EMI Filter Design: C, LC, Pi & T-Section Topology Guide

A power supply design requires input filtering to suppress conducted EMI across a 10 MHz to 100 MHz frequency range. An engineer specifies a simple LC filter but finds the measured attenuation falls short of the 60 dB requirement at 10 MHz. The design must be revised, but the engineer is uncertain whether to add components to the existing LC filter or redesign with a Pi configuration. This scenario illustrates a common challenge: many engineers understand that filters are necessary but lack clarity on how different configurations affect performance. Understanding filter types, their characteristics, and selection criteria enables designs that meet performance requirements efficiently.

Understanding Filter Configurations

Fundamental Concept

A passive EMI filter uses capacitors (and often inductors) to suppress unwanted frequencies while passing desired frequencies. Different configurations achieve different attenuation characteristics. Selecting the appropriate configuration type requires understanding how each topology impacts performance and cost.

Filter configuration determines:

• Attenuation slope:How quickly attenuation increases with frequency
• Frequency response:Attenuation at specific frequencies
• Insertion loss:Overall signal attenuation introduced by the filter, covering both pass-band and stop-band ranges
• Impedance matching:Interaction with source and load impedance
• Cost and complexity:Component count and design effort

Filter Type 1: Simple Capacitor (C Filter)

A simple C filter consists of a single capacitor connected to ground, providing a shunt impedance path at the point where filtering is needed.

Characteristics:

• Lowest cost and simplest design
• Attenuation improves with increasing frequency
• Poor attenuation at low frequencies
• Effective only below the capacitor’s self-resonance frequency (SRF); performance degrades above SRF due to parasitic inductance

Typical Attenuation: Approximately 20 dB/decade within the capacitive region below SRF

When Appropriate: Local decoupling, high-frequency bypass, applications where low-frequency attenuation is not critical

Limitations: Cannot meet demanding EMI suppression requirements across wide frequency ranges

Filter Type 2: LC Filter

An LC filter combines a series inductor with shunt capacitors, forming a resonant network that provides improved attenuation compared to C filters.

Configuration: Series inductor in the main signal path, with capacitor(s) connected to ground (shunt)

Filtering Principle: The inductor and capacitor form a second-order low-pass resonant network. Below the cutoff frequency, the series inductor presents low impedance and the shunt capacitor presents high impedance, allowing desired signals to pass through. Above the cutoff frequency, the inductor offers high series impedance while the shunt capacitor provides a low-impedance path to ground for noise. The LC natural resonant frequency is where resonance-related peaking may occur.

Typical Attenuation: Better than C filter across frequency range, with a theoretical -40 dB/decade slope

Advantages:

• Improved low-frequency attenuation compared to C filter
• Reasonable cost with moderate complexity
• Suitable for many power supply applications

Disadvantages:

• May exhibit resonance peaks if not properly damped
• Performance depends on impedance matching
• Requires careful component selection

When Appropriate: Power supply input filtering, basic EMI suppression, moderate attenuation requirements

Filter Type 3: Pi (Π) Filter

A Pi filter configuration uses three elements: an input capacitor, a series inductor, and an output capacitor — arranged in the shape of the Greek letter π.

Filtering Principle: Dual-stage filtering provides attenuation at both input and output sides. The input capacitor addresses noise from the source; the output capacitor provides additional attenuation and voltage smoothing to the load.

Typical Attenuation: Approximately 60 dB/decade theoretical slope, providing superior attenuation compared to LC

Advantages:

• Better attenuation than LC filter
• Symmetrical topology, adaptable to different source-load impedance conditions
• Suitable for demanding EMI suppression
• Commonly used in power supply design

Disadvantages:

• More components than LC (higher cost)
• Larger physical size
• Requires careful impedance analysis

When Appropriate: Power supply entry filtering, demanding EMI suppression, wide frequency range attenuation

Filter Type 4: T-Section Filter

A T-section filter is an independent third-order low-pass topology built with two series inductors and one intermediate shunt capacitor. It delivers excellent noise attenuation under high-impedance source-load conditions.

Configuration: Series Inductor-Shunt Capacitor-Series Inductor (L-C-L) arrangement

Characteristics:

• High attenuation slope comparable to Pi filter
• Excellent noise rejection under high-impedance source-load conditions
• Relatively complex and costly compared with C-filter and LC filter

Typical Attenuation: Approximately 60 dB/decade theoretical slope

When Appropriate: Extreme EMI suppression requirements, military/aerospace applications, situations where Pi filter attenuation is insufficient, and high-impedance source-load interfaces

Limitations: High cost and complexity mean it is only justified for highly demanding requirements

Design Methodology

Step 1: Define Attenuation Requirement

Specify the target attenuation (dB) at each critical frequency based on EMI suppression or power conditioning needs. Example: “60 dB attenuation at 1 MHz to 100 MHz”

Step 2: Determine Frequency Range

Identify the frequency band requiring suppression. EMI suppression typically spans 0.15 MHz to 30 MHz (conducted) or higher frequencies for radiated emissions.

Step 3: Calculate Component Values

For a second-order LC low-pass filter, the general cutoff-frequency formula is:

fc = 1 / (2π√(LC))

The following set of equations applies only when both source and load impedances are matched to characteristic impedance Z₀ (e.g. 50 Ω test-system environment):

L = Z₀ / (2πfc) and C = 1 / (2πZ₀fc)

Where Z₀ is the characteristic impedance.

Note: Practical power-supply EMI filters often work under non-50-Ω operating conditions; direct application of these formulas may yield non-optimal results, simulation is strongly recommended.

Step 4: Verify Impedance Matching

Confirm how the filter impedance interacts with source and load impedances in your application. Impedance mismatch can degrade attenuation or introduce resonant gain peaks.

Step 5: Analyze Frequency Response

Calculate attenuation at frequencies throughout the operating range. Use simulation tools or analytical methods to verify performance.

Step 6: Select Components and Verify

Choose specific capacitors and inductors that meet the calculated values, considering tolerances, temperature effects, self-resonance frequency, and availability.

Step 7: Implement and Verify

Design PCB layout with proper grounding and isolation. Measure actual performance using a network analyzer or equivalent equipment.

Conclusion

Capacitor filter configurations range from simple single-capacitor designs to complex multi-stage networks. Configuration type fundamentally affects attenuation characteristics, cost, and complexity. C filters are simplest but provide limited attenuation. LC filters deliver better low-frequency attenuation via second-order topology, while requiring careful suppression of resonant peaks. Pi filters provide dual-stage attenuation suitable for demanding EMI suppression. T-section filters deliver excellent noise attenuation under high-impedance conditions for extreme-specification applications. Impedance matching is critical to all filter designs — impedance mismatch degrades performance significantly. Component selection, PCB layout, and proper implementation are essential to achieving designed performance in actual systems.

Frequently Asked Questions

Q: Why does filter performance depend on impedance matching? A filter is designed for specific source and load impedance (typically 50 Ω). If actual impedance is different, reflections and standing waves degrade attenuation. A Pi filter designed for 50 Ω will underperform in a 1 kΩ circuit. Always verify impedance matching for proper filter performance.

Q: What is the difference between Pi and T-section filters? Pi filter uses two shunt capacitors with one series inductor (C-L-C topology). T-section filter uses two series inductors with one intermediate shunt capacitor (L-C-L topology). Both are third-order topologies with identical theoretical 60 dB/dec roll-off. T-section delivers superior practical attenuation under high-impedance source-load conditions. Use Pi for most low-to-medium impedance applications; use T-section for high-impedance interfaces and extreme-specification scenarios.

Q: How do I know if my filter design is working? Measure frequency response using a spectrum analyzer or network analyzer. Verify attenuation at target frequencies, insertion loss at passband, and impedance matching (return loss). Compare measurements to theoretical design. Significant discrepancies suggest layout problems or component tolerance issues.

Q: Why is PCB layout important for filters? Parasitic inductance introduced by traces, vias, and component leads degrades filter performance, especially at high frequencies. Poor layout can shift resonant frequency by 10-20 % or more. Keep component leads short, adopt wide traces to minimize parasitic inductance, and maintain solid ground-plane continuity.

Q: Should I use Pi or T-section filter for my power supply? It depends on your operating-impedance environment and attenuation requirements. If a Pi filter meets your attenuation target, select Pi — it is simpler and lower-cost for typical low-impedance power interfaces. T-section is preferred when operating with high-impedance source-load interfaces, rather than for pursuing higher roll-off slope. Always select the simplest topology that satisfies your EMI requirements.

Next Steps

To systematically design and optimize capacitor filter configurations for your application, you may refer to filter type comparisons, design equations, component selection procedures, impedance analysis methods, frequency response plotting examples, and design worksheets for C, LC, Pi, and T-section configurations. For further assistance, please contact LCA’s filter design specialists.

Technical guidance in this article reflects general filter design principles and methodology. Specific requirements depend on your frequency range, attenuation target, impedance, and system architecture. Always verify component performance and filter frequency response through measurement before finalizing production design.