Piezo vs. Stepper: Resolution, Noise, and Why Steps Are the Enemy of Precision

Stepper motors are everywhere. They are cheap, reliable, and easy to drive. They require no encoder for basic operation, they hold position at rest with full torque, and they are available in standardized NEMA frame sizes from dozens of manufacturers. For 3D printers, CNC routers, and general automation, steppers are the obvious choice.

But in precision positioning, steppers carry baggage that is often underestimated. The very mechanism that makes them simple (discrete angular steps) introduces vibration, resonance, and resolution limits that are fundamental to the motor's physics. This article examines these limitations with specific numbers and explains when piezoelectric motors provide a quantitatively superior alternative.

Image: Nanomotion Ltd.

The Step: A Feature That Becomes a Bug

A standard hybrid stepper motor has 200 full steps per revolution (1.8 degrees per step). Each step is a discrete jump in rotor position as the magnetic detent snaps from one pole pair to the next. This produces:

1. Step-induced vibration. Every step excites the rotor-load mechanical system with an impulse. The rotor overshoots, oscillates around the new position, and eventually settles. The amplitude of this oscillation is typically 3% to 10% of one full step, which for a 1.8-degree motor corresponds to 0.05 to 0.18 degrees of angular oscillation, or (with a 2 mm pitch lead screw) 0.3 to 1.0 micrometers of linear vibration.

2. Torque ripple. The electromagnetic torque is not constant during a step. It follows a sinusoidal-like profile, creating cyclic variations in available torque. At constant velocity, this manifests as speed ripple (periodic variation in instantaneous velocity) of 5% to 15% for full-step operation.

3. Acoustic noise. The step-induced vibration is audible. A stepper running at 200 to 2,000 steps per second produces tonal noise at 200 Hz to 2 kHz, exactly in the frequency range where human hearing is most sensitive. Sound levels of 50 to 70 dBA at 1 meter distance are common for NEMA 17 and NEMA 23 motors.

Piezoelectric ultrasonic motors, by contrast, operate by continuous friction coupling driven at ultrasonic frequencies (20 to 200 kHz). Motion is smooth and continuous with no discrete steps. Vibration is at frequencies well above the mechanical bandwidth of any practical load. Acoustic noise is minimal (typically below 30 dBA) because the driving frequency is above the audible range.

Vibration Spectrum Analysis

The vibration signature of a stepper motor is not random; it is highly structured and predictable. Understanding this spectrum is essential for evaluating whether stepper-induced vibration will affect a given application.

Stepper Motor Vibration Spectrum

A stepper motor running at constant velocity produces vibration at several characteristic frequencies:

Fundamental step frequency: f_step = (steps/rev x velocity) / (screw_pitch). For a 200 step/rev motor driving a 2 mm pitch lead screw at 1 mm/s: f_step = (200 x 1) / 2 = 100 Hz. This is the dominant vibration frequency.

Harmonics: The step waveform is not sinusoidal; it is a series of impulses. The Fourier decomposition produces harmonics at 2x, 3x, 4x, and higher multiples of f_step. Typically the 2nd and 3rd harmonics are 6 to 12 dB below the fundamental, with higher harmonics dropping off at roughly 6 dB per octave.

Detent torque frequency: The motor's inherent cogging produces vibration at 4x the fundamental step frequency (for a 2-phase, 50-pole-pair motor). This component is present even during microstepping and cannot be eliminated by drive electronics.

Mechanical resonance excitation: If any harmonic of the step frequency coincides with a mechanical resonance of the load structure, that component is amplified by the resonance Q-factor (typically 10 to 50 for metal structures, 5 to 20 for polymer structures).

The following table shows measured vibration amplitudes for a NEMA 17 stepper motor (0.4 Nm holding torque) driving a 2 mm pitch lead screw with a 200 g payload, at 1 mm/s velocity:

Frequency Component	Frequency (Hz)	Full-Step Amplitude (um)	1/8 Microstep Amplitude (um)	1/256 Microstep Amplitude (um)
Fundamental (f_step)	100	1.2	0.15	0.005
2nd harmonic	200	0.6	0.08	0.003
3rd harmonic	300	0.3	0.05	0.002
4th harmonic (detent)	400	0.4	0.35	0.30
5th harmonic	500	0.1	0.02	0.001
Sum of all components	broadband	1.5	0.40	0.31

Critical observation: microstepping dramatically reduces the fundamental and low-order harmonics (by 10x to 200x), but the detent torque component (4th harmonic) is barely affected. This is because detent torque is a magnetic property of the motor's permanent magnets and iron geometry; it exists regardless of the drive waveform. The detent vibration at 0.30 to 0.35 micrometers sets a hard floor on achievable vibration, even with 1/256 microstepping.

Piezo Motor Vibration Spectrum

An ultrasonic piezo motor driving the same load at the same velocity produces a fundamentally different vibration signature:

Frequency Range	Amplitude (um)	Source
0 to 100 Hz	0.001 to 0.005	Controller noise, encoder quantization
100 to 1,000 Hz	0.002 to 0.01	Friction contact dynamics
1 to 10 kHz	0.001 to 0.005	Structural resonance (minimal excitation)
20 to 200 kHz	0.01 to 0.1	Motor drive frequency (ultrasonic, not transmitted to load)
Total (0 to 10 kHz)	0.003 to 0.015	All sources combined

The piezo motor's vibration in the mechanically relevant frequency range (0 to 10 kHz) is 20x to 100x lower than the microstepped stepper's vibration. The motor does vibrate at ultrasonic frequencies (20 to 200 kHz), but these frequencies are far above the mechanical bandwidth of any practical load structure and are attenuated by the contact interface. They do not transmit to the workpiece.

Image: Nanomotion Ltd.

Impact on Sensitive Applications

For scanning probe microscopy, the relevant vibration band is 1 to 1,000 Hz, with a typical noise floor requirement of 0.1 nm RMS. A stepper motor at 1/256 microstepping produces roughly 310 nm of vibration in this band. Even with vibration isolation (10x to 100x attenuation), the residual 3 to 30 nm is far above the 0.1 nm requirement. The stepper is categorically disqualified.

A piezo motor produces 3 to 15 nm of vibration in this band, and with moderate vibration isolation, achieves 0.03 to 0.15 nm. This meets the requirement with margin.

For optical metrology (interferometric surface measurement), the relevant band is 10 to 500 Hz, with a typical requirement of 1 to 5 nm RMS. The stepper at 1/256 microstepping produces roughly 300 nm in this band (dominated by detent torque vibration). Completely inadequate. The piezo motor produces 2 to 10 nm, which meets or approaches the requirement without supplementary isolation.

Microstepping: The Resolution Myth

Microstepping is the standard technique for improving stepper motor resolution. Instead of driving full current to one phase at a time, the controller divides each full step into smaller "microsteps" by applying sinusoidally varying currents to both phases simultaneously. Common microstep ratios are 1/4, 1/8, 1/16, 1/32, 1/64, and 1/256.

With 1/256 microstepping on a 200-step motor, the theoretical angular resolution is 1.8 / 256 = 0.007 degrees, or (with a 2 mm pitch lead screw) 0.04 micrometers per microstep. This sounds impressive. It is also largely fictional.

Why High Microstep Ratios Don't Deliver Proportional Resolution

The problem is that microstep accuracy degrades rapidly beyond about 1/8 to 1/16 microstepping. The reasons are fundamental:

Motor magnetic non-ideality. Microstepping assumes that the motor's torque varies sinusoidally with rotor position. Real motors have significant harmonic content in their torque profiles (5% to 20% third harmonic, plus higher orders). These harmonics cause the rotor to not follow the commanded microstep positions accurately. The position error at each microstep can be 5% to 30% of a full step, regardless of the microstep ratio.

Detent torque. Hybrid stepper motors have a residual detent torque (cogging torque) even with no current applied. This detent torque has a spatial period of one full step and an amplitude of 1% to 10% of the motor's holding torque. At fine microstep levels, the detent torque is comparable to or larger than the torque difference between adjacent microsteps. The rotor simply cannot resolve the fine microstep positions against this background torque.

Friction. Both the motor bearings and the drivetrain (lead screw, coupling, linear guide) have static friction (stiction). The torque required to overcome stiction is typically 2% to 10% of the motor's holding torque. If the torque change between adjacent microsteps is less than the stiction torque, the rotor does not move. For a NEMA 17 motor with 0.4 Nm holding torque and 0.02 Nm stiction, the minimum resolvable step requires approximately 5% of full-step torque, corresponding to roughly 1/16 to 1/32 microstepping. Finer microsteps produce commanded positions that the motor cannot physically reach.

Measured vs. Theoretical Microstep Accuracy

Multiple independent studies have measured actual microstep positioning accuracy. The following table presents a consolidated view of measured results for a NEMA 17 motor (0.4 Nm, 1.8 deg/step) with a 2 mm pitch lead screw:

Microstep Setting	Theoretical Step Size (um)	Measured Average Step Size (um)	Measured Step-to-Step Variability (um)	Effective Resolution (um)	Ratio: Measured/Theoretical
Full step (1/1)	10.0	10.0	0.3 to 1.0	10.0	1.0x
1/2	5.0	4.8 to 5.2	0.3 to 0.8	5.0	1.0x
1/4	2.5	2.2 to 2.8	0.3 to 0.7	2.5	1.0x
1/8	1.25	0.9 to 1.6	0.3 to 0.6	1.25	1.0x
1/16	0.625	0.3 to 0.9	0.3 to 0.6	0.6 to 0.9	1.0 to 1.4x
1/32	0.3125	0.1 to 0.5	0.2 to 0.5	0.5 to 1.0	1.6 to 3.2x
1/64	0.156	0 to 0.4	0.2 to 0.5	0.5 to 1.0	3.2 to 6.4x
1/128	0.078	0 to 0.3	0.2 to 0.5	0.5 to 1.0	6.4 to 12.8x
1/256	0.039	0 to 0.2	0.2 to 0.5	0.5 to 1.0	12.8 to 25.6x

The pattern is unmistakable. Up through 1/8 microstepping, the measured step size tracks the theoretical value closely (within the step-to-step variability). At 1/16, the measured steps begin to deviate. By 1/64 and beyond, the measured step size is essentially random within a 0 to 0.5 micrometer band; commanding finer microsteps produces no additional positioning resolution. The effective resolution plateau is 0.5 to 1.0 micrometer, regardless of the microstep setting.

This plateau is set by the combination of detent torque, friction, and magnetic non-ideality. No amount of microstepping refinement can push below it. The only escape is closed-loop control with an external encoder, which defeats the primary advantage of stepper motors.

Quantifying the Resolution Lie

A vendor claims "0.04 micrometer resolution" for a stepper axis based on 1/256 microstepping with a 2 mm pitch screw. What is the actual resolution?

Theoretical: 2,000 um / (200 x 256) = 0.039 um

Actual: 0.5 to 1.0 um (measured), a factor of 13x to 25x worse than claimed

If a customer specifies 0.1 micrometer resolution based on the vendor's claim, the system will fail. The stepper axis cannot achieve it in open-loop operation. Period.

Ultrasonic piezo motors have no steps, no detent torque, and no torque ripple. Minimum incremental motion is determined by the friction contact dynamics and the encoder resolution, with typical values of 5 to 50 nm for commercial stages. This is 10 to 100 times finer than a microstepped stepper can reliably achieve.

Resonance: The Stepper Motor's Hidden Failure Mode

Stepper motors have a dangerous characteristic that is often overlooked in specification sheets: mid-frequency resonance.

Low-Frequency Resonance

At low step rates (typically 50 to 200 full steps per second), stepper motors enter a natural resonance where the step frequency matches the rotor-load mechanical resonant frequency. At this point, step-induced oscillations accumulate rather than damp, and the motor can lose synchronization (miss steps). The resonant frequency depends on motor inertia, load inertia, and the motor's electromagnetic stiffness:

f_resonance = (1 / (2 x pi)) x sqrt(K_hold / J_total)

For a typical NEMA 17 motor (holding torque 0.4 Nm, rotor inertia 5.4 x 10^-6 kg*m^2) with a load inertia equal to the rotor inertia:

f_resonance ~ 200 to 400 Hz (full steps per second)

This corresponds to linear speeds of 0.4 to 0.8 mm/s with a 2 mm pitch lead screw. Many precision positioning applications operate exactly in this speed range.

Mid-Frequency Instability

Above the low-frequency resonance, stepper motors can also exhibit mid-frequency instability (sometimes called "mid-band resonance") at 1,000 to 5,000 steps per second. This is caused by interaction between the electrical time constant of the motor windings and the mechanical dynamics. The motor may vibrate excessively, lose steps, or stall unpredictably.

Mitigation techniques include microstepping (which reduces the excitation amplitude), viscous damping (which adds drag), and current profiling (which reshapes the step waveform). All of these help but do not eliminate the underlying resonance; they merely reduce its amplitude.

Resonance Map: Danger Zones for Common Motor Sizes

The following table identifies the resonance danger zones for common stepper motor frame sizes, assuming a load inertia equal to the rotor inertia and a 2 mm pitch lead screw:

Motor Size	Holding Torque (Nm)	Rotor Inertia (kg*m^2)	Low-Freq Resonance (Hz)	Equivalent Linear Speed (mm/s)	Mid-Band Resonance (Hz)	Equivalent Linear Speed (mm/s)
NEMA 11	0.06 to 0.12	1.8 x 10^-7	400 to 600	0.8 to 1.2	2,000 to 4,000	4 to 8
NEMA 14	0.10 to 0.20	1.0 x 10^-6	250 to 350	0.5 to 0.7	1,500 to 3,000	3 to 6
NEMA 17	0.20 to 0.55	3.5 to 7.0 x 10^-6	150 to 300	0.3 to 0.6	1,000 to 3,000	2 to 6
NEMA 23	0.5 to 2.0	1.5 to 6.0 x 10^-5	80 to 200	0.16 to 0.4	500 to 2,000	1 to 4
NEMA 34	2.0 to 8.0	5.0 to 20 x 10^-5	50 to 150	0.1 to 0.3	300 to 1,500	0.6 to 3

Notice that the low-frequency resonance falls squarely in the 0.1 to 1.2 mm/s range, exactly where many precision positioning applications operate. The mid-band resonance zone extends to 8 mm/s for small motors, covering even moderate-speed precision scanning.

The Practical Impact

A stepper motor that loses steps is worse than one that moves slowly. Lost steps produce cumulative position error with no indication to the controller (in open-loop operation). The system believes it is at position X while the actual position is X plus some unknown error. In precision applications, this is catastrophic.

Worked example: A NEMA 17 stepper driving a microscope Z-focus at 0.5 mm/s (within the resonance zone). The motor occasionally loses 1 to 3 steps at direction reversals. Each lost step corresponds to 10 micrometers of position error (full-step mode) or 1.25 micrometers (1/8 microstep mode). Over 100 reversal cycles, the cumulative error could reach 125 to 300 micrometers if steps are consistently lost in one direction. The microscope's autofocus algorithm compensates by running extra approach cycles, but this wastes time and does not guarantee convergence. Adding an encoder for closed-loop step verification costs $100 to $500 and adds complexity. At this point, the simplicity advantage of the stepper motor is significantly diminished.

Ultrasonic piezo motors have no discrete steps and therefore no step-induced resonance. The ultrasonic driving frequency is far above any mechanical resonance of the load. Motion is smooth and continuous across all operating speeds. Closed-loop operation with an encoder is standard, and lost-step events are not possible.

Acoustic Noise Comparison

In laboratory, clinical, and office environments, acoustic noise matters. The comparison is dramatic.

Stepper motors generate tonal noise at the step frequency and its harmonics. At 1,000 full steps per second (a common operating point), the fundamental frequency is 1 kHz, with harmonics at 2, 3, 4 kHz. Measured sound levels:

• Full stepping, NEMA 17, no load: 55 to 65 dBA at 1 m
• 1/8 microstepping, NEMA 17, no load: 40 to 55 dBA at 1 m
• 1/256 microstepping, NEMA 17, no load: 30 to 45 dBA at 1 m

Microstepping reduces noise substantially (this is one of its genuine benefits), but even at 1/256, the motor is still audible in a quiet room.

Detailed Acoustic Measurements Across Operating Conditions

The following table consolidates acoustic measurements for a NEMA 17 stepper (0.44 Nm) and a NEMA 23 stepper (1.26 Nm), compared with a representative ultrasonic piezo linear motor, all measured at 1 meter distance with an A-weighted sound level meter:

Condition	NEMA 17 (dBA)	NEMA 23 (dBA)	Piezo Motor (dBA)	Reference: Office Background (dBA)
Stationary, energized (holding)	25 to 30	28 to 35	< 25 (inaudible)	35 to 45
Full-step, 0.5 mm/s	55 to 62	60 to 68	26 to 30	35 to 45
Full-step, 2 mm/s	58 to 65	63 to 70	27 to 32	35 to 45
1/8 microstep, 0.5 mm/s	38 to 48	42 to 52	26 to 30	35 to 45
1/8 microstep, 2 mm/s	42 to 52	48 to 58	27 to 32	35 to 45
1/256 microstep, 0.5 mm/s	28 to 38	32 to 42	26 to 30	35 to 45
1/256 microstep, 2 mm/s	32 to 42	38 to 48	27 to 32	35 to 45
At resonance speed	65 to 75	70 to 80	N/A (no resonance)	35 to 45

Key observations:

1. At full stepping, steppers are 25 to 40 dB louder than piezo motors. Since dBA is logarithmic, this corresponds to perceived loudness roughly 6 to 16 times greater.
2. At 1/256 microstepping, the gap narrows to 2 to 15 dB, but the stepper remains audible in a quiet room, while the piezo motor is at or below the typical office background noise.
3. At resonance, stepper noise spikes to 65 to 80 dBA, comparable to a vacuum cleaner. This spike occurs at specific speeds and can be startling and disruptive.
4. The NEMA 23 is consistently 5 to 8 dB louder than the NEMA 17 due to its larger mass and greater step impulse energy.
5. Piezo motor noise is nearly independent of operating speed, varying by only 1 to 2 dB across the speed range. The noise is dominated by structural vibration in the mounting, not the motor itself.

Ultrasonic piezo motors operate at 20 to 200 kHz, which is above human hearing (nominally 20 kHz). Measured sound levels are typically 25 to 35 dBA at 1 m, dominated by structural vibration transmitted through the stage body and mounting, not by the motor itself. In practice, a piezo motor stage is essentially silent to human ears.

For applications in medical imaging (MRI compatible stages), microscopy (vibration-sensitive optical systems), or general laboratory use, the noise difference alone can be a deciding factor.

Thermal Comparison

Stepper motors consume significant power at rest. Unlike servo motors, which can reduce current when holding a static position, stepper motors are typically driven at full rated current at all times (including at standstill) to maintain holding torque and prevent step loss. A NEMA 17 motor rated at 1.5 A per phase and 2.1 ohms per phase dissipates:

P_hold = 2 x I^2 x R = 2 x 1.5^2 x 2.1 = 9.45 W

This heat is generated continuously, even when the motor is not moving. Motor case temperatures of 60 to 90 degrees Celsius are normal. In enclosed spaces, thermal management becomes a concern.

Current reduction at standstill (reducing holding current to 50% to 70% of run current) is common practice and reduces heat by 50% to 75%, but also reduces holding torque by the same ratio, increasing the risk of step loss under external disturbance.

Thermal Dissipation Calculations Across Motor Sizes

The following table calculates static hold power for common stepper motor sizes at full rated current and at typical reduced standby current:

Motor Size	Rated Current/Phase (A)	Resistance/Phase (ohm)	Hold Power at Full Current (W)	Hold Power at 70% Current (W)	Hold Power at 50% Current (W)	Typical Case Temp at Full Current (C)
NEMA 11	0.5	6.0	3.0	1.47	0.75	45 to 60
NEMA 14	0.8	3.5	4.5	2.2	1.1	50 to 65
NEMA 17 (std)	1.5	2.1	9.45	4.6	2.4	60 to 85
NEMA 17 (high torque)	2.0	1.4	11.2	5.5	2.8	65 to 90
NEMA 23 (single stack)	2.8	0.9	14.1	6.9	3.5	55 to 75
NEMA 23 (double stack)	2.8	1.5	23.5	11.5	5.9	65 to 90
NEMA 34	4.0	0.6	19.2	9.4	4.8	60 to 85

Worked thermal example: A microscope XY stage using two NEMA 17 motors in an enclosed housing. The motors operate at full current continuously (to ensure position retention during imaging). Total heat generation: 2 x 9.45 = 18.9 W. The enclosed housing has a thermal resistance to ambient of approximately 1.5 degrees C/W (natural convection from outer surfaces). Steady-state temperature rise of the housing: 18.9 x 1.5 = 28.4 degrees C above ambient. If the room is 22 degrees C, the housing reaches 50.4 degrees C, and the motor cases reach 70 to 90 degrees C.

This heat creates thermal gradients in the microscope frame. With a steel frame having a thermal expansion coefficient of 12 ppm/degree C and a 200 mm optical path length, a 10 degree C gradient causes 200 x 12 x 10^-6 x 10 = 0.024 mm = 24 micrometers of thermal drift. For a microscope claiming 1 micrometer positioning accuracy, this thermal drift is 24x the specified accuracy. The motors are the dominant error source.

Mitigation strategies and their costs:

Strategy	Temperature Reduction	Additional Cost	Drawbacks
Current reduction to 50% at hold	50% to 75% power reduction	$0 (driver feature)	50% torque reduction, step loss risk
Heat sink on motor	10 to 20 C	$10 to $30 per motor	Adds size and mass
Fan cooling	15 to 30 C	$20 to $50 plus power	Vibration, noise, airflow disturbance
Thermal isolation (G10 spacers)	Reduces gradient, not temp	$50 to $200	Adds compliance, reduces stiffness
Liquid cooling jacket	25 to 40 C	$200 to $500 per motor	Plumbing, complexity, cost
Peltier cooling	20 to 35 C	$100 to $300 per motor	Power consumption, condensation risk

Each mitigation adds cost, complexity, or performance compromise. The total cost of thermal management for two NEMA 17 motors can easily reach $200 to $1,000, narrowing the cost gap with piezo motors.

Piezo Motor Thermal Profile

Ultrasonic piezo motors have zero power consumption at standstill. The friction preload holds the stage in position passively. During motion, power consumption is typically 1 to 5 W for a small linear stage. The motor does not heat up significantly during normal operation.

For applications requiring long hold times (hours to days), in thermally sensitive environments (optical metrology, electron microscopy), or in confined spaces without active cooling, the thermal advantage of piezo motors is substantial.

Thermal Impact on Positioning Accuracy

The following table quantifies the positioning error caused by thermal expansion from motor self-heating, for a typical precision stage with a 100 mm frame dimension:

Motor Type	Hold Power (W)	Frame Temp Rise (C)	Thermal Drift (um, aluminum frame)	Thermal Drift (um, steel frame)	Thermal Drift (um, Invar frame)
NEMA 17 stepper, full current	9.45	8 to 15	18 to 35	10 to 18	0.1 to 0.2
NEMA 17 stepper, 50% current	2.4	2 to 5	5 to 12	3 to 6	0.03 to 0.06
Piezo motor, at hold	~0	< 0.1	< 0.2	< 0.1	< 0.001
Piezo motor, during motion	2 to 4	0.5 to 1.5	1 to 3.5	0.6 to 1.8	0.006 to 0.018

For a precision stage on an aluminum frame, the stepper motor at full current introduces 18 to 35 micrometers of thermal drift, which completely dominates the error budget. Even at reduced current, 5 to 12 micrometers of thermal drift remains. The piezo motor at hold contributes less than 0.2 micrometers.

Velocity Smoothness

For scanning applications (microscopy, metrology, printing), velocity smoothness is often as important as positional accuracy. The relevant metric is velocity ripple: the peak-to-peak variation in instantaneous velocity as a percentage of commanded velocity.

Stepper motors exhibit velocity ripple from two sources:

1. Step-induced ripple at the step frequency. Amplitude: 5% to 15% for full stepping, 1% to 5% for 1/8 microstepping, 0.2% to 1% for 1/256 microstepping.
2. Lead screw ripple from pitch error periodicity (typically at once-per-revolution frequency). Amplitude: depends on screw quality, typically 0.5% to 3% for precision screws.

Total velocity ripple for a microstepped stepper with lead screw is typically 1% to 5%.

Ultrasonic piezo motors with closed-loop velocity control achieve velocity ripple of 0.1% to 1%, depending on the feedback sensor and controller bandwidth. The absence of discrete steps eliminates the step-induced ripple component entirely. The remaining ripple comes from encoder quantization, controller bandwidth limitations, and friction variations at the contact interface.

For applications requiring velocity uniformity better than 1%, piezo motors are the more straightforward solution.

When Steppers Are Fine

Despite the limitations discussed above, stepper motors are the correct choice for many applications. Specifically:

General automation and robotics. When positioning accuracy of 5 to 50 micrometers is sufficient, speeds of 10 to 500 mm/s are needed, and cost must be minimized, stepper motors with lead screws are hard to beat. The system cost (motor + driver + lead screw) can be as low as $50 to $200.

3D printing and CNC. These applications need millimeter-scale accuracy at high speed. Stepper motors are the established solution and are supported by mature, open-source controller ecosystems.

Low-duty-cycle positioning. When the motor moves infrequently (adjusting a fixture, setting an initial position), stepper motor limitations in vibration and noise are irrelevant. The motor moves, settles, and stays.

High-force, long-stroke applications. A NEMA 34 stepper motor with a ball screw produces 1,000 to 5,000 N of linear force over strokes of 100 to 1,000 mm. No piezo motor can match this combination of force and stroke.

Education and prototyping. Steppers are cheap, well-documented, and easy to drive with commodity electronics. They are the right learning tool.

When Steppers Are Not

Stepper motors become the wrong choice when:

Sub-micrometer resolution is required. Microstepping cannot reliably deliver below 1 micrometer, and adding a closed-loop encoder negates the stepper's simplicity advantage.

Acoustic noise must be minimized. Even with aggressive microstepping, steppers are audible. Piezo motors are not.

Vibration-free motion is needed. Optical metrology, scanning probe microscopy, and live-cell imaging cannot tolerate the step-induced vibration of a stepper motor.

Continuous scanning at constant velocity is required. Velocity ripple from stepping limits stepper performance in scanning applications.

The motor must operate in vacuum or cleanroom. Stepper motors generate heat continuously, create magnetic fields, and (with lead screw transmission) may generate particles. Piezo motors avoid all three issues.

Size and weight are critical. Piezo motor stages are typically 50% to 80% smaller and lighter than equivalent stepper-driven stages for the same stroke and resolution.

Hold time is long and thermal stability matters. A stepper motor dissipating 10 W continuously into a precision optical bench introduces thermal gradients that degrade measurement accuracy.

Detailed Quantitative Comparison

Parameter	Stepper (full step + lead screw)	Stepper (1/16 microstep + lead screw)	Stepper (1/256 microstep + lead screw)	Piezo motor (closed-loop)
Theoretical step size (2 mm pitch)	10 um	0.625 um	0.039 um	N/A (continuous)
Actual min. incremental motion	10 um	1 to 5 um	0.5 to 1.0 um	0.005 to 0.05 um
Bidirectional repeatability	10 to 30 um	2 to 10 um	2 to 10 um	0.05 to 0.5 um
Velocity ripple	5% to 15%	2% to 5%	0.2% to 1%	0.1% to 1%
Acoustic noise at 1 mm/s	55 to 65 dBA	38 to 52 dBA	28 to 42 dBA	26 to 32 dBA
Vibration (0 to 1 kHz)	1.5 um	0.5 um	0.3 um	0.01 um
Power at hold (no load)	5 to 20 W	5 to 20 W	5 to 20 W	0 W
Max practical speed	50 to 500 mm/s	50 to 500 mm/s	50 to 500 mm/s	5 to 200 mm/s
Continuous force	10 to 5,000 N (with screw)	10 to 5,000 N	10 to 5,000 N	1 to 20 N
System cost (complete axis)	$100 to $500	$150 to $800	$200 to $1,000	$2,000 to $8,000
Size (for 25 mm stroke)	80 x 80 x 150 mm	80 x 80 x 150 mm	80 x 80 x 150 mm	30 x 30 x 60 mm
Resonance risk	Severe (100 to 400 Hz)	Moderate (reduced by microstepping)	Low (further reduced)	None
Encoder required	No (open-loop)	No (open-loop)	No (open-loop)	Yes (closed-loop standard)
Thermal drift (100 mm frame, Al)	18 to 35 um	18 to 35 um	18 to 35 um	< 0.2 um (at hold)

The Resolution-Cost Frontier

The decision between stepper and piezo often comes down to a resolution-cost tradeoff. There is a resolution threshold below which steppers cannot go, regardless of microstepping ratio, and a cost floor below which piezo motors are not available. The crossover region is defined by specific price points.

Resolution-Cost Data Points

Target Resolution (um)	Stepper Solution	Stepper Cost	Piezo Solution	Piezo Cost	Winner
50	NEMA 17, 1/4 step, lead screw	$80 to $200	Overkill; not applicable	N/A	Stepper
20	NEMA 17, 1/8 step, lead screw	$100 to $250	Overkill; not applicable	N/A	Stepper
10	NEMA 17, full step, lead screw	$100 to $250	Overkill; not applicable	N/A	Stepper
5	NEMA 17, 1/8 step, precision lead screw	$200 to $500	Basic piezo stage, 1 um encoder	$2,000 to $3,500	Stepper
2	NEMA 17, 1/16 step, precision lead screw	$300 to $700	Basic piezo stage, 0.5 um encoder	$2,500 to $4,000	Stepper (on cost)
1	NEMA 17, 1/32 step, precision screw + tuning	$500 to $1,200	Standard piezo stage, 0.1 um encoder	$3,000 to $5,000	Gray zone
0.5	Stepper + linear encoder + closed-loop	$1,200 to $2,500	Standard piezo stage, 0.1 um encoder	$3,000 to $5,000	Gray zone; piezo more reliable
0.1	Stepper + linear encoder + closed-loop + anti-backlash	$2,000 to $4,500	Precision piezo stage, 20 nm encoder	$4,000 to $7,000	Piezo (stepper marginal)
0.05	Not achievable with stepper	N/A	Precision piezo stage, 5 nm encoder	$5,000 to $8,000	Piezo (only option)
0.01	Not achievable with stepper	N/A	High-end piezo stage, 1 nm encoder	$7,000 to $12,000	Piezo (only option)
0.005	Not achievable with stepper	N/A	Ultra-precision piezo stage	$10,000 to $20,000	Piezo (only option)

The crossover zone is approximately 0.5 to 2 micrometers. Below 0.5 micrometers, the stepper solution either cannot deliver the resolution or requires so many add-ons (encoder, closed-loop controller, anti-backlash mechanisms) that it costs nearly as much as a piezo stage while being larger, noisier, and less reliable.

The Hidden Costs That Shift the Crossover

The raw motor and stage cost tells only part of the story. When targeting 0.5 to 2 micrometer resolution, the stepper system accumulates hidden costs:

Hidden Cost Item	Stepper System	Piezo System
External linear encoder	$300 to $1,500	Included in stage
Closed-loop stepper driver	$200 to $800 (vs. $50 for open-loop)	Standard (included)
Anti-backlash lead screw nut	$100 to $400	N/A (direct drive)
Thermal management (heat sink, isolation)	$50 to $300	$0 (no heat at hold)
Vibration isolation from motor	$100 to $500	$0 (ultrasonic, no transmission)
Resonance avoidance (dampers, profiling)	$50 to $200 (engineering time + parts)	$0 (no resonance)
Integration and tuning labor	8 to 16 hours at $100/hr = $800 to $1,600	2 to 4 hours at $100/hr = $200 to $400
Total hidden costs	$1,600 to $5,300	$200 to $400

When these hidden costs are included, the "cheap stepper" at 1 micrometer resolution costs $2,100 to $6,500 total, while the "expensive piezo" costs $3,200 to $5,400 total. The cost gap vanishes, and the piezo solution delivers better vibration, noise, and thermal performance.

Above 5 micrometers resolution: Steppers almost always win on cost. The resolution is easily achievable, and the system is simple and inexpensive.

1 to 5 micrometers resolution: The gray zone. A high-quality stepper axis with fine microstepping and a precision lead screw can sometimes deliver this. A piezo motor stage delivers it more reliably but at higher cost.

Below 1 micrometer resolution: Steppers cannot reliably deliver this without external encoders and closed-loop control, which eliminates their main advantage. Piezo motors deliver sub-micrometer resolution routinely and cost-effectively.

If your resolution requirement falls below 1 micrometer, the stepper motor path leads to a system that costs nearly as much as a piezo stage, is larger, noisier, and less reliable. At that point, the piezo motor is the technically and economically superior choice.