Lumped Element Filters: L-C Design and Selection

Lumped element filters use inductors and capacitors, with resistive losses or deliberate damping, to pass wanted frequencies and attenuate unwanted signals. This article explains lumped element L-C filter construction, response types, practical limitations and selection criteria for engineers designing RF and signal-path circuits, with additional cautions for power-filter applications.

Knowles A-Series RF Attenuators Target Compact 20 GHz Designs

Electrification Raises Demands on DC-Link Capacitors

Knowles Cornell Dubilier 105C Flatpack Aluminum Electrolytic Capacitors Target Low-Profile High-Density Power Designs

Key Takeaways

A useful lumped element filter specification describes the complete installed response, not just nominal inductance, capacitance or operating frequency.

Lumped element L-C networks can implement lowpass, highpass, bandpass and bandreject responses; component arrangement, termination impedances and losses determine the actual response.
Ideal LC resonance is a starting point, not a universal formula for the cutoff frequency of a loaded filter.
Component quality factor, self-resonance, PCB parasitics and input-to-output coupling limit insertion loss and stopband rejection.
Published manufacturer frequency and bandwidth ranges describe design capabilities, not guaranteed performance at every combination of frequency, bandwidth, power and size.
Validate the chosen components and layout using realistic models, tolerance analysis and prototype measurements under the intended operating conditions.

The Basics of Lumped Element L-C Filter Construction

Lumped element filters are passive networks built from inductors, capacitors and, where required, resistors. Their circuit model treats these elements as localised impedances; the model must be extended when component or interconnection parasitics materially affect the response. This distinction becomes increasingly important as operating frequency rises. [1, 2]

Inductors, capacitors and resonators

Inductors and capacitors provide complementary frequency-dependent impedances. Combining them in series and shunt branches produces the required filtering behaviour. Resonant LC sections are useful building blocks, but a complete filter must also account for how those sections interact with each other and with the source and load. Figure 1 retains the original generic structure. [1, 2]

Generic lumped element L-C filter circuit structure using inductors and capacitors — Figure 1. An example of a generic lumped element L-C filter structure. Source: Knowles Precision Devices.

For an ideal isolated LC resonator, its resonant frequency f₀, in hertz, is given by Eq. (1), where L is inductance in henries and C is capacitance in farads. This relation describes ideal resonance; the cutoff frequency and passband shape of a terminated filter additionally depend on topology, damping and termination impedances. [3]

f_{0} = \frac{1}{2\pi\sqrt{LC}} \tag{1}

Poles and zeros control the response

A filter transfer function H(s) describes the relationship between its output and input using the complex-frequency variable s. After common numerator and denominator factors have been cancelled, zeros are roots of the numerator and poles are roots of the denominator. A pole is therefore not simply a frequency at which the measured response reaches a maximum. The sinusoidal frequency response is obtained by evaluating the transfer function at s = jω, where j is the imaginary unit and ω is angular frequency. [3]

Component values, topology and loading establish the pole and zero locations. Suitable transmission zeros can improve rejection at selected frequencies, while pole locations govern the transition response and influence transient behaviour. Not every LC network has finite-frequency transmission zeros. The companion article on filter poles and zeros introduces this design approach. [3]

Lumped Element Filter Characteristics

Lumped element construction offers considerable flexibility in electrical response and physical implementation. Its advantages must nevertheless be evaluated against the required bandwidth, loss, power handling and environmental conditions rather than assumed from the technology name alone. [1, 2]

Size, construction and environmental limits

At lower RF frequencies, lumped element filters can provide a compact alternative to implementations based on physically larger resonators, such as cavity or waveguide structures. This is a relative comparison between implementations, not a claim that lumped element filters are always the smallest solution or that they operate only at low frequencies. [1]

Component materials, assembly methods, packaging and thermal design can be selected to suit the application. Custom construction may accommodate demanding input-power, temperature or environmental requirements, but these capabilities require an explicit specification and qualification. They do not follow automatically from using an LC network. [1]

Loss, self-resonance and parasitic coupling

Real inductors and capacitors introduce losses and parasitic reactances. Inductor winding resistance and other frequency-dependent losses increase insertion loss, while parasitic capacitance and capacitor parasitic inductance alter the intended impedance behaviour. Component quality factor, Q, must therefore be considered at the operating frequencies rather than inferred solely from nominal component values. [2]

PCB traces, pads, vias and return paths also form part of the filter. At higher frequencies, a circuit assembled from nominally correct L and C values can differ substantially from an ideal simulation. Use measurement-based component models where available, include relevant interconnections, and check the stopband as well as the passband. [2]

Keep input and output routing separated sufficiently to avoid unintended capacitive or inductive coupling around the filter. Short connections and controlled return paths help preserve the intended network response. The broader RF inductors and filters guide discusses component behaviour and implementation considerations.

Source and load impedances are design inputs

Insertion loss and passband shape depend on the impedances connected to the filter. A response specified in a 50 Ω source and 50 Ω load system does not automatically describe performance between different or frequency-dependent impedances. Define the intended terminations before selecting component values, and preserve those conditions when comparing simulations, datasheets and measurements. [4]

The RF and signal-path discussion here is not a complete power-converter input-filter design method. A power filter also needs current, voltage, thermal and dynamic-interaction checks, including the possibility that the filter changes converter stability. Mains-connected filtering additionally requires application-appropriate safety-rated components and an assessment of the complete equipment safety requirements.

What Filter Types Can Be Built Using Lumped Element Construction?

Lumped element networks can realise all four common magnitude-response categories: lowpass, highpass, bandpass and bandreject. The response category describes which frequencies pass; the response approximation describes the trade-off between flatness, selectivity and time-domain behaviour. [2, 3]

Lowpass filters pass lower frequencies and attenuate frequencies above the specified transition region.
Highpass filters attenuate lower frequencies and pass frequencies above the specified transition region, within the practical operating range of the implementation.
Bandpass filters pass a defined frequency band and attenuate frequencies on both sides of that band.
Bandreject filters attenuate a defined frequency band while passing frequencies below and above it; a narrow rejection band is commonly called a notch.

A bandpass response can be formed by combining highpass and lowpass behaviour when their passbands overlap. Directly connecting passive sections, however, changes their loading, so the combined network must be analysed rather than treated as two independent ideal responses. Resonator-based bandpass networks provide another implementation route. See Basic Filter Circuits Explained for the underlying circuit arrangements. [2, 3]

Choose the response approximation

Butterworth responses prioritise a maximally flat passband magnitude. Chebyshev Type I responses trade passband ripple for a sharper transition. Elliptic responses use transmission zeros to obtain sharp transitions, with ripple in both passband and stopband. Bessel responses prioritise group-delay flatness and waveform preservation rather than the sharpest magnitude cutoff. Choose according to the permitted amplitude and phase distortion, not selectivity alone. [2, 3]

Define bandwidth and center frequency consistently

For a bandpass filter, define the lower and upper band edges f_L and f_H using an agreed attenuation criterion. Bandwidth BW is their difference. Fractional bandwidth, FBW, is bandwidth divided by the stated center frequency f₀, as shown in Eq. (2). The edge frequencies and center frequency must use the same frequency units. [4]

\mathrm{BW} = f_{H} – f_{L}, \qquad \mathrm{FBW} = \frac{f_{H} – f_{L}}{f_{0}}, \qquad \mathrm{FBW}_{\%} = 100\,\mathrm{FBW} \tag{2}

Center frequency may be specified using an arithmetic or geometric convention. State which convention applies, particularly for wideband filters, because the resulting fractional bandwidths differ. Fractional bandwidth can exceed 100%; this describes the ratio between bandwidth and the chosen center frequency, not an impossible operating condition. [4]

Lumped Element Filter Capabilities

The original Knowles Precision Devices capability illustration is retained in Figure 2. The accompanying ranges are manufacturer-published examples of available design capabilities, not general physical limits for lumped element filters or guarantees that every listed extreme is available simultaneously. [1]

Manufacturer chart comparing filter frequency coverage and fractional bandwidth — Figure 2. Filter frequency and fractional-bandwidth capability chart from the original article. Source: Knowles Precision Devices.

Bandpass filters with narrow to moderate bandwidths: 1% to 70% FBW, with center frequencies from 10 MHz to 7 GHz.
Bandpass filters with extra-wide bandwidths: 70% to 175% FBW, with center frequencies from 20 MHz to 11 GHz.
Lowpass filters: a published frequency range from 10 MHz to 22 GHz.
Highpass filters: a published frequency range from 10 MHz to 10 GHz.
Bandreject filters: a published frequency range from 20 MHz to 6 GHz, with narrowband or wideband implementations.

These envelopes do not specify insertion loss, return loss, rejection depth, transition width, package size or power rating. Lowpass and highpass filters should be specified through their passband and stopband edges rather than a center frequency. For bandreject filters, state the rejected band and the required attenuation explicitly. Obtain a part-specific datasheet or an agreed custom specification before using any capability range as a design commitment. [1, 4]

Similarly, high-power, high-temperature or harsh-environment operation requires an implementation designed and qualified for those conditions. The frequency range alone provides no evidence of the applicable power, thermal or environmental rating. [1]

Design and Selection Checklist

Start with an electrical response specification, then select components and packaging that can meet it under realistic conditions. The following checklist applies primarily to RF and signal-path lumped element filters; power and safety requirements need additional application-specific assessment. [1, 2, 4]

Define the useful signal band, passband edges, allowable insertion loss and ripple, stopband frequencies and required rejection. Add phase or group-delay limits where waveform fidelity matters.
Specify source and load impedances, permitted mismatch and the measurement reference conditions. Compare return loss and insertion loss under consistent conditions.
Choose the response approximation and order. Use ideal synthesis as the starting point, then substitute available component values and rerun the analysis.
Compare inductance and capacitance tolerances, component Q and losses over frequency, self-resonant behaviour, temperature dependence and available simulation models.
Check voltage, current, input-power and temperature limits for the assembled filter, not only individual components. For power applications, include bias, saturation, ripple heating and damping considerations.
Include PCB traces, pads and return paths in the model where they affect the result. Prevent coupling that allows signals to bypass the intended filter network.
Compare catalog and custom options against the same specification. For custom filters, agree the response limits, mechanical interface, environmental tests, acceptance measurements and change-control requirements before ordering.
Confirm qualification and safety requirements for the actual application. An RF filter performance specification is not evidence of suitability for mains connection or a safety-critical installation.
Build and measure a representative prototype. Verify passband loss, matching and stopband rejection, then check component tolerances and relevant operating conditions before releasing the design.

Conclusion

Lumped element L-C filters offer a flexible route to lowpass, highpass, bandpass and bandreject responses. Successful implementation depends on the complete network: component losses, source and load impedances, parasitic effects, layout and operating conditions all influence the installed response.

Use ideal resonance and synthesis calculations to establish a starting point, not as substitutes for a terminated circuit model and prototype verification. Manufacturer capability charts help identify candidate technologies, while part-specific specifications and qualification establish whether a particular filter is suitable for the application.

FAQ

These questions address common design and procurement decisions for lumped element filters.

What is a lumped element L-C filter?

A lumped element L-C filter is a passive network that uses inductors and capacitors to pass selected frequencies and attenuate others. Resistors or component losses provide damping. The approach models components as localised impedances, but practical designs must include relevant component and PCB parasitics when these affect the required frequency response.

Does the LC resonance equation give the filter cutoff frequency?

The ideal LC resonance equation gives the resonant frequency of an ideal isolated inductor-capacitor combination. It does not automatically give the cutoff frequency of a complete filter. The terminated network also depends on topology, source and load impedances, and damping, so the required passband and stopband edges must be checked using the complete circuit response.

Can lumped element filters operate at microwave frequencies?

Yes. Manufacturer capability ranges include lumped element implementations at microwave frequencies, but those ranges are not universal technology limits or guarantees for every specification combination. At higher frequencies, component losses, self-resonance, PCB interconnections and unintended coupling become important design constraints. Confirm the required response using a part-specific specification and representative measurements.

Why does an LC filter behave differently from its ideal simulation?

Ideal simulations can omit winding resistance, capacitor and inductor parasitics, PCB traces, pads and return paths. The actual source and load impedances may also differ from the assumed terminations. These differences change loss, matching, resonance and stopband behaviour. Use realistic component models and a representative layout, then verify the assembled filter experimentally.

When should I choose a custom filter instead of a catalog part?

Choose between catalog and custom filters by comparing both against the same electrical, mechanical and environmental requirements. A custom implementation is worth considering when available catalog parts cannot meet the required combination. Agree response limits, termination conditions, input power, packaging, qualification tests and acceptance measurements before ordering rather than relying on a broad supplier capability chart.

How to Specify and Validate a Lumped Element Filter

Use this procedure to turn the required frequency response into a testable filter specification.

Specify and validate an RF or signal-path lumped element filter using realistic components, terminations and prototype measurements.

Define the response and terminations
Specify the useful signal band, passband loss and ripple, stopband rejection, source impedance and load impedance. State the band-edge criterion and center-frequency convention for bandpass filters.
Select the topology and components
Choose the response category, approximation and order. Calculate initial values, substitute available components and compare losses, tolerances, self-resonant behaviour and operating ratings.
Model the practical implementation
Simulate the terminated network with realistic component models and relevant PCB interconnections. Examine passband matching, stopband behaviour and the effect of component tolerances.
Verify the assembled filter
Measure a representative prototype under the specified termination conditions. Confirm insertion loss, return loss and rejection, then evaluate the required power, temperature and environmental conditions before design release.

References

The following sources support the construction principles, design guidance and attributed manufacturer capability ranges discussed in this article.

Peter Matthews, Lumped Element Filters: A Compact, Low-Frequency Filtering Option, Knowles Precision Devices Blog, 2026.
Coilcraft, Passive LC Filter Design and Analysis, application note, Document 1021, publication year not stated.
Analog Devices, A Filter Primer, Analog Devices technical article, 2002.
Peter Matthews, Filter Basics Part 3: Five Key Filter Specifications to Understand, Knowles Precision Devices Blog, 2022.

Latest technical review and revision: 5 November 2026. Updated definitions, application scope, practical design guidance, capability interpretation and references; original figures retained.

Source: Knowles Precision Devices