SMD Chip Inductors

SMD surface mount inductors can be easily placed by high‑speed pick‑and‑place machines and thus reduce assembly cost compared to through‑hole coils.

SMD inductors are surface-mount magnetic components used for energy storage, filtering, resonance and RF impedance matching. This article explains multilayer and wire-wound constructions, compares their practical limitations, and gives design and component engineers a selection workflow covering inductance under load, current ratings, losses, self-resonance and qualification.

Key Takeaways

Select an SMD inductor by its circuit function and operating conditions, not by package size and nominal inductance alone.

• Multilayer, wire-wound and thin-film constructions address different combinations of inductance, current, frequency, size and loss.
• Saturation current describes a specified inductance reduction; thermal current describes a specified temperature rise. Both limits must be checked independently.
• For power converters, use effective inductance at operating current and temperature, and assess DC winding loss together with AC winding and core losses.
• Magnetic shielding depends on the magnetic structure. Plastic encapsulation alone does not establish shielding or hermetic sealing.
• RF matching normally requires inductive behaviour below self-resonance, whereas an RF choke may deliberately exploit high impedance near self-resonance.
• Validate the selected part on the intended PCB. Current-rating test conditions, copper area, airflow, assembly and component qualification affect the final design.

SMD Multilayer Inductors

A multilayer SMD inductor contains conductive patterns embedded in an insulating or magnetic body and interconnected to form a coil. Figures 1 and 2 illustrate the construction and a ferrite multilayer component; this is a different manufacturing approach from winding a discrete wire around a core.

In a typical ferrite multilayer construction, printed internal conductors are formed on layers that are stacked and processed into a component body. The surrounding magnetic material helps confine flux, making suitable ferrite multilayer series useful in densely packed signal filters and compact power circuits. The degree of field confinement and the usable frequency range remain properties of the particular component, not guarantees attached to the word multilayer.

Multilayer RF inductors can instead use non-magnetic ceramic bodies. These must not be grouped indiscriminately with ferrite multilayer power inductors: their magnetic behaviour, loss mechanisms and intended operating conditions differ. Likewise, a ferrite bead intended to provide lossy noise suppression is not automatically interchangeable with an inductor selected for energy storage or a high-Q resonant network.

Controlled internal geometry supports repeatable component characteristics, but the applicable inductance tolerance, quality factor and parasitic behaviour must still come from the datasheet. Multilayer production should not be described universally as photolithographic, and modern wire-wound components should not be assumed to have hand-wound variability.

Practical Tips

• Check inductance, impedance and quality factor at the operating frequency rather than relying only on a low-frequency nominal value.
• For ferrite parts carrying DC, examine inductance versus current and temperature together with the thermal-current limit.
• Compare actual maximum DC resistance. Low resistance is valuable in low-voltage systems, but it is not an inherent property of every multilayer inductor.
• For resonant and matching applications, maintain the required inductive behaviour over the complete operating band. For RF choking, select the impedance response required by the circuit, including deliberate use of resonance where appropriate. [1]

Power Multilayer Inductors

Power multilayer inductors extend the embedded-conductor approach to compact energy-storage applications. Figure 3 shows examples. Low height and small footprint make these components relevant to portable equipment, but current capability and loss must be evaluated at the required inductance rather than from a family headline rating.

For a given converter voltage ratio and target ripple current, a higher switching frequency reduces the inductance required. It does not produce a universal proportional reduction in component volume: winding resistance, magnetic loss, saturation, temperature rise and manufacturable geometry constrain the achievable size. The converter IC also imposes an acceptable inductance range and control-loop requirements. [2]

As a construction example, the WE-PMI catalogue includes low-profile, low-resistance and higher-saturation variants. Its 1008 versions have a nominal body length and width of 2.5 mm × 2.0 mm, while height and electrical ratings depend on the ordering code. These variants illustrate why a package designation is not a complete mechanical or electrical specification. [3]

Figures 4 and 5 retain the original illustrations of inductance versus DC current and frequency. They show why a nominal inductance cannot be treated as constant under all operating conditions. Design acceptance must use the curves for the selected ordering code and the relevant temperature, rather than extrapolating from an illustrative family plot.

Compare power multilayer parts with shielded wire-wound ferrite and metal-composite alternatives at the same effective inductance, current waveform, temperature and height constraint. Neither the smallest footprint nor the lowest room-temperature DC resistance alone establishes the most efficient solution.

SMD Wire-Wound and RF Inductors

Wire-wound SMD inductors use a discrete conductor wound around a core or support. Their construction ranges from exposed windings to encapsulated RF parts and magnetically shielded power components; these categories should be compared by their electrical and environmental specifications rather than appearance alone.

Figure 6 retains the original comparison between the open WE-LQ construction and plastic-encapsulated WE-GF construction. Encapsulation provides a physical covering, but it does not by itself demonstrate hermetic sealing, magnetic shielding or a higher current rating. Humidity resistance and cleaning compatibility require component-specific evidence.

An exposed winding may require particular attention to handling, contamination and assembly processes. An encapsulated component still requires review of reflow limits, permitted cleaning agents and mechanical stresses. Current capability depends on conductor cross-section, turns, core geometry, losses and heat removal; it cannot be inferred from open versus enclosed construction alone.

Figures 7 and 8 illustrate ferrite-core and encapsulated wire-wound arrangements. Ferrite can provide useful inductance in a compact structure, while non-magnetic ceramic and air-core RF constructions serve applications where frequency response, linearity and Q are more important than inductance density. RF selection therefore requires the impedance and Q curves, and where available application-relevant network models, rather than nominal inductance alone. [1]

Moulded Power Inductors Are a Separate Category

In a metal-composite moulded power inductor, magnetic material surrounds the winding and forms part of the magnetic structure. This is not equivalent to simply placing a plastic cover over an RF coil. Such parts belong in the comparison for compact switching converters, not solely in an RF category.

Contemporary moulded power-inductor families include devices intended for converter operation from hundreds of kilohertz into the megahertz range. For example, the Coilcraft XGL family is specified for applications extending to 5 MHz and above. The practical selection implication is to compare measured application losses and bias behaviour, not to assume that metal-composite inductors are restricted to lower switching frequencies. [4]

Selection and Application Guidelines for SMD Inductors

Start by deciding whether the component stores energy, establishes a reactive impedance, or suppresses unwanted signals. These functions lead to different priorities for current, loss, Q, self-resonance and environmental performance.

Overview of SMD Inductor Types

Table 1 compares the principal construction families. Frequency limits are deliberately not assigned to whole technologies: the usable band depends on the inductance value, geometry, material formulation and required circuit behaviour. The construction and application distinctions are illustrated by the power and RF references. [1][3][4]

Inductor type	Typical construction	Typical applications	Selection advantages	Principal limitations to check
Ferrite multilayer signal	Embedded printed conductors in a ferrite body	Signal filters, bias networks and compact chokes	Compact geometry and magnetic field confinement in suitable series	DC bias, frequency-dependent loss, Q and current rating
Ceramic multilayer RF	Embedded conductors in a non-magnetic ceramic body	RF matching and resonant networks	Small dimensions and controlled geometry	Q, tolerance, parasitics and self-resonance
Power multilayer	Embedded conductors designed for energy-storage duty	Compact DC/DC converters	Low-profile options and small footprint	Effective inductance, saturation and thermal limits at the chosen value
Wire-wound ferrite power	Wire on a ferrite core, open or magnetically shielded	Power conversion and differential filtering	Choice of conductor and magnetic geometry	Core loss, saturation, leakage field and thermal behaviour
Metal-composite moulded power	Winding surrounded by magnetic composite	Compact switching converters and high-current rails	Shielded constructions and application-specific loss optimisation	Bias-dependent inductance, AC loss, working voltage and temperature
Wire-wound RF	Wire on ferrite, ceramic or another suitable support	RF matching, resonant circuits and bias chokes	High-Q options and a broad choice of impedance responses	Performance at operating frequency, parasitics and DC-current limits
Thin-film and air-core RF	Patterned conductor or self-supporting winding	RF matching and precision reactive networks	Non-magnetic options avoid magnetic-core saturation	Available inductance, conductor heating, Q and layout sensitivity

Core Materials and Frequency Range

Core material affects permeability, magnetic loss, saturation and temperature dependence, but a material name is not a complete operating specification. Compare the finished component under the intended excitation, including DC bias and ripple, rather than ranking materials only by initial permeability or saturation flux density.

Ferrite formulations differ substantially. NiZn and MnZn labels describe broad material families rather than fixed frequency cut-offs. For energy-storage components, magnetic geometry and any deliberate gap also influence current capability. Metal-powder composites use distributed magnetic gaps, but their bias and loss characteristics still depend on composition, processing and geometry.

Table 2 replaces universal material rankings with the evidence needed for selection. A low-loss claim is meaningful only when frequency, flux excursion, waveform, temperature and bias conditions are comparable. Manufacturer loss models can help, provided they represent the intended operating conditions. [3][5]

Material or core class	Relevant behaviour	Design evidence to request	Scope limit
NiZn ferrite	Used in compact multilayer and signal magnetics	Inductance versus bias, impedance, Q and application loss data	No universal power-frequency ceiling applies to all formulations
MnZn ferrite	Used in power and filtering magnetics with formulation-dependent losses	Loss versus frequency, flux excursion and temperature; gap and bias information	Initial permeability alone does not establish converter suitability
Iron and alloy powder composites	Distributed-gap behaviour and current-dependent inductance	Full inductance-current curve, winding loss, core loss and thermal data	Different powders and binders are not interchangeable material grades
Nanocrystalline or amorphous systems	Properties depend strongly on alloy, physical form and processing	Finished-component loss, bias, temperature, reliability and production data	Research composite results do not automatically establish catalogue availability
Non-magnetic ceramic or air core	No magnetic-core saturation	Q, self-resonance, tolerance, conductor loss and temperature rise	Absence of magnetic saturation does not remove electrical or thermal limits

Nanocrystalline soft magnetic composites are an active research direction for reducing loss in miniature high-frequency magnetics. Their relevance to a production design depends on demonstrated finished-component performance and manufacturing maturity. They should not be presented as universally superior substitutes for established ferrite or metal-composite catalogue parts.

Key Electrical Parameters for Selection

The following parameters should be compared using their stated test conditions. Electrical and thermal ratings are interdependent, and manufacturer terminology is not completely uniform. [5]

• Inductance, L: Check tolerance, measurement frequency and amplitude, temperature dependence, and the inductance retained at operating current. In switching converters, the effective inductance determines current ripple and current slope.
• Saturation current, I_sat: Identify the specified inductance reduction and test temperature. This rating is a defined point on a magnetic characteristic, not necessarily an absolute damage threshold or a guarantee of acceptable ripple.
• Thermal current, I_rms, or rated current, I_R: Read the definition. A thermal rating may correspond to a stated self-heating rise on a specified test fixture; another rated-current designation may use a different limiting criterion.
• DC resistance, R_DC: Use the maximum value and account for operating temperature. This determines the resistive voltage drop and the DC-resistance contribution to winding loss.
• Self-resonant frequency, SRF: Check the chosen inductance value and measurement conditions. Near self-resonance, parasitic capacitance materially alters impedance and phase. Above the first resonance, a simple inductive model is generally inadequate.
• Quality factor, Q: Evaluate it at the operating frequency. Q is useful for matching and resonant networks, but small-signal Q does not replace a large-signal power-loss assessment.
• Voltage and temperature limits: Check working-voltage restrictions, winding insulation where specified, ambient range and maximum component temperature. A current rating does not establish voltage suitability.

As a concrete example of rating definitions, the WE-PMFI 74479332257110 datasheet specifies typical saturation currents of 9 A for a 10% inductance reduction and 21.5 A for a 30% reduction. The same device has a performance-rated current of 11.4 A at a 40 K rise under its stated test conditions. These are three different criteria for one part, not contradictory ratings. [6]

For a first estimate, the winding-loss contribution associated with DC resistance is given by Eq. (1), where P_DC is dissipated power, I_L,rms is the total RMS inductor current and R_DC(T) is winding resistance at component temperature T. Additional AC winding and magnetic losses must be evaluated separately. [5]

Do not use an ordinary small-signal SPICE model as a validated efficiency model unless its loss representation covers the application current, frequency and temperature. Small-signal equivalent series resistance measured without DC bias may not represent converter operation. Use suitable manufacturer loss tools or validated models, then confirm temperature and efficiency on hardware. [5]

Self-Resonance Depends on Circuit Function

For RF matching and reactive filters, select a component whose inductance and Q remain suitable across the operating band. For converter energy storage, evaluate switching frequency, significant harmonics and parasitic behaviour; a single fixed ratio between switching frequency and SRF is not a substitute for the component and converter requirements.

For a narrowband RF bias choke, high impedance near self-resonance may be exactly the desired property. Broadband choking instead requires adequate impedance over the complete rejection band and acceptable DC loss. The useful quantity is the measured impedance response, not an absolute prohibition on operation near SRF. [1]

A larger package can reduce DC resistance or improve thermal performance, but it does not necessarily increase SRF. Winding geometry and parasitic capacitance must also be considered. Verify all relevant curves when substituting a nominally equivalent part.

Application Priorities

Table 3 summarises the parameters that usually dominate each application. Inductance ranges are not imposed because the required value follows from circuit conditions, not the application label alone.

Application	How inductance is established	Priority checks	Candidate families
Buck or boost energy-storage choke	Converter topology, voltage range, switching frequency, ripple and control requirements	Effective L, peak current, RMS current, DCR, AC loss and temperature	Power multilayer, shielded wire-wound ferrite and metal-composite power
Input or output EMI filtering	Required impedance, source and load impedances, and attenuation band	Bias, impedance versus frequency, damping, DCR and thermal behaviour	Suitable signal or power inductors; ferrite suppressors where lossy suppression is intended
RF matching	Required reactance at the signal frequency	Q, tolerance, SRF, temperature stability and PCB parasitics	Wire-wound RF, ceramic multilayer and thin-film
Resonant LC filters	Target resonance or transfer function together with capacitance	L accuracy, Q, loss, linearity and loading	Suitable wire-wound or multilayer signal parts
Bias choke or RF decoupling	Required blocking impedance over the rejection band	Impedance response, DC current, DCR and interaction with bypass capacitors	RF chokes, ferrite-core or non-magnetic RF inductors

Practical Design-In Guidelines

For power chokes, check the minimum effective inductance at the highest relevant current and temperature. Include initial tolerance, switching-frequency tolerance, input-voltage range and converter current-limit behaviour. A fixed percentage margin below a catalogue saturation-current number cannot replace these checks.

Check RMS current and the complete thermal budget independently. Ambient temperature plus self-heating must remain within the permitted component temperature. PCB copper, nearby heat sources, enclosure conditions and airflow can make the application substantially different from the manufacturer’s current-rating fixture. [5]

For layout, follow the converter manufacturer’s reference arrangement, keep the switch-node connection compact, minimise the actual high-di/dt commutation loop, and separate sensitive signal routing from noisy conductors and magnetic fields. Do not enlarge a noisy switch-node copper area merely to improve heat spreading. Shielded inductors can reduce magnetic coupling, but they do not eliminate capacitive coupling or guarantee EMC compliance.

For demanding power designs, current-dependent characterisation can help verify whether catalogue curves represent the required operating region. Pulse testing and converter waveform measurements complement small-signal measurements; they do not remove the need to check repetitive losses and steady-state temperature.

Design, Procurement and Qualification Checklist

• Define the function: energy storage, reactive matching, resonance, differential filtering or lossy noise suppression. Do not substitute a ferrite bead for an energy-storage inductor solely because both have similar package dimensions.
• Compare nominal and effective inductance, tolerance, test frequency, DC-bias curves, temperature curves, maximum DCR, current definitions, loss data, SRF, Q and working-voltage restrictions.
• For catalogue parts, verify the complete ordering code, dimensions, maximum height, land pattern, termination finish, packaging, production status and change-notification arrangements. Approve alternate sources using electrical curves and application tests, not nominal L and package code alone.
• Consider a custom component only when catalogue choices cannot meet a defined combination of electrical, thermal, mechanical or qualification requirements. Agree test methods, acceptance limits, tolerances, traceability and change control before qualification.
• Check reflow profile, moisture handling where applicable, cleaning compatibility, board flex, vibration and permitted potting or coating processes. Encapsulation does not remove these requirements. [6]
• For automotive applications, review the supplier’s applicable AEC-Q200 qualification report, temperature range and customer-specific requirements. AEC-Q200 Rev E is a stress-test qualification document; AEC does not operate a component certification board. [7]
• For aerospace and other safety-critical applications, apply the programme-specific approval, screening, traceability and reliability requirements. Automotive qualification alone does not establish suitability for those applications. [6][7]
• Review system fault behaviour and the manufacturer’s application restrictions. Do not infer protective isolation, flame performance or safety approval from an ordinary chip-inductor current rating.

Quick Selection Example

Consider the original example: a buck converter operating from 5 V to 3.3 V at 2 MHz and supplying 1 A. Use an initial peak-to-peak ripple target of 30–40% of output current, corresponding to 0.30–0.40 A. The calculation below assumes ideal steady-state continuous conduction and approximately constant inductance over the ripple excursion; it is a starting point, not a complete converter design. [2]

Define V_in as input voltage, V_out as output voltage, f_s as switching frequency, D as duty ratio and ΔI_L as peak-to-peak inductor ripple current. For the ideal buck stage, Eq. (2) gives duty ratio and required inductance.

Eq. (2) gives D = 0.66 and L = 1.870 µH for 0.30 A ripple, or L = 1.403 µH for 0.40 A ripple. Thus the original recommendation of approximately 1–2 µH is directionally correct, but the calculated interval for the stated ripple target is approximately 1.4–1.9 µH.

Choose 1.5 µH as the nominal trial value. Eq. (3) calculates ripple and peak current, where I_out is average output current and I_L,pk is peak inductor current. The peak-current relation follows the buck power-stage method in Reference [2].

At the nominal conditions, ΔI_L = 0.374 A and I_L,pk = 1.187 A. For triangular ripple superimposed on the DC output current, Eq. (4) gives the total RMS current.

Eq. (4) gives I_L,rms = 1.006 A. A thermal-current rating of exactly 1 A is therefore not sufficient as a numerical acceptance criterion, even before allowing for application temperature, losses and tolerances. A preliminary saturation-current screening target of 1.5 A may be useful, but acceptance still depends on the specified inductance drop, temperature and effective L at the required peak current.

Repeat the calculation using minimum effective inductance. As an illustrative sensitivity check, if L is 1.2 µH rather than 1.5 µH, with the other nominal conditions unchanged, ripple increases to 0.468 A peak-to-peak. Then include maximum input voltage, minimum switching frequency, temperature, startup, load transients and the controller’s current-limit tolerances. Remain within the converter IC’s permitted inductance range. [2]

A 1008 or 1210 package is only an initial search constraint; enlarge it if the required inductance, current or temperature cannot be achieved. A useful caution appears in the WE-PMI catalogue: ordering code 74479787215A is listed at 1.5 µH with 1.4 A rated current, but only 1.1 A saturation current. Its saturation-current figure is below the calculated 1.187 A nominal peak, despite the apparently adequate rated-current figure. It therefore fails the preliminary saturation-current screen for this example. [3]

That same catalogue entry lists 70 mΩ DC resistance. Using this resistance only as an illustrative value, Eq. (1) gives approximately 71 mW of DC-resistance winding loss at the calculated RMS current. This is not total inductor loss: temperature-dependent resistance, AC winding loss and magnetic loss must still be included. [3][5]

Finally, compare suitable power multilayer, shielded ferrite wire-wound and metal-composite alternatives at the required effective inductance. Use application-relevant loss data or models, then measure current ripple, efficiency, component temperature and transient behaviour on the intended PCB. Retain the smallest part only if it also meets the electrical, thermal and reliability requirements.

Conclusion

SMD construction supports automated assembly and compact layouts, but the package does not determine the component’s electrical role. Multilayer, wire-wound, ceramic RF and metal-composite power inductors must be compared within the intended application class.

Successful selection combines effective inductance, peak and RMS current, loss, self-resonance, Q, voltage, temperature and assembly constraints. Treat magnetic shielding separately from encapsulation, and qualification separately from approval for the final application.

The worked example demonstrates the essential distinction: a part can meet a nominal inductance and thermal-current requirement yet fail the saturation-current screen. Datasheet review, worst-case calculation and validation on the intended PCB are all necessary before approving the component or an alternative source.

FAQ

These answers summarise the construction, rating and selection distinctions that most often affect design and purchasing decisions.

How to Select and Validate an SMD Inductor

This procedure turns the construction comparison into a repeatable design-in workflow. Use it for initial selection and repeat the relevant checks when qualifying an alternate source.

Further Reading

These articles expand the underlying principles, material-development context and validation methods without replacing component-specific datasheet review.

• Inductor and Choke: Principles and Applications — background on inductance, energy storage, equivalent circuits and choke functions.
• Nanocrystalline Cores for Low-Loss MHz Chip Inductors — research context for emerging soft magnetic composites and their manufacturing constraints.
• Experimental Evaluation of Wear Failures in SMD Inductors — accelerated-stress observations relevant to mission-profile and termination-reliability discussions.
• One-Pulse Characterization of Nonlinear Power Inductors — a time-domain approach to evaluating current-dependent inductance.

References

The numbered sources support the application guidance, rating definitions, catalogue examples and worked calculation.

1. Coilcraft, Inductors as RF Chokes, Coilcraft application note, undated.
2. Brigitte Hauke, Basic Calculation of a Buck Converter’s Power Stage, Texas Instruments, SLVA477C, 2026 revision.
3. Würth Elektronik, WE-PMI Power Multilayer Inductor, Würth Elektronik product catalogue, undated.
4. Coilcraft, XGL Family Moulded Power Inductors, Coilcraft product catalogue, undated.
5. Coilcraft, Current and Temperature Ratings, Coilcraft application note, undated.
6. Würth Elektronik, WE-PMFI Power Molded Flatwire Inductor, Order Code 74479332257110, Würth Elektronik datasheet, revision 001.000, 2025.
7. Automotive Electronics Council, AEC-Q200 Rev E: Stress Test Qualification for Passive Components, Automotive Electronics Council, 2023.