins as an independent challenge. Define the reference operating point for characteristic equations and pole locations in the process case and record a small safe input sequence, the actual actuator trajectory, the measurement times and instrument range. A step, a pulse and a slow ramp probe different properties, so do not collapse them into one claimed validation. If the fitted output differs from the recorded trace, inspect the timestamp alignment and sensor dynamics before increasing model order.

Where possible, obtain at least two input amplitudes. Normalize the measured output by the change in input and test whether the curves approximately collapse. Failure to collapse can indicate nonlinearity, saturation, temperature-dependent coefficients or changed initial conditions. A second holdout operating point is especially important when the main formula was derived by linearization. Report the residual error and uncertainty rather than relabeling an imperfect approximation as a universal physical law.

Uncertainty, troubleshooting, and design decisions

Failure mode to investigate: An unstable feedback loop may initially look quiet if the disturbance does not sufficiently excite the growing mode. The first diagnostic question is which conservation law, measurement chain or actuator path the proposed explanation implicates. Reconstruct the actual manipulated-variable history and compare it with what the controller requested; a large difference is a plant-side constraint, not a parameter-estimation error that can be fixed by fitting a more elaborate process transfer function.

Decision from the evidence: Check poles or Routh conditions and confirm by numerical response to more than one initial condition. State what the experiment supports and what remains unidentifiable. The decision should include whether the next useful action is to add a sensor, run a second input condition, change an assumed model structure or restrict the admissible operating range. When model outputs could guide actual equipment, independent engineering review and safety analysis are necessary before any implementation.

A challenge to the simplest explanation

Compare the present explanation with Negative feedback and closed-loop sensitivity, which appears elsewhere in this unit. Imagine obtaining a smooth measured response that both models could plausibly produce. Choose one additional measurement that would change differently under the two explanations. Define the expected sign and time scale under each. For characteristic equations and pole locations, record not just peak error but also the total actuator travel, disturbances, measurement noise and how the reference signal was generated. This exercise forces you to separate apparent goodness of fit from the ability to predict new conditions.

Numerical implementation and independent checks

The companion [interactive model, Laboratory 04](../labs/lab-04.html) is Second-order damping and overshoot. Change one slider at a time, export the series and independently reproduce at least one value from a governing equation. The browser tool is deterministic and offline. It represents a defined educational model, not an approved process simulator. Its chart scale may rescale as parameters change, so compare numerical coordinates rather than visual slope alone.

For a numerical transient, verify step-size sensitivity by recomputing with half the time increment. If a predicted instability disappears merely when the integration time step shrinks, it was likely numerical rather than physical. On the other hand, an unstable physical model may diverge more quickly when more accurately integrated. Preserve initial conditions, parameter units and exact software version in the report so another student can repeat the test.

Practice and oral-examination questions

1. Starting from 1+C(s)G(s)H(s)=0, identify its state(s), manipulated input(s) and measured output(s). State how you would verify all dimensions.
2. Recalculate the example after a 10% increase of the principal process parameter. What changes in the predicted steady value and in the transient, and why are these not necessarily the same?
3. Propose a sensor or actuator failure that would make an otherwise correct characteristic equations and pole locations model appear wrong. What extra logged variable would distinguish that failure?
4. Explain how Negative feedback and closed-loop sensitivity might confound the identification of this lesson’s mechanism and propose a discriminating test.
5. State one safety or operating limit that the simplified 1+C(s)G(s)H(s)=0 relation does not capture. How should an engineer document that omission?

Answer guide and reflection

Question 1 is assessed by correct units and by separating the actual physical inventory from its measured signal; a memorized symbolic expression without those definitions is insufficient. For Question 2, the correct method uses the original equation twice and recomputes the equilibrium or time constant where applicable; simple proportional scaling is not accepted without showing that the equation is linear in the varied coefficient. For Question 3, the measurement or actuator diagnosis must lead to a different predicted trace. For Question 4, a proposed experiment is credible only if it can distinguish two candidate mechanisms. For Question 5, a constraint must be named and its omission explicitly stated.

Takeaway for Feedback, Transfer Functions, and Stability. The lesson is complete only after the model can be challenged with a changed initial condition, a disturbance and a non-ideal measurement. A plotted agreement under one nominal set of synthetic conditions is the start of investigation, not its end. The next step is to test whether the assumptions retain their meaning when bode response and stability margins is included in the model.

Unit 3, Lesson 4: Bode response and stability margins

Unit: Feedback, Transfer Functions, and Stability. Physical setting: temperature-loop performance in a continuous process unit. Reference model: L(jw)=C(jw)G(jw)H(jw).

Learning outcomes

By the end of this lesson you should be able to (i) explain bode response and stability margins in physical terms, (ii) identify the variables and assumptions of L(jw)=C(jw)G(jw)H(jw), (iii) reproduce and challenge a numerical benchmark, (iv) recognize how measurement, delay and actuator limits modify the model, and (v) design a specific experiment that can distinguish the current model from an alternative interpretation.

Physical reasoning, not just a formula

Gain and phase margins quantify proximity to the −1 point in a nominal loop. They do not form a universal stability guarantee under arbitrary model mismatch, and margins must use the correct crossover frequencies.

Consider the unit-wide equipment case: temperature-loop performance in a continuous process unit. In this setting the same numerical trend may have more than one physical cause. A change observed at the analyzer may reflect a real process response, a changed flow residence time or the instrument dynamics. The analyst must first specify which quantity is manipulated, which is observed and which variables are treated as external disturbances. This is particularly important for bode response and stability margins, where confusing a state with a measured output can give a mathematically solvable but physically incorrect model.

The chosen governing expression is L(jw)=C(jw)G(jw)H(jw). It is a compact model, not a complete plant specification. Terms on opposite sides of an equality must have compatible units and consistent time bases. When a parameter is estimated from one transient, its result belongs to that operating range unless new data demonstrate transferability. The task is to explain the meaning of the relation before substituting numbers and then to compare its predictions with a checkable boundary condition or steady state.

Working the numerical reference case

Synthetic worked example. A pure 5 s delay contributes −omega*5 radians of phase. At omega=0.20 rad/s this is −1 rad=−57.3°, although its magnitude remains exactly one.

First transcribe every numerical input into a short table with its physical unit. Second, compute the baseline result without a software solver. Third, identify which output should change and which should remain unchanged when the manipulated input is zero. In this bode response and stability margins exercise, the reported figures are not fitted plant data: they are an internally stated teaching benchmark. Any extension to different equipment requires new coefficients and new validation.

A useful sensitivity check is to change the selected coefficient or input by +10% while retaining all other conditions. This is not equivalent to assuming the output must also rise by 10%; exponential responses, ratios, saturation and feedback can change the mapping. Write the original expression, substitute the perturbed value and compare the sign of the result against physical intuition. Next repeat with a −10% change. An asymmetric response is evidence of nonlinearity in the selected mapping, not necessarily a numerical error.

Derivation and model restrictions

A frequency-response model is valid only over the tested frequency region; sampled-data effects alter high-frequency behavior. To turn a physical inventory into a differential state equation, write accumulation as incoming rate minus outgoing rate plus net production, if relevant. To turn an instrument or actuator into a dynamical element, specify its output, time constant and initial condition. To build a transfer function, take deviations about a steady operating point and verify that the assumed initial deviations are zero. Do not apply frequency-domain algebra blindly to a strongly nonlinear or switching regime.

For bode response and stability margins, the limiting checks should include a zero-change input, a long-time equilibrium where one exists and at least one infeasible input that the physical equipment cannot realize. If a parameter has dimensions of reciprocal time, use it only with a time expressed in matching units. If a parameter multiplies a temperature difference, its unit must convert the difference into a heat rate before the quantity is added to the energy balance. These simple checks detect mistakes that a graph renderer will not.

How to identify the model experimentally

Use the previously studied topic Characteristic equations and pole locations as a source of prior knowledge and the next topic Bode response and stability margins as an independent challenge. Define the reference operating point for bode response and stability margins in the process case and record a small safe input sequence, the actual actuator trajectory, the measurement times and instrument range. A step, a pulse and a slow ramp probe different properties, so do not collapse them into one claimed validation. If the fitted output differs from the recorded trace, inspect the timestamp alignment and sensor dynamics before increasing model order.

Where possible, obtain at least two input amplitudes. Normalize the measured output by the change in input and test whether the curves approximately collapse. Failure to collapse can indicate nonlinearity, saturation, temperature-dependent coefficients or changed initial conditions. A second holdout operating point is especially important when the main formula was derived by linearization. Report the residual error and uncertainty rather than relabeling an imperfect approximation as a universal physical law.

Uncertainty, troubleshooting, and design decisions

Failure mode to investigate: Ignoring process delay can make an otherwise sensible controller design have far less phase margin in practice. The first diagnostic question is which conservation law, measurement chain or actuator path the proposed explanation implicates. Reconstruct the actual manipulated-variable history and compare it with what the controller requested; a large difference is a plant-side constraint, not a parameter-estimation error that can be fixed by fitting a more elaborate process transfer function.

Decision from the evidence: Overlay nominal and delayed loop phase curves, then identify how a crossover-frequency shift alters reserve stability. State what the experiment supports and what remains unidentifiable. The decision should include whether the next useful action is to add a sensor, run a second input condition, change an assumed model structure or restrict the admissible operating range. When model outputs could guide actual equipment, independent engineering review and safety analysis are necessary before any implementation.

A challenge to the simplest explanation

Compare the present explanation with Block diagrams and transfer functions, which appears elsewhere in this unit. Imagine obtaining a smooth measured response that both models could plausibly produce. Choose one additional measurement that would change differently under the two explanations. Define the expected sign and time scale under each. For bode response and stability margins, record not just peak error but also the total actuator travel, disturbances, measurement noise and how the reference signal was generated. This exercise forces you to separate apparent goodness of fit from the ability to predict new conditions.

Numerical implementation and independent checks

The companion [interactive model, Laboratory 05](../labs/lab-05.html) is Sinusoidal forcing and phase lag. Change one slider at a time, export the series and independently reproduce at least one value from a governing equation. The browser tool is deterministic and offline. It represents a defined educational model, not an approved process simulator. Its chart scale may rescale as parameters change, so compare numerical coordinates rather than visual slope alone.

For a numerical transient, verify step-size sensitivity by recomputing with half the time increment. If a predicted instability disappears merely when the integration time step shrinks, it was likely numerical rather than physical. On the other hand, an unstable physical model may diverge more quickly when more accurately integrated. Preserve initial conditions, parameter units and exact software version in the report so another student can repeat the test.

Practice and oral-examination questions

1. Starting from L(jw)=C(jw)G(jw)H(jw), identify its state(s), manipulated input(s) and measured output(s). State how you would verify all dimensions.
2. Recalculate the example after a 10% increase of the principal process parameter. What changes in the predicted steady value and in the transient, and why are these not necessarily the same?
3. Propose a sensor or actuator failure that would make an otherwise correct bode response and stability margins model appear wrong. What extra logged variable would distinguish that failure?
4. Explain how Block diagrams and transfer functions might confound the identification of this lesson’s mechanism and propose a discriminating test.
5. State one safety or operating limit that the simplified L(jw)=C(jw)G(jw)H(jw) relation does not capture. How should an engineer document that omission?

Answer guide and reflection

Question 1 is assessed by correct units and by separating the actual physical inventory from its measured signal; a memorized symbolic expression without those definitions is insufficient. For Question 2, the correct method uses the original equation twice and recomputes the equilibrium or time constant where applicable; simple proportional scaling is not accepted without showing that the equation is linear in the varied coefficient. For Question 3, the measurement or actuator diagnosis must lead to a different predicted trace. For Question 4, a proposed experiment is credible only if it can distinguish two candidate mechanisms. For Question 5, a constraint must be named and its omission explicitly stated.

Takeaway for Feedback, Transfer Functions, and Stability. The lesson is complete only after the model can be challenged with a changed initial condition, a disturbance and a non-ideal measurement. A plotted agreement under one nominal set of synthetic conditions is the start of investigation, not its end. The next step is to test whether the assumptions retain their meaning when bode response and stability margins is included in the model.

Unit 3 applied project: Feedback, Transfer Functions, and Stability

Decision problem

You are an analyst assigned to temperature-loop performance in a continuous process unit. A proposed controller or model change claims better performance. Your job is not to accept the claim but to define a reproducible counterfactual using the physical assumptions of this unit, quantify expected benefits, identify failure modes and explain what evidence would change the decision. Use deliberately synthesized teaching data unless real data are supplied with permission.

Required technical workflow

Stage 1 — Negative feedback and closed-loop sensitivity. Use Y/R = CG/(1+CG) to predict a measurable effect. Trace a positive sensor perturbation through controller, valve and plant and verify the predicted corrective direction. Explain how changing from heating to cooling equipment without verifying control action can create reinforcing rather than correcting feedback. Use the example conditions in the lesson only as a preliminary reference case; the final project should vary at least two different operating conditions so that conclusions do not depend on one curve.

Stage 2 — Block diagrams and transfer functions. Use G_cl(s)=G_c G_p/(1+G_c G_p H) to predict a measurable effect. Label units on each block; transform at least one flow and one temperature signal into dimensionally compatible variables. Explain how mistaking a sensor lag as plant lag can lead to tuning that overreacts to noisy or delayed measurements. Use the example conditions in the lesson only as a preliminary reference case; the final project should vary at least two different operating conditions so that conclusions do not depend on one curve.

Stage 3 — Characteristic equations and pole locations. Use 1+C(s)G(s)H(s)=0 to predict a measurable effect. Check poles or Routh conditions and confirm by numerical response to more than one initial condition. Explain how an unstable feedback loop may initially look quiet if the disturbance does not sufficiently excite the growing mode. Use the example conditions in the lesson only as a preliminary reference case; the final project should vary at least two different operating conditions so that conclusions do not depend on one curve.

Stage 4 — Bode response and stability margins. Use L(jw)=C(jw)G(jw)H(jw) to predict a measurable effect. Overlay nominal and delayed loop phase curves, then identify how a crossover-frequency shift alters reserve stability. Explain how ignoring process delay can make an otherwise sensible controller design have far less phase margin in practice. Use the example conditions in the lesson only as a preliminary reference case; the final project should vary at least two different operating conditions so that conclusions do not depend on one curve.

Evidence package

Provide a diagram with input, state, output and disturbance labels; a parameter dictionary with units and evidence status; the precise model equations; a csv or spreadsheet of numerical results; two labelled charts; and a structured comparison of nominal and perturbed cases. The input changes must be safe in the synthetic scenario and no physical equipment actuation is authorized by this assignment.

Model comparison and failure analysis

Compare a simple low-order model with a second model that adds one relevant limitation from feedback structure; block algebra; closed-loop poles; frequency margins. Keep the initial conditions and disturbance realization consistent. Calculate peak deviation, time to regain a chosen tolerance, integrated absolute error and the total excursion of the manipulated input when a controller is present. Explain any model disagreement using equation terms, not subjective plot descriptions.

Deliverables and evaluation

Submit a 3–5 page technical memo, runnable offline files or documented equations, a small data table and an audit statement. The recommended change must be limited to the modeled conditions. State at least two uncertainties, one additional validation experiment, and any safety question outside the authority of a classroom model. The grading priority is transparent reasoning and reproducibility, not a claimed industrial optimum.

Unit 3 assessment: Feedback, Transfer Functions, and Stability

Assessment type: open-book problem set. Total: 16 tasks plus one integrative evaluation. Show reasoning, units, model assumptions, and at least one verification check for each numerical answer. Data are illustrative rather than plant measurements.

Part 1: Negative feedback and closed-loop sensitivity

Question 1. Derive or interpret Y/R = CG/(1+CG). Define its coefficient units, the state variable, its initial condition and the operating point if linearized.

Question 2. Using the synthetic benchmark described in the corresponding lesson (For a static positive plant gain G=2 and proportional controller C=3, setpoint closed-loop gain is 6/7=0.857 and steady setpoint error is 1/7 of the commanded change.), repeat the arithmetic, then alter one explicitly stated input by +15% and explain the sign of the difference.

Question 3. Challenge the approximation: The simple fraction assumes stable linear elements and unity measurement dynamics; actuator saturation breaks superposition. Describe an experiment, instrument placement and observation that would invalidate this assumption.

Question 4. Troubleshoot the known risk: Changing from heating to cooling equipment without verifying control action can create reinforcing rather than correcting feedback. Propose a controlled test and a quantitative metric by which two possible causes can be separated.

Part 2: Block diagrams and transfer functions

Question 5. Derive or interpret G_cl(s)=G_c G_p/(1+G_c G_p H). Define its coefficient units, the state variable, its initial condition and the operating point if linearized.

Question 6. Using the synthetic benchmark described in the corresponding lesson (For Gp=2/(10s+1), C=4 and H=1, Gcl=8/(10s+9). The closed-loop time constant is 10/9=1.11 s, while the DC setpoint gain is 8/9.), repeat the arithmetic, then alter one explicitly stated input by +15% and explain the sign of the difference.

Question 7. Challenge the approximation: This nominal response assumes no transport delay and unbounded actuator authority; fast mathematical poles are not proof of fast equipment. Describe an experiment, instrument placement and observation that would invalidate this assumption.

Question 8. Troubleshoot the known risk: Mistaking a sensor lag as plant lag can lead to tuning that overreacts to noisy or delayed measurements. Propose a controlled test and a quantitative metric by which two possible causes can be separated.

Part 3: Characteristic equations and pole locations

Question 9. Derive or interpret 1+C(s)G(s)H(s)=0. Define its coefficient units, the state variable, its initial condition and the operating point if linearized.

Question 10. Using the synthetic benchmark described in the corresponding lesson (The polynomial s²+4s+3 factors as (s+1)(s+3), so both modes decay. Replacing the constant term by −3 instead yields roots −2+sqrt(7)≈0.646 and −2−sqrt(7)≈−4.646, demonstrating one unstable mode.), repeat the arithmetic, then alter one explicitly stated input by +15% and explain the sign of the difference.

Question 11. Challenge the approximation: Linear stability around an operating point does not guarantee safety in the presence of hard stops, delays or nonlinear runaway. Describe an experiment, instrument placement and observation that would invalidate this assumption.

Question 12. Troubleshoot the known risk: An unstable feedback loop may initially look quiet if the disturbance does not sufficiently excite the growing mode. Propose a controlled test and a quantitative metric by which two possible causes can be separated.

Part 4: Bode response and stability margins

Question 13. Derive or interpret L(jw)=C(jw)G(jw)H(jw). Define its coefficient units, the state variable, its initial condition and the operating point if linearized.

Question 14. Using the synthetic benchmark described in the corresponding lesson (A pure 5 s delay contributes −omega*5 radians of phase. At omega=0.20 rad/s this is −1 rad=−57.3°, although its magnitude remains exactly one.), repeat the arithmetic, then alter one explicitly stated input by +15% and explain the sign of the difference.

Question 15. Challenge the approximation: A frequency-response model is valid only over the tested frequency region; sampled-data effects alter high-frequency behavior. Describe an experiment, instrument placement and observation that would invalidate this assumption.

Question 16. Troubleshoot the known risk: Ignoring process delay can make an otherwise sensible controller design have far less phase margin in practice. Propose a controlled test and a quantitative metric by which two possible causes can be separated.

Integrative design problem

For the case of temperature-loop performance in a continuous process unit, select two lessons from this unit and build a consistent process-and-measurement model. Include one realistic actuator bound, one sensor imperfection and one disturbance. Explain the trade-off between fast response and robust operation. Present at least one table of simulated values and one limiting-value check. Label synthetic data distinctly from measured data.

Marking rubric

• 30% governing equations and physical interpretation; 20% internally consistent numerical reasoning; 20% model/assumption checks; 20% alternative hypothesis and experiment design; 10% clear figures, units, and communication.

Examiner notes

Answers are not unique. A valid derivation states a control volume, assumptions, and dimensions before substituting numbers. Parameter changes should be recomputed using the actual model rather than asserted by proportion. If the student finds that the available inputs do not identify a required quantity, a defensible bound and an appropriate additional measurement earn more credit than an invented precise number.