What Happens Inside a Micro Servo Motor When It Moves?

If you’ve ever built a robot, flown a drone, or tinkered with an RC car, you’ve almost certainly used a micro servo motor. These tiny, inexpensive actuators are the unsung heroes of the maker world—small enough to fit in the palm of your hand, yet capable of precise, repeatable motion that brings projects to life. But have you ever stopped to wonder what actually happens inside that little plastic case when you command it to move from 0° to 90°?

It looks simple from the outside: you send a signal, and the shaft turns. But beneath the surface, a fascinating chain reaction of electronics, mechanics, and control theory unfolds in milliseconds. In this deep dive, we’ll crack open the micro servo—figuratively and literally—and trace every step of that motion, from the moment your microcontroller sends a pulse to the moment the output shaft locks into position.

The Anatomy of a Micro Servo: More Than Just Gears

Before we can understand the motion, we need to understand the parts. A typical micro servo—like the ubiquitous SG90 or MG90S—contains four main subsystems working in tight harmony.

The DC Motor: The Muscle

At the core of every micro servo is a small, high-speed DC motor. These motors are typically 3V to 6V rated and can spin at 10,000 to 20,000 RPM with no load. That’s absurdly fast for a device that only needs to move 180 degrees. But here’s the trick: the motor isn’t directly connected to the output shaft. Instead, it’s geared down dramatically.

The motor itself is a standard brushed DC motor. It has a permanent magnet stator and a wound rotor (armature) with brushes that deliver current to the spinning commutator. When you apply power, current flows through the armature coils, creating a magnetic field that interacts with the stator’s field, generating torque. The direction of rotation depends on the polarity of the voltage—reverse the leads, and the motor spins the other way.

The Gear Train: The Translator

The motor’s high speed is useless for precise positioning. That’s where the gear train comes in. A micro servo typically uses a multi-stage planetary or spur gear reduction system. In an SG90, for example, you’ll find four or five plastic gears in series, each with a specific tooth count.

The total gear ratio is usually around 300:1 to 400:1. This means the motor must spin 300 to 400 times to turn the output shaft just once. This massive reduction does two things: it multiplies torque (giving the servo enough strength to lift a small control surface or robot arm) and it converts high-speed rotation into slow, controllable motion.

But there’s a hidden consequence. With such a high gear ratio, the output shaft has a lot of mechanical inertia and friction. If you try to force the shaft by hand, you’re essentially trying to spin the motor through the gears—which is why servos feel “stiff” when powered off.

The Potentiometer: The Feedback Eye

This is the component that makes a servo a servo. Attached to the output shaft (usually via a small plastic coupling) is a potentiometer—a variable resistor. As the shaft rotates, the wiper of the pot moves, changing the resistance between its terminals.

The potentiometer is wired as a voltage divider. One end is connected to a stable reference voltage (typically 5V), the other to ground, and the wiper outputs a voltage that’s proportional to the shaft’s angular position. If the servo has a 180° range, the pot might output 0.5V at 0° and 2.5V at 180°. This voltage is the servo’s “position sense” signal—its way of knowing exactly where it is at all times.

Not all micro servos use potentiometers. Some high-end or continuous-rotation servos use magnetic encoders or Hall effect sensors. But for the vast majority of hobbyist servos, it’s a simple, cheap, and surprisingly accurate 5kΩ or 10kΩ pot.

The Control Board: The Brain

The final piece is a small printed circuit board (PCB) embedded in the servo case. This board contains a few key components:

• A comparator or microcontroller: A dedicated IC that compares the incoming command signal to the feedback from the potentiometer.
• An H-bridge driver: A transistor circuit that can drive the DC motor in both directions.
• Supporting passives: Resistors, capacitors, and sometimes a voltage regulator.

This control board is the unsung hero. It’s what makes the servo “smart” enough to hold a position against external forces.

The Signal: How You Talk to a Servo

Every micro servo motion starts with a signal. The standard control method is Pulse Width Modulation (PWM) , but it’s not the same PWM used to dim an LED. Servo PWM uses a very specific timing protocol.

The Pulse Train Protocol

A servo expects a repeating pulse every 20 milliseconds (50 Hz). The pulse width—the duration the signal stays high—determines the target position. For a standard 180° servo:

• 1.0 ms pulse: 0° (full left)
• 1.5 ms pulse: 90° (center)
• 2.0 ms pulse: 180° (full right)

Between these extremes, the relationship is roughly linear. A 1.25 ms pulse might give 45°, and a 1.75 ms pulse might give 135°. The exact mapping varies between manufacturers, which is why you often need to calibrate servos in your code.

But here’s the critical detail: the servo doesn’t care about the duty cycle in the traditional sense. It only cares about the absolute pulse width. Even if the period changes slightly (say, 18 ms instead of 20 ms), the servo will still respond to the pulse width. This is why you can control servos with simple delayMicroseconds() calls on an Arduino, rather than needing a hardware PWM timer (though hardware PWM is more precise).

What the Control Board Sees

The control board’s input pin receives this PWM signal. Inside the board, a circuit measures the pulse width—usually by charging a capacitor through a resistor and comparing the voltage to a reference. More sophisticated servos use a dedicated servo decoder IC or a small microcontroller that directly measures the high time.

The measured pulse width is stored as a target position value. Simultaneously, the board reads the potentiometer voltage and converts it (via an analog-to-digital converter, or ADC) into a current position value. Now the board has two numbers: where you want to be and where you are.

The Closed-Loop Control: The Magic of Feedback

This is the heart of servo operation. The control board runs a closed-loop control system—specifically, a proportional controller (P-controller) in most cheap micro servos, though some use PI or PID.

The Error Signal

The board subtracts the current position from the target position. This difference is called the error signal.

Error = Target Position - Current Position

If the error is positive, the shaft needs to move clockwise. If negative, counterclockwise. If zero, the shaft is already at the target.

Driving the Motor

The error signal is amplified and fed to the H-bridge. But it’s not a simple on/off signal. The control board applies a voltage to the motor that’s proportional to the error. A large error (say, commanding 180° when the shaft is at 0°) results in full battery voltage being applied to the motor. The motor spins at maximum speed toward the target.

As the shaft approaches the target, the error shrinks. The control board reduces the motor voltage proportionally. This is why servos slow down as they near their destination—they’re not “braking”; they’re reducing drive power.

The Damping Problem

A pure proportional controller has a problem: overshoot. If the motor has momentum, it can swing past the target before the controller can react. The result is a buzzing, oscillating servo that never settles. To fix this, most servo control boards include a dead band.

The dead band is a small range of error (typically 1-2 degrees) where the controller does nothing. If the error is within this band, the motor is turned off. The servo relies on gear friction and static torque to hold position. This is why servos don’t oscillate in practice—they have a built-in “good enough” zone.

Some higher-end servos add derivative control (D in PID) to actively damp overshoot by opposing rapid changes in error. But in a $3 micro servo, you’re getting a simple P-controller with a dead band.

The Motion: Step by Step

Let’s walk through a real movement from start to finish. Imagine you command a servo to go from 0° to 90°.

Phase 1: Signal Reception (0-1 ms)

Your microcontroller sends a 1.5 ms pulse (for 90°). The control board captures this pulse and stores it as the target. The potentiometer reads 0.5V (0°). The error is +90°.

Phase 2: Full Acceleration (1-10 ms)

The control board sees a large positive error. It applies full forward voltage to the motor. The motor screams to life, spinning at thousands of RPM. The gear train multiplies torque but reduces speed. The output shaft begins to rotate toward 90°.

Phase 3: Constant Velocity (10-150 ms)

The shaft is moving at a steady speed. The error is decreasing linearly. The control board gradually reduces motor voltage as the error shrinks. The motor is still spinning, but slower. The potentiometer voltage rises smoothly.

Phase 4: Deceleration (150-200 ms)

The error is now small—maybe 10°. The control board is applying low voltage to the motor. The shaft is moving slowly. Gear friction is becoming significant relative to the drive torque.

Phase 5: Settling (200-250 ms)

The error enters the dead band. The control board cuts motor power. The shaft coasts to a stop, held in place by gear friction and the slight magnetic detent of the motor. The potentiometer reads 2.5V (90°). The error is zero (within tolerance). The servo is “holding” position.

Phase 6: Holding (250 ms onward)

The servo is now in a holding state. The control board continuously monitors the potentiometer. If an external force tries to push the shaft away from 90°, the error becomes non-zero. The board immediately applies corrective voltage to the motor to fight the force. This is why a powered servo feels “stiff” when you try to move it—it’s actively resisting you.

The Hidden Physics: What You Can’t See

While the electrical and control logic is fascinating, there’s a lot of physics happening inside that tiny plastic box that you might not notice.

Electrical Noise and Ground Loops

The DC motor inside the servo is a massive source of electrical noise. Brushed motors generate sparks at the commutator, which create wideband radio frequency interference (RFI). In a micro servo, this noise can radiate back through the power wires and affect your microcontroller. This is why you often see a capacitor (typically 10-100 µF) soldered across the servo’s power terminals on the control board—it’s a crude filter.

But there’s a more subtle issue. The motor’s current draw is not constant. When the servo starts moving, it can draw a surge of 500 mA to 1 A (for a typical SG90). This sudden current spike causes a voltage drop on the power supply line. If your microcontroller shares that supply, it can brown out and reset. This is why experienced makers always power servos from a separate regulator or battery.

Gear Backlash

No gear train is perfect. There’s always a tiny amount of play between gear teeth—called backlash. In a micro servo, backlash is usually 1-3 degrees. This means if you command the servo to 90° from the left, it might stop at 89.5°. If you command it from the right, it might stop at 90.5°. The difference is backlash.

This is not a defect; it’s a mechanical reality. Plastic gears (like in the SG90) have more backlash than metal gears (like in the MG90S) because plastic can deform slightly under load. For most applications, 1-2 degrees of hysteresis is acceptable. But for precise positioning (like a camera gimbal), backlash is a major concern.

Cogging Torque

The DC motor inside the servo has permanent magnets. As the rotor spins, the magnetic attraction between the rotor poles and the stator magnets creates a periodic variation in torque called cogging. You can feel this if you spin a powered-down servo by hand—it feels “notchy.”

When the servo is holding a position, cogging can cause tiny vibrations. The control board is constantly fighting these magnetic bumps. In cheap servos, this can manifest as a faint high-frequency hum or buzz, especially when the servo is holding a position that aligns with a magnetic pole.

Thermal Effects

Micro servos get hot. The motor’s copper windings have resistance, and current flow generates heat (I²R losses). The H-bridge transistors also dissipate heat. In a continuous-use scenario (like a robot walking), the internal temperature can rise 20-30°C above ambient.

Heat affects the potentiometer. The resistive element in the pot changes value with temperature. This means the servo’s position feedback drifts as it warms up. A servo that holds 90° when cold might drift to 92° after 10 minutes of operation. Higher-end servos compensate for this with temperature sensors or better potentiometers, but cheap micro servos simply drift.

Beyond Standard Micro Servos: Variations on a Theme

Not all micro servos follow the exact same internal logic. Some interesting variations exist.

Continuous Rotation Servos

A standard servo is limited to 180° or 270° of rotation. A continuous rotation servo (often called a “360° servo”) removes the mechanical stop on the output shaft and modifies the control logic. Instead of interpreting the potentiometer as an absolute position, it’s used to measure speed.

In a continuous rotation servo, the potentiometer is still present, but the control board treats the center position (1.5 ms pulse) as “stop.” A pulse shorter than 1.5 ms makes the servo spin one direction, with speed proportional to the pulse deviation. A longer pulse spins the other direction. The potentiometer provides feedback to maintain a constant speed, not a position. These are popular for simple robot drive bases.

Digital vs. Analog Servos

Most cheap micro servos are analog servos. They use a comparator IC and a simple analog control loop. The motor is driven with a continuous voltage proportional to the error. They’re simple, cheap, and have relatively slow response times.

Digital servos (like the popular MG90D or Tower Pro MG90S Digital) contain a small microcontroller that runs the control loop in software. They can update the motor drive signal at much higher frequencies (300 Hz vs. 50 Hz). This allows for faster response, higher holding torque, and smoother motion. However, they consume more power and can generate more electrical noise.

The difference is internal: the control board has a microcontroller instead of a comparator. The physical motor, gears, and potentiometer are often identical.

Coreless Motors

Some high-performance micro servos (like those used in FPV drone gimbals) use coreless DC motors. Instead of a wound iron rotor, coreless motors have a self-supporting copper winding with no iron core. This eliminates cogging torque and reduces rotor inertia dramatically. The result is buttery-smooth motion with instant acceleration.

But coreless motors are more expensive and less durable. They’re also more sensitive to voltage spikes. You won’t find them in $3 servos—they’re reserved for $30+ applications.

Practical Implications: What This Means for Your Projects

Understanding what happens inside a micro servo isn’t just academic. It has real consequences for how you design and debug your projects.

Power Supply Design

Given the current surges and noise, never power a servo directly from a microcontroller’s 5V pin. Use a separate regulator (like a 5V 2A module) or a dedicated battery. Add a large electrolytic capacitor (470 µF or more) near the servo’s power input to absorb transients.

Avoiding Brownouts

If your project has multiple servos moving simultaneously, the total current draw can exceed 3-4 amps. This will overwhelm most breadboard power supplies. Use a servo driver board (like the PCA9685) that has its own power input and can handle the load.

Mechanical Loading

The gear train is the weak point. Micro servos have plastic gears that strip easily under shock loads. If your robot arm catches on something, the servo will try to force through it, potentially stripping gears. Use metal gear servos (like the MG90S) for applications with high torque or impact risk.

Signal Integrity

Long signal wires (more than 12 inches) can pick up noise and cause erratic servo behavior. If you need long runs, use twisted pair wire or add a 100Ω resistor in series with the signal line at the servo end to damp reflections.

Calibration

Because potentiometers and pulse width mapping vary, always calibrate your servos in software. Measure the actual pulse width that gives 0° and 180° for each servo, and use map() functions to convert your desired angle to the correct pulse width.

The Final Frame: A System in Miniature

When you send that 1.5 ms pulse to a micro servo, you’re not just “turning a motor.” You’re initiating a symphony of electrical, mechanical, and control processes that happen in less time than it takes to blink. The pulse is measured, compared to a feedback voltage, amplified, and converted into torque. That torque is reduced by a factor of 300 through precision-cut gears, overcoming friction and inertia to move a shaft to within a degree of your command. And then it holds that position against gravity, vibration, and external forces, constantly correcting in a silent, invisible loop.

All of this happens inside a package smaller than a matchbox, powered by a single wire carrying a signal that’s nothing more than a carefully timed on-off pulse. It’s a testament to how far miniaturization and control theory have come—and a reminder that even the simplest components in our projects are engineering marvels in their own right.

The next time you see a tiny robot arm swing into position or a RC car’s steering servo twitch, remember: there’s a lot more going on inside than meets the eye.