Quiet Servo Motors: How to Select Low-Noise Servo Motors for Precision Automation

Noise has quietly become a specification. Medical device makers, semiconductor equipment builders, and packaging OEMs now arrive at the design review with a dB(A) target alongside torque and settling time — and the motion axis is usually the first thing they point at.

This guide is written for engineers and buyers specifying industrial AC and brushless servo systems, covering small to medium power ranges commonly used in precision automation equipment. It is not about RC hobby servos: the plastic-gear, three-wire units discussed on maker forums fail for entirely different reasons, and the advice there (regreasing gear trains, changing PWM refresh rates) does not transfer to a closed-loop industrial axis.

What follows is how noise is actually generated in a servo system, what a realistic silent servo motor target looks like in decibels, which datasheet parameters predict quiet running, and how to write the requirement so a supplier can be held to it.

What Makes a Servo Motor "Quiet"? Acoustic Noise vs. Electrical Noise

The single biggest source of confusion in this topic is that the industry uses "servo motor noise" for two unrelated phenomena.

Both matter, but they are separate engineering problems with separate solutions. This article addresses acoustic noise only. If your symptom is dropped encoder counts rather than an audible tone, you are looking for an EMI article, not this one.

Acoustic energy reaches the operator by three paths, and the third one is where most machines lose:

• Airborne — the motor housing radiates sound directly.
• Structure-borne — vibration conducts into the mounting plate, frame, or enclosure panel, which then radiates as a much larger surface area. A small motor mounted on a flexible sheet-metal bracket can produce higher system-level noise because the structure amplifies vibration and radiates airborne sound.
• Resonance amplification — a modest excitation at a structural natural frequency produces a disproportionately loud tone.

The practical consequence: there is no such thing as a quiet servo motor in isolation. There are quiet servo systems — motor, drive, feedback device, transmission, and mechanical mounting, considered together. Buying a low-cogging motor and bolting it to a resonant frame through a rigid coupling will not deliver a quiet machine.

Why Acoustic Noise Matters Beyond Operator Comfort

Four reasons noise earns engineering attention rather than just an apology:

1. Noise is unwanted vibration made audible. Some sources of audible noise, such as torque ripple and mechanical vibration, are also disturbances that can affect settling time, surface finish, and repeatability.
2. Noise is a leading fault indicator. A change in acoustic signature usually precedes a bearing failure by weeks or months, which is why acoustic monitoring appears in predictive maintenance programs.
3. Noise is an acceptance criterion. Machines destined for hospitals, laboratories, and cleanrooms are increasingly tested against a contractual dB(A) limit at handover.
4. Noise is regulated. Where operators work near the machine for a full shift, occupational exposure limits apply — and the machine builder inherits the problem.

Where Servo Motor Noise Comes From: Four Root-Cause Categories

Effective troubleshooting starts with correct attribution. In practice, many noise complaints attributed to the servo motor originate from transmission components or control tuning rather than the motor alone.e motor itself.

Electromagnetic Noise: Torque Ripple, Cogging, and PWM Switching

Inside the motor, radial magnetic forces act between the rotor magnets and the stator teeth. These forces deform the stator slightly, and the housing radiates the resulting vibration as sound. Several mechanisms drive this:

• Cogging torque — caused by the interaction between rotor magnets and stator slots, creating periodic torque ripple as the rotor moves through different positions.
• Back-EMF distortion — if the motor's back-EMF waveform is not truly sinusoidal, even a perfect sine-wave drive produces torque ripple. This is a motor design property that no amount of tuning fully removes.
• PWM switching excitation — the drive's carrier frequency creates ripple current, which creates ripple force at the carrier frequency and its sidebands. PWM carrier frequencies in the audible range can excite electromagnetic forces and produce the familiar servo whine.

Mechanical Noise: Bearings, Imbalance, and Runout

• Bearings dominate at moderate to high speed. Bearing grade, preload, grease type and fill quantity, and cage design all contribute. Excessive preload produces a tonal whine; a degraded raceway produces broadband growl or squeal.
• Rotor imbalance appears as a tone at rotational frequency and grows roughly with the square of speed. Precision servo motors may use tighter rotor balancing requirements, with grades selected according to speed and application requirements.
• Shaft runout and airgap eccentricity create an unbalanced magnetic pull that adds first- and second-order components.
• Fan cooling, on larger frames, is frequently the loudest single element — broadband airflow noise plus a blade-passing tone. Higher motor efficiency reduces heat generation and can remove the need for forced-air cooling entirely.

Transmission Noise: Gearboxes, Couplings, and Belts

This is the most commonly misdiagnosed category.

• Gear mesh generates a strong tone at mesh frequency (tooth count × shaft speed). Spur planetary gear stages generally produce stronger tonal gear-mesh excitation than well-designed helical gear systems, although actual noise depends heavily on gear accuracy, preload, lubrication, and manufacturing quality. Expect a measurable rise in sound level and a shift to tonal character when any gearbox is added to a bare motor.
• Backlash produces impact noise on every direction reversal, and a persistent ticking if the axis dithers around a position.
• Couplings matter more than expected. A rigid coupling transmits torsional vibration and mechanical disturbances directly to the load. Elastomeric jaw couplings provide torsional damping and reduce shock transmission, while bellows and Oldham couplings are selected when low backlash and specific misalignment compensation are required. Misalignment introduces additional vibration components, increases bearing loads, and accelerates wear.
• Belts and screws contribute transmission noise through tooth-entry impacts in timing belt drives (affected by belt tension, tooth profile, and pulley engagement) and ball recirculation noise in ball screws.

Control-Induced Noise: Gains, Filters, and Motion Profiles

A mechanically sound axis can still be loud if it is badly tuned.

• Velocity loop gain set too high causes the axis to hunt around the target position. The ear detects this dither long before it looks significant on a position trace — it is the usual explanation for a servo that hums or buzzes while standing still.
• Excessive derivative or high-frequency gain causes chatter, sometimes described as sounding like loose ball bearings inside the motor.
• Servo bandwidth overlapping a structural resonance sustains a tone at the resonant frequency. The correct fix is a notch filter tuned to that frequency, not a blanket gain reduction.
• Low feedback resolution creates velocity quantization noise that the gain stage amplifies into audible buzz at standstill.
• Aggressive acceleration and jerk impulsively excite the whole structure. S-curve or jerk-limited profiles reduce this at negligible cycle-time cost.

A useful technique for noise-sensitive equipment: many applications tolerate some noise during motion but must be silent while holding position. Most modern drives support gain switching between two parameter sets — an aggressive set for moves and a low-gain holding set for standstill, where the axis has little work to do anyway.

How Quiet Is Quiet? dB(A) Benchmarks and Measurement Standards

Three properties of decibel arithmetic drive every noise budget:

• +3 dB = double the acoustic power. Two identical motors running together produce about 3 dB more than one.
• +10 dB sound pressure level ≈ perceived roughly twice as loud
• Doubling distance in a free field drops sound pressure by about 6 dB.

The multi-axis implication is the one designers miss. Eight axes at 55 dB(A) each land near 64 dB(A) combined — the motor met its spec and the machine still failed.

Typical Factors Affecting Servo Motor Noise Levels

Actual sound pressure levels vary significantly depending on motor design, operating speed, load condition, cooling method, mounting structure, and measurement distance. In general, noise increases with higher speed and load, while mounting conditions can significantly affect structure-borne vibration.

Three standards govern the measurement:

• ISO 1680 — test code for airborne noise from rotating electrical machinery. This is the motor-level standard.
• ISO 3744 — determination of sound power level from sound pressure, engineering grade, free field over a reflecting plane.
• ISO 11201 — emission sound pressure level at an operator work station. This is the machine-level standard your customer will cite.

Why datasheet dB figures are rarely comparable. Vendors vary the measurement distance (1 m vs. 0.3 m), the load condition (no-load figures are flattering), the speed, the mounting fixture, the ambient floor level, and whether the cooling fan was running. A 52 dB(A) claim measured no-load at 0.3 m on a massive steel block may well be louder in your machine than a 58 dB(A) claim measured at rated load. Always ask for conditions before comparing numbers.

Regulatory context for procurement. Noise requirements may also be influenced by applicable occupational safety regulations and customer acceptance criteria. Machine builders should define system-level noise limits early in the design process and allocate a realistic noise budget across motors, drives, transmissions, and mechanical structures rather than addressing noise only during final acceptance testing.

Servo vs. Stepper Motor: Which Is Quieter for Low-Noise Applications?

This comparison drives a large share of quiet-motor searches, so it is worth answering honestly rather than promotionally.

Stepper motors carry two intrinsic noise mechanisms. They are driven by discrete current commands that create periodic torque excitation, and they are susceptible to mid-band resonance caused by the open-loop control architecture.

Servo motors significantly reduce these effects through continuous closed-loop sinusoidal commutation, although electromagnetic torque ripple, cogging torque, and control-related vibration can still occur. Unlike open-loop stepper systems, servo motors do not rely on discrete step commands for motion generation. Instead, the feedback loop continuously adjusts current commands to maintain position and suppress disturbances. Engineers migrating from stepper systems typically notice smoother motion, lower resonance sensitivity, and more consistent performance across different speeds and loads.

The honest conclusion: a well-microstepped stepper at low speed under light, constant load can be quiet enough, and it will be cheaper. A servo wins clearly wherever speed is moderate to high, load varies, or motion smoothness is itself a requirement. A closed-loop stepper sits between the two and is often the right compromise for cost-sensitive noise-sensitive axes.

Seven Design Features to Look For in a Low-Noise Servo Motor

Each item below pairs the parameter with the mechanism and with a way to verify it on paper.

1. Sinusoidal commutation and low torque ripple. Skewed stator laminations, optimized magnet geometry, and a well-chosen pole/slot combination. Verify: ask for a torque ripple figure as a percentage of rated torque, and confirm the drive uses sine-wave rather than trapezoidal commutation.
2. Low cogging torque. Verify: require detent torque as an explicit number, not an adjective. Low-cogging servo motors are designed to minimize cogging torque relative to rated torque, but acceptable levels depend on the application requirements.
3. High-resolution absolute encoder. Higher feedback resolution means a cleaner velocity estimate, which allows higher gain without quantization buzz, and much smoother low-speed motion. Verify: 17-bit resolves roughly 131,000 counts per revolution; 23-bit exceeds 8 million. Absolute feedback also removes homing routines after power cycling.
4. Bearing grade and rotor balance grade. Verify: ask for the rotor balance grade according to ISO 21940 and select the appropriate grade based on speed requirements. Suppliers who cannot answer usually have not specified it.
5. Drive PWM carrier frequency. Higher PWM carrier frequencies can shift switching-related acoustic components toward less audible frequency ranges, although sidebands and mechanical resonances may still create audible noise. Verify: check whether carrier frequency is user-adjustable, and note the switching-loss trade-off at the top of the range.
6. Vibration suppression and notch filtering in the drive. Often the difference between a quiet machine and a loud one, and almost always overlooked at selection time. Verify: confirm the drive offers adjustable notch filters, a torque command filter, and auto-detection of resonance frequency.
7. Cooling method and efficiency class. Verify: prefer naturally cooled frames where thermally viable. Higher efficiency means less heat, which is what allows the fan to be deleted.

Trade-offs to Watch: Does Quieter Mean Slower or Weaker?

Sometimes, and it is better to plan for it than discover it.

Aggressive cogging reduction can cost a small amount of torque density, so a low-cogging motor may need a slightly larger frame for the same output. Reducing gain to silence a standstill hum costs stiffness and settling time — acceptable on a laboratory pipetting head, unacceptable on a high-throughput packaging axis. Adding compliant couplings or isolation pads lowers structural transmission but also lowers mechanical bandwidth.

The workable approach is a noise budget allocated by axis criticality: spend the acoustic performance where the operator is, and leave the high-bandwidth axes deep in the machine tuned for performance.

System-Level Noise Reduction: Mounting, Gearboxes, and Tuning

Mechanical Integration

Align the motor and load shafts properly — misalignment is both a noise source and a bearing killer. Mount to stiff, massive structure rather than thin panels, and add mass or damping treatment to any panel large enough to radiate. Where isolation is acceptable, elastomeric pads between motor flange and frame break the structure-borne path, though they reduce mechanical stiffness and are unsuitable for high-bandwidth axes.

Transmission Selection

Choose helical or ground planetary gearing over spur-cut stages for noise-sensitive axes. Keep backlash low, because impact noise on reversal scales with it. Select couplings deliberately: bellows and Oldham types tolerate misalignment and damp somewhat, while rigid couplings pass everything straight through.

Tuning and Motion Profile Optimization

Work in this order: identify the resonance frequency, apply a notch filter at that frequency, then raise gains to the level the mechanics support. Add a torque command filter to attenuate high-frequency content. Replace trapezoidal moves with S-curve or jerk-limited profiles. Where standstill silence matters, configure a separate low-gain holding parameter set. Retuning is free; changing hardware is not, which is why it should be attempted first.

When to Replace vs. Retune: A Decision Path

Sound character is a reliable first diagnostic.

Replace the motor when the noise is intrinsic — high cogging, poor balance, a fan you cannot remove — or when a bearing has degraded. Retune when noise is speed-band specific, standstill-only, or appeared after a mechanical change.

Applications That Demand Quiet Servo Motors

Quiet Servo Motor Selection Checklist and FAQ

Work through these eight items before releasing a purchase order:

1. Torque and inertia correctly matched — an undersized axis is always a loud axis.
2. Detent torque stated as a number.
3. Encoder resolution sufficient for smooth low-speed velocity control (17-bit or higher feedback resolution is commonly preferred for demanding low-speed smoothness applications.).
4. Bearing designation and rotor balance grade confirmed.
5. Cooling method known; fan avoided if thermally viable.
6. Gearbox tooth form and backlash specified, and measured with the motor.
7. Drive offers adjustable notch filters, torque command filtering, and gain switching.
8. dB(A) limit written into the RFQ with test conditions, per ISO 1680 or ISO 11201.

Frequently Asked Questions

How many decibels is a quiet servo motor?

There is no universal dB(A) threshold that defines a "quiet" industrial servo motor. Published noise levels should only be compared when speed, load, cooling method, mounting conditions, measurement distance, and test method are equivalent.

Are brushless servo motors quieter than brushed ones?

Yes. Brushed motors add commutator and brush friction noise plus electrical arcing, and they wear, so noise increases over life. Most modern industrial servo motors are brushless, especially AC servo systems used in precision automation.

Why is my new servo motor humming or whining at standstill?

Almost always control-related rather than a defect. A hum or buzz at standstill usually means velocity or position gain is too high for the mechanics, causing the axis to hunt. A high-pitched whine that is constant regardless of position is more likely the drive's PWM carrier. Try a lower holding gain set first, then the carrier frequency.

Are servo motors quieter than stepper motors?

Generally yes, especially at moderate to high speed and under varying load, because sinusoidal closed-loop commutation avoids the stepping excitation and resonance behaviour inherent to open-loop steppers. A well-microstepped stepper at low constant load can be comparably quiet.

Does a gearbox make a servo motor louder?

Usually, yes — gear mesh adds a tonal component that raises the measured level and is often more noticeable than a broadband rise of the same magnitude. Helical or ground gearing and low backlash reduce the penalty. Always specify and measure the motor and gearbox as an assembly.

Conclusion

Quiet servo motor selection is not simply a matter of choosing a motor with a lower dB(A) rating. The final noise performance depends on the complete motion system — including the motor design, servo drive tuning, encoder resolution, transmission components, and mechanical structure.

For noise-sensitive automation, engineers should evaluate key factors such as torque ripple, cogging torque, feedback resolution, cooling method, gearbox selection, and vibration control features before releasing a purchase specification. A properly tuned servo system with suitable mechanical integration will often outperform a theoretically quieter motor installed in a poorly designed structure.

When specifying a low-noise servo motor, define measurable requirements early: operating speed, load condition, measurement distance, mounting method, and acceptable sound level. This allows suppliers to provide comparable data and helps prevent costly redesigns after machine assembly.

The quietest servo system is not the one with the lowest motor noise alone — it is the one where electromagnetic, mechanical, and control-related disturbances are designed and managed together.