7.1 Open-Loop and Closed-Loop Control Architecture

7.1 Open-Loop and Closed-Loop Control Architecture

In airborne electrical and avionic systems, precision control of mechanical mechanisms is paramount to flight safety and navigational accuracy. Flight control computers, automatic flight guidance systems, and engine electronic controls must command physical displacements—such as deflecting an aileron, modulating a turbine fuel metering valve, or stabilizing a weather radar dish—with exceptional repeatability under fluctuating aerodynamic loads and severe environmental vibrations. Under European Aviation Safety Agency (EASA) Part-66 Module 04, aircraft maintenance engineers must master the architectural foundations, operating principles, error detection methods, and dynamic responses that distinguish basic open-loop actuation from closed-loop servomechanisms.

---

Fundamental Principles: Open-Loop vs. Closed-Loop Systems

Control systems are broadly categorized into two fundamental architectures based on the presence or absence of an active measurement feedback loop:

1. Open-Loop Control Systems

An open-loop control system operates strictly on a feedforward basis: an input command is applied to an actuator or controller, which alters the output state without measuring, monitoring, or verifying the actual physical result. There is no feedback path connecting the controlled output back to the input command.

graph LR
    IN["Input Demand<br/>(Command θ_i)"] --> AMP["Controller /<br/>Amplifier"]
    AMP --> ACT["Actuator /<br/>Motor"]
    ACT --> LOAD["Mechanical Output<br/>(Load Position θ_o)"]
    DIST["External Disturbance<br/>(Friction, Air Loads)"] -.-> LOAD

• Operational Characteristics: The output displacement is entirely dependent on the initial calibration and steady-state transfer characteristics of the forward path components.
• Vulnerabilities: Open-loop systems are fundamentally unmonitored. If an external disturbance occurs—such as increased aerodynamic hinge drag on a flap, mechanical stiction in a bearing, thermal expansion of control rods, or a sag in the aircraft DC bus voltage—the actual output deviates from the desired state without the system recognizing or correcting the error.
• Aircraft Applications: Limited to non-critical auxiliary functions where positional precision is secondary and external load variations are negligible, such as cockpit ventilation fan blowers, manual windshield wiper speed switches, open-loop emergency hydraulic dump valves, and raw rheostat cabin accent lighting.

2. Closed-Loop Control Systems (Servomechanisms)

A closed-loop control system incorporates an active feedback path that continuously measures the actual output state and feeds a proportional signal back to a comparator. The comparator algebraically subtracts the feedback signal from the input demand, generating an error signal ($\epsilon$) that drives the actuator in the direction necessary to eliminate the error.

graph LR
    IN["Input Demand (θ_i)"] --> COMP{"Comparator /<br/>Error Detector"}
    COMP -->|"Error Signal (ε)"| AMP["Error Amplifier &<br/>Compensator"]
    AMP --> ACT["Actuator /<br/>Servomotor"]
    ACT --> LOAD["Mechanical Output<br/>(Load θ_o)"]
    LOAD --> FB["Feedback Transducer<br/>(H)"]
    FB -->|"Feedback Signal (H·θ_o)"| COMP
    DIST["Disturbance (T_d)"] -.-> LOAD

• Definition of a Servomechanism: Per EASA Part-66 specifications, a servomechanism (often abbreviated as servo) is defined specifically as an automatic, closed-loop, power-amplifying control system in which the controlled output variable is mechanical position, velocity, or acceleration.
• Power Amplification: A vital hallmark of the servomechanism is power amplification. A low-energy electrical input command (e.g., a few millivolts or microamperes from a flight management computer or digital autopilot processor) is amplified to command high-power mechanical outputs (delivering hundreds of newton-meters of torque or kilonewtons of linear force) to displace massive flight control surfaces.
• Inherent Self-Correction: Closed-loop servomechanisms continuously compensate for internal component drift, environmental temperature variations, supply voltage transients, and external aerodynamic disturbance loads.

---

Essential Architecture of a Closed-Loop Servomechanism

A standard aviation position servomechanism consists of seven interconnected functional building blocks:

1. Command / Demand Input Element ($\theta_i$ or $R(s)$): The device that generates the reference setpoint representing the desired mechanical position. In an aircraft, this may be an analog DC voltage from a pilot's side-stick displacement transducer, a 400 Hz AC synchro voltage from an attitude gyro, or a digital word transmitted across an ARINC 429 or AFDX avionics data bus from the flight guidance computer.
2. Error Detector / Comparator / Summing Junction: The differential element that algebraically compares the input demand signal ($\theta_i$) with the feedback signal ($H \theta_o$). It outputs an electrical error signal ($\epsilon$): $\epsilon = \theta_i - H \theta_o$ In analog DC servos, this is implemented using an operational amplifier differential subtractor; in 400 Hz AC synchro systems, it is formed by a Control Transformer (CT); in modern fly-by-wire flight control computers (FCC), it is computed digitally.
3. Error Amplifier: The raw error signal produced by the comparator is typically in the millivolt or microampere range—insufficient to drive an electro-mechanical prime mover. The error amplifier provides high voltage and current gain ($K_a$), scaling the error signal into a robust drive voltage.
4. Controller / Compensator Network: An active electronic filtering network placed in the forward loop (or embedded within the amplifier) to shape the system's dynamic response. It provides proportional, integral, and derivative (PID) processing or phase-lead/phase-lag compensation to ensure fast rise time, zero steady-state error, and adequate stability phase margins without oscillatory overshoot.
5. Actuator / Servomotor: The electro-mechanical, electro-hydraulic, or electro-hydrostatic power converter that transforms amplified electrical energy into mechanical motion. Common aircraft actuators include two-phase AC induction servomotors, brushless DC torque motors, and Electro-Hydraulic Servo Valves (EHSVs) coupled to hydraulic rams.
6. Reduction Gearbox & Mechanical Load ($\theta_o$ or $C(s)$): The physical flight control surface (elevator, aileron, rudder) or mechanism being positioned. Because servomotors operate most efficiently at high rotational velocities ($3,000\text{ to } 10,000\text{ RPM}$) with low torque, a precision step-down reduction gear train converts this high-speed/low-torque output into the low-speed/high-torque motion required to overcome flight surface hinge moments ($T_d$), moment of inertia ($J$), and bearing friction ($B$).
7. Feedback Path Transducer ($H$): A precision sensing device mechanically coupled to the output load (or motor drive shaft) that converts actual physical position into a proportional electrical signal. The transducer's transfer function ($H$) matches the physical unit, voltage scale, and carrier frequency of the input demand element, enabling direct algebraic subtraction at the comparator.

---

Follow-Up Systems & The Defining Null-Seeking Action

In aerospace literature, position servomechanisms are universally described as follow-up systems. In a follow-up system, the mechanical output position is slaved to track the input command displacement with high fidelity across both static setpoints and dynamic trajectories.

The Null-Seeking Operating Principle

The foundational operating principle governing all closed-loop servomechanisms is null-seeking action. This automatic closed-loop cycle proceeds through five distinct operational states:

graph TD
    S1["1. Static Equilibrium (Null): θ_o matches θ_i; ε = 0; Motor power is ZERO."] --> S2["2. Input Demand Change: Pilot / Autopilot alters θ_i; ε = θ_i - H·θ_o ≠ 0."]
    S2 --> S3["3. Drive Signal Generation: Error amplifier amplifies ε to command actuator."]
    S3 --> S4["4. Mechanical Actuation: Servomotor rotates, driving load θ_o toward θ_i."]
    S4 --> S5["5. Null Restoration: As θ_o approaches θ_i, ε decreases toward zero."]
    S5 --> S1

1. Null State: When the output position exactly matches the commanded demand ($\theta_o = \theta_i / H$), the feedback signal cancels the input signal at the comparator. The error voltage is precisely zero ($\epsilon = 0$). With zero error input, the error amplifier outputs zero drive voltage to the servomotor. The actuator remains stationary, holding the flight surface at rest.
2. Displacement off Null: When the pilot or autopilot introduces a new demand (e.g., pitching the elevator up by $+5^\circ$), an instantaneous difference appears between $\theta_i$ and the feedback signal. The comparator instantly produces a non-zero error voltage ($\epsilon > 0$).
3. Drive Phase: The error amplifier magnifies $\epsilon$, applying a high-voltage drive signal to the servomotor. The polarity (in DC systems) or phase angle (in 400 Hz AC systems) of the amplified error determines the motor's direction of rotation.
4. Follow-Up Motion: The servomotor rotates, driving the flight surface toward the commanded $+5^\circ$ deflection via the reduction gearbox. Simultaneously, the feedback transducer rotates, causing the feedback voltage to climb.
5. Null Convergence: As the control surface approaches $+5^\circ$, the difference between command and feedback shrinks, causing $\epsilon$ to attenuate proportionally. Consequently, motor drive torque decelerates smoothly. When the surface reaches exactly $+5^\circ$, the error returns to zero ($\epsilon = 0$), drive power is cut, and the servomotor stops cleanly at the commanded position.

If an external aerodynamic gust knocks the elevator off its $+5^\circ$ commanded position, the feedback transducer immediately registers the displacement. An opposing error signal is instantly generated at the comparator without any pilot intervention, driving the actuator to restore the surface to its $+5^\circ$ null position.

---

Critical Aircraft Servomechanism Applications

Closed-loop servomechanisms are deployed throughout modern transport category aircraft across mechanical, hydraulic, and avionics disciplines:

1. Autopilot Primary Flight Control Surface Actuators: Autopilot computers drive electro-mechanical or electro-hydraulic servo actuators coupled to the elevator, aileron, and rudder control runs. In fly-by-wire (FBW) aircraft, Actuator Control Electronics (ACE) units continuously close the loop around hydraulic main control valves using Linear Variable Differential Transformers (LVDTs).
2. Autothrottle / Autothrust Cable Drives: Dedicated autothrottle servomotors drive the cockpit throttle thrust levers and engine fuel control linkages via capstans, clutches, and cable loops, maintaining target calibrated airspeed (CAS) or Mach number commanded by the Flight Management System (FMS).
3. Engine Bleed Air & Environmental Control Valves: High-temperature modulating valves in the pneumatic system position butterfly gates to regulate 5th-stage and 9th-stage compressor bleed air pressure and temperature before entering the air conditioning packs.
4. Cockpit Weather Radar Antenna Stabilization: Weather radar antenna dishes must remain locked to the horizon regardless of aircraft pitch and roll attitudes during turbulent climbs, descents, and bank maneuvers. Fast-acting closed-loop servos driven by attitude gyro signals continuously torque the antenna gimbal platform in pitch (elevation) and roll (tilt).
5. Fly-by-Wire Nose-Wheel Steering: Cockpit steering tiller inputs and rudder pedal deflections are compared against nose-gear steering angle feedback sensors (rotary variable differential transformers) to modulate dual hydraulic steering actuators during ground taxi operations.

---

Open-Loop versus Closed-Loop Comparative Engineering Matrix

---

Maintenance Safety & Operational Callouts

[!NOTE] Dynamic Tracking Error (Follow-Up Lag): In a Type 0 proportional position servomechanism, tracking a continuously moving input ramp (such as a constant pitch rate command) requires a sustained, non-zero error signal ($\epsilon$) to keep the servomotor spinning against viscous friction. This necessary dynamic offset between demanded position and actual position is called follow-up lag or velocity error. Avionic flight computers eliminate this lag by incorporating integral action (PID controllers) or feedforward velocity compensation.

[!WARNING] Feedback Disconnect Hazard (Actuator Hard-Over Runaway): If a mechanical linkage fracture, loose wiring terminal, or open-circuit failure severs the feedback path transducer signal ($H\theta_o \to 0$) in an active flight control servomechanism, the comparator will perceive that the output has never moved. The full input demand is interpreted as a persistent, massive error signal ($\epsilon = \theta_i$). The error amplifier will supply continuous maximum drive power to the servomotor, driving the flight control surface violently into its mechanical limit stops—a catastrophic flight condition known as actuator hard-over. Modern avionics incorporate dual-redundant or triplex feedback channels with independent cross-monitoring to isolate severed feedback paths and immediately disengage the servo clutch.

---

Worked Engineering Calculations

Calculation 1: Closed-Loop Transfer Function and Loop Gain

An autopilot elevator position servomechanism has the forward loop transfer function $G(s)$ and a unity feedback path ($H = 1.0\text{ V/V}$):

• Preamplifier and error amplifier forward voltage gain: $K_a = 40.0\text{ V/V}$
• Servomotor torque and velocity transfer function: $G_m(s) = \frac{K_m}{s(J s + B)}$
  • Motor constant: $K_m = 2.50\text{ rad/(V}\cdot\text{s)}$
  • Equivalent load inertia: $J = 0.05\text{ kg}\cdot\text{m}^2$
  • Equivalent viscous friction: $B = 1.25\text{ N}\cdot\text{m}\cdot\text{s/rad}$

graph LR
    R["Demand θ_i(s)"] --> SUM{"+\n-"}
    SUM -->|"ε(s)"| G["G(s) = K_a · G_m(s)"]
    G --> C["Output θ_o(s)"]
    C -->|H = 1.0| SUM

Step 1: Derive the Forward Path Gain ($K_{fwd}$)

$K_{fwd} = K_a \times K_m = 40.0\text{ V/V} \times 2.50\text{ rad/(V}\cdot\text{s)} = 100.0\text{ s}^{-1}$ $G(s) = \frac{K_{fwd}}{s(J s + B)} = \frac{100.0}{s(0.05 s + 1.25)}$

Step 2: Formulate the Closed-Loop Transfer Function ($T(s) = \frac{\theta_o(s)}{\theta_i(s)}$)

$T(s) = \frac{G(s)}{1 + G(s) H(s)} = \frac{\frac{100.0}{0.05 s^2 + 1.25 s}}{1 + \frac{100.0}{0.05 s^2 + 1.25 s}} = \frac{100.0}{0.05 s^2 + 1.25 s + 100.0}$

Step 3: Standardize to Second-Order Canonical Form ($T(s) = \frac{\omega_n^2}{s^2 + 2\zeta\omega_n s + \omega_n^2}$)

Divide numerator and denominator by the leading coefficient ($0.05$): $T(s) = \frac{\frac{100.0}{0.05}}{s^2 + \frac{1.25}{0.05} s + \frac{100.0}{0.05}} = \frac{2000}{s^2 + 25 s + 2000}$

Step 4: Calculate System Natural Frequency ($\omega_n$) and Damping Ratio ($\zeta$)

$\omega_n^2 = 2000 \implies \omega_n = \sqrt{2000} \approx 44.72\text{ rad/s}$ $2 \zeta \omega_n = 25 \implies \zeta = \frac{25}{2 \omega_n} = \frac{25}{2 \times 44.72} = \frac{25}{89.44} \approx 0.280$

Avionics Engineering Assessment: With a damping ratio of $\zeta \approx 0.28$, this position servo is significantly underdamped and will exhibit substantial overshoot ($> 40%$) in response to a step command, requiring rate feedback damping compensation to achieve the industry standard $\zeta \approx 0.707$.

---

Calculation 2: Steady-State Positioning Error under Aerodynamic Disturbance Torque

During high-speed cruise, a continuous aerodynamic hinge moment (disturbance torque $T_d$) acts on the elevator surface, opposing the actuator output.

• Total forward loop proportional stiffness gain: $K_{loop} = 80.0\text{ N}\cdot\text{m/rad}$
• Aerodynamic disturbance torque: $T_d = 6.40\text{ N}\cdot\text{m}$

In a proportional closed-loop servomechanism, the steady-state position error ($\epsilon_{ss}$) produced by an external load disturbance is given by: $\epsilon_{ss} = \frac{T_d}{K_{loop}}$

Step 1: Calculate Steady-State Angular Error in Radians

$\epsilon_{ss} = \frac{6.40\text{ N}\cdot\text{m}}{80.0\text{ N}\cdot\text{m/rad}} = 0.080\text{ rad}$

Step 2: Convert Angular Error to Degrees

$\epsilon_{ss}(^\circ) = 0.080\text{ rad} \times \left( \frac{180^\circ}{\pi} \right) = 0.080 \times 57.2958^\circ \approx 4.58^\circ$

Step 3: Evaluate Effect of Increasing Forward Gain ($K_a$) by a Factor of 5

If the flight control computer increases the amplifier gain such that $K_{loop}' = 400.0\text{ N}\cdot\text{m/rad}$: $\epsilon_{ss}' = \frac{6.40\text{ N}\cdot\text{m}}{400.0\text{ N}\cdot\text{m/rad}} = 0.016\text{ rad} \approx 0.92^\circ$

Technical Conclusion: Increasing forward loop gain reduces steady-state position error caused by aerodynamic blow-back from $4.58^\circ$ down to $0.92^\circ$, demonstrating that high loop stiffness is essential for precise flight control surface positioning, provided adequate damping is added to prevent loop instability.