A closed-loop system is a control circuit that continuously measures its actual output and feeds that data back to adjust its input, automatically correcting any errors to maintain a target setpoint. If you are searching for what is closed loop in the context of electrical engineering and electronics, you are looking at control theory, not basic continuity. Before we break down the math and the hardware, we need to clear up the single most common point of confusion for beginners: a "closed loop" is not the same thing as a "closed circuit." A closed circuit simply means a switch is turned on and current can flow from the power source to the load. A closed loop means information flows backward from the output to the input to regulate the system.
Open-Loop vs. Closed-Loop Systems
To understand what a closed loop changes in a real installation, you have to compare it to an open-loop system. In an open-loop system, the controller sends a fixed command and blindly hopes the output matches. If a disturbance occurs (like a mechanical load increasing or input voltage sagging), the open-loop system has no way to know the output failed, and it takes no corrective action. A closed-loop system uses sensors and an error amplifier to constantly compare the actual output against the desired setpoint, dynamically adjusting the drive signal to eliminate the error.
| Criteria | Open-Loop System | Closed-Loop System |
|---|---|---|
| Feedback Path | None (unidirectional signal flow) | Present (output measured and fed back to input) |
| Disturbance Rejection | Poor (output drifts with load/temp changes) | Excellent (automatically compensates for disturbances) |
| Component Count | Low (just the driver and power stage) | High (adds sensors, ADC, error amplifiers, compensators) |
| Stability Risk | Unconditionally stable (cannot oscillate) | Conditionally stable (poor tuning causes oscillation/ringing) |
| Typical Hardware | Basic 555 timer PWM motor driver, space heater | ODrive BLDC controller, PID temperature controller, SMPS |
As shown in the table above, the trade-off for the precision of a closed loop is complexity and the risk of instability. If the feedback loop reacts too slowly, the system underperforms; if it reacts too aggressively, it will overshoot the target and oscillate wildly—a concept deeply tied to phase margin and loop compensation.
How Feedback Changes a Real Circuit: A Numeric Example
Let’s look at exactly what closed-loop feedback does in a physical circuit by examining a DC-DC buck converter, such as the ubiquitous LM2596 or a modern equivalent like the TPS5430. The goal of the circuit is to step down a 12V nominal battery voltage to a strict 5.0V logic rail for a microcontroller.
The Baseline (Open-Loop Calculation): In a theoretical, lossless buck converter, the required PWM duty cycle ($D$) to achieve 5V from 12V is simply $V_{out} / V_{in}$. Therefore, $D = 5.0 / 12.0 = 41.6\%$. If the controller just output a fixed 41.6% duty cycle, it would be running open-loop.
Now, let's introduce a real-world disturbance. The microcontroller wakes up a high-power WiFi module, and the load current instantly steps from 0.5A to 3.0A. In a real power stage, the inductor has a DC resistance (DCR) of roughly $50m\Omega$, and the high-side MOSFET has an $R_{DS(on)}$ of $100m\Omega$.
When the current spikes by 2.5A, the parasitic resistances cause an internal voltage drop ($V = I \times R$) of about $0.375V$. Furthermore, the input battery voltage sags slightly under the sudden load. Without feedback, the output voltage would droop to roughly 4.6V, potentially causing a brownout reset on the microcontroller.
The Closed-Loop Correction:
The buck converter uses a resistor divider network (e.g., $R1 = 10k\Omega$, $R2 = 3.16k\Omega$) to scale the 5.0V output down to the internal 1.221V reference voltage of the error amplifier. When the output sags to 4.6V, the feedback pin sees only 1.12V instead of 1.221V. The error amplifier detects this $0.1V$ deficit and immediately commands the PWM controller to increase the duty cycle from 41.6% to approximately 44.5%. This wider pulse width forces more energy through the inductor, overcoming the parasitic voltage drops and pulling the output back to exactly 5.0V, typically within a few hundred microseconds. This dynamic, automatic adjustment is the core value of closed-loop control.
Where You Meet Closed Loops in Practice
You are likely already using closed-loop systems daily on the bench or in the field, even if you don't actively think about the control theory behind them. Here is where this topology dominates modern electrical and electronic design:
- Switch-Mode Power Supplies (SMPS): Every modern phone charger, PC power supply, and LED driver uses closed-loop voltage or current regulation. Without it, output voltage would swing wildly with changes in wall voltage or device load.
- Variable Frequency Drives (VFDs) and Servo Motors: When you need a 3-phase AC motor to hold a precise position or speed under varying mechanical loads, the VFD uses closed-loop vector control. It reads the rotor position via encoders or Hall-effect sensors and dynamically adjusts the stator current phase angle and frequency.
- Maximum Power Point Tracking (MPPT): Advanced solar charge controllers use a closed loop to constantly perturb the array voltage and measure the resulting current. The algorithm continuously hunts for the exact knee of the I-V curve where power output ($V \times I$) is maximized, adjusting for passing clouds and temperature shifts.
- PID Temperature Controllers: Whether you are tuning a 3D printer hotend or running an industrial kiln, a PID (Proportional-Integral-Derivative) controller uses closed-loop feedback from a thermocouple to modulate a solid-state relay, preventing the massive temperature overshoot you would get from a simple mechanical thermostat.
Common Confusions and Troubleshooting
When working with closed-loop systems, misunderstanding the terminology or the tuning parameters will lead to frustrating hardware behavior. Keep these distinctions and troubleshooting paths in mind.
"Closed Loop" vs. "Closed Circuit" vs. "Ground Loop"
As established, a closed circuit just means current has a complete path to flow. A ground loop is an unwanted, often noisy, alternate path for current to flow between two grounded points due to a potential difference. A closed loop (in control theory) strictly refers to the informational feedback path. If an inspector or datasheet mentions "closed-loop testing," they are referring to testing a system while its automated feedback regulation is active, as opposed to open-loop bench testing where you force specific inputs manually.
Troubleshooting Loop Oscillation (Hunting)
If your closed-loop system is "hunting"—meaning the output continuously overshoots and undershoots the target in a sine-wave pattern—your loop gain is too high or your phase margin is too low.
Diagnostic Step: Hook up an oscilloscope to the output rail or motor speed tachometer. If you see a continuous, fixed-frequency ringing after a step-load change, the system is marginally stable or unstable. To fix this in an analog SMPS, you must adjust the Type II or Type III compensation network (usually by increasing the compensation capacitor to roll off the high-frequency gain). In a digital PID system, you must reduce the Proportional (P) and Derivative (D) gains until the ringing stops, then slowly increase the Integral (I) gain to eliminate steady-state error.
Understanding what a closed loop is—and more importantly, how to measure and tune its feedback network—is what separates a technician who can only swap out broken modules from an engineer who can design, debug, and optimize power and motion systems from scratch.
