EMI Filter Design Calculator

Calculator inputs

This page uses a stacked layout for sections, while the input fields switch to three columns on large screens, two on medium screens, and one on mobile.

Topology

LC uses a second-order estimate. π uses a third-order estimate.

Capacitor selection mode

Auto mode keeps the design under the entered capacitor-current limit.

Line voltage RMS (V)

Load current (A)

Line frequency (Hz)

Noise frequency (kHz)

Target attenuation (dB)

Design margin (dB)

Source impedance (Ω)

Load impedance (Ω)

Preferred shunt capacitance (nF)

For π filters, this is the capacitance of each shunt branch.

Maximum capacitor current (A)

Example data table

Case	Topology	Noise frequency	Goal + margin	Shunt capacitance	Approx. inductance	Cutoff	Comment
Offline converter input	LC	150 kHz	46 dB	220 nF	1.03 mH	10.6 kHz	Balanced first-pass filter for differential noise reduction.
Higher current front end	π	250 kHz	54 dB	100 nF each	2.27 mH	31.5 kHz	Steeper slope, but damping and parasitics matter more.
Industrial AC stage	LC	300 kHz	50 dB	47 nF	3.79 mH	16.9 kHz	Smaller capacitance pushes the inductor value upward.

Formula used

Total attenuation goal: A_total = A_target + A_margin
Required cutoff frequency: f_c = f_noise / 10^{A_total / (20n)}, where n = 2 for LC and n = 3 for π.
Series inductance: L = 1 / ((2πf_c)² C_eff)
Equivalent capacitance: C_eff = C for LC, and C_eff = C / 2 for equal-capacitance π filters.
Resonant frequency: f_r = 1 / (2π√(LC_eff))
Characteristic impedance: Z₀ = √(L / C_eff)
Capacitor current estimate: I_C = 2πf_lineV_lineC for each shunt branch.
Ideal insertion loss estimate: IL ≈ 10 log₁₀(1 + (f / f_c)²ⁿ)
Damping resistance estimate: R_d ≈ √(L / C_eff)

These formulas create a practical first-pass estimate. Real compliance results depend on parasitics, layout, capacitor ESR and ESL, inductor winding resistance, saturation, enclosure behavior, and the exact measurement setup.

How to use this calculator

Select the intended topology. Use LC for a simpler stage, or π for a steeper estimated roll-off.
Enter the line voltage and frequency so the page can estimate capacitor current at the operating mains condition.
Enter the dominant conducted-noise frequency, usually around the switching frequency or one troublesome harmonic band.
Set the attenuation target and a margin. The margin helps absorb modeling error and component tolerance.
Add source and load impedance values. They affect loading, damping, and how closely the practical response follows the ideal curve.
Choose auto capacitor mode if you want the design limited by allowable capacitor current. Choose manual if you already know the capacitor value.
Press Calculate filter. The result panel appears above the form with component values, design notes, and the attenuation graph.
Download CSV or PDF after a calculation if you want to save the design snapshot for review.

FAQs

1) What does this calculator design?

It provides a first-pass differential-mode EMI input filter estimate. It sizes shunt capacitance, series inductance, cutoff frequency, damping resistance, and ideal insertion loss from your attenuation goal and electrical limits.

2) Can this replace compliance testing?

No. It is a design aid, not a compliance verdict. Laboratory results still depend on layout, cable routing, parasitic coupling, magnetic leakage, common-mode paths, enclosure behavior, and the exact LISN and analyzer setup.

3) When should I choose LC instead of π?

Choose LC when you want a simpler filter with fewer parts and less resonance management. Choose π when you need a steeper estimated slope and can validate damping, capacitor stress, and parasitic behavior carefully.

4) Why is capacitor current included?

Large shunt capacitance can create significant line-frequency reactive current. This matters for component heating, ratings, and practical mains behavior. Auto mode uses your current limit to keep the suggested capacitance more realistic.

5) What if the required inductance looks too large?

That usually means the selected capacitance is small, the attenuation goal is aggressive, or the cutoff must be pushed very low. Consider a π filter, more damping, staged filtering, or a higher allowable shunt capacitance.

6) Does it handle common-mode noise?

Not directly. This page is aimed at first-pass differential-mode filtering. Common-mode issues usually require common-mode chokes, Y capacitors, careful grounding strategy, shielding, and measurement data from the real hardware.

7) Why is a damping resistance estimate shown?

High-Q filters can ring, overshoot, or shift response when real source and load impedances change. The damping estimate gives a starting point for an RC damping branch or practical loss target during prototype tuning.

8) Why can measured attenuation differ from the graph?

The graph is an idealized estimate. Real parts add ESR, ESL, saturation, winding resistance, and stray coupling. Source impedance, load impedance, PCB layout, and noise current paths can change the measured response significantly.