The LC filter corner frequency (also called the cutoff or -3 dB frequency) is the specific point where the filter's output power drops by half, marking the boundary between the frequencies that pass through and those that are attenuated. In a real circuit or installation, setting this frequency correctly dictates whether your switch-mode power supply outputs clean DC or a noisy mess, and whether your Class-D audio amplifier produces crisp sound or destroys a tweeter with high-frequency switching noise. If you set it too high, ripple and noise slip through; set it too low, and you choke the transient response or trigger control loop instability.
Where f_c is the corner frequency in Hertz, L is inductance in Henries, and C is capacitance in Farads.
The Math and a Worked Numeric Example
To understand how this works on the bench, let's design the output filter for a 500 kHz synchronous buck converter stepping 12V down to 3.3V to power a sensitive RF module. A standard power electronics rule of thumb is to set the LC filter corner frequency to roughly 1/10th of the switching frequency to adequately attenuate the switching ripple while maintaining a fast transient response.
- Switching Frequency (f_sw): 500 kHz
- Target Corner Frequency (f_c): 50 kHz
- Chosen Inductor (L): 10 µH (a standard shielded drum-core part)
We need to find the required capacitance (C). Rearranging the core formula to solve for C gives us:
C = 1 / [(2π × f_c)² × L]
Plugging in our real-world values:
- Calculate the angular frequency squared: (2 × π × 50,000)² ≈ 9.8696 × 10¹⁰
- Multiply by inductance: 9.8696 × 10¹⁰ × (10 × 10⁻⁶) = 986,960
- Divide 1 by the result: 1 / 986,960 ≈ 1.013 × 10⁻⁶ Farads
The math tells us we need 1.013 µF. However, here is where bench experience overrides textbook math. If you select a standard 0805 1 µF Y5V ceramic capacitor, the DC bias effect will decimate your capacitance. At a 3.3V DC bias, a Y5V dielectric can lose up to 80% of its nominal capacitance, dropping to 0.2 µF. This shifts your actual corner frequency up to roughly 112 kHz, ruining your ripple attenuation. Instead, you must select a 2.2 µF X7R ceramic capacitor in a 1206 or larger package, which retains over 90% of its capacitance at 3.3V bias, keeping your corner frequency firmly near the 50 kHz target.
What People Commonly Confuse It With
When reading datasheets or application notes from manufacturers like Analog Devices or Texas Instruments, engineers frequently mix up three related but distinct concepts:
1. Corner Frequency vs. Resonant Frequency
In an ideal, completely lossless LC circuit, the corner frequency and the resonant frequency are mathematically identical. But in reality, equivalent series resistance (ESR) in the capacitor, DCR in the inductor, and the actual load resistance damp the circuit. The corner frequency defines the -3 dB power drop point of the filter's transfer function, while the resonant frequency is where the inductive and capacitive reactances perfectly cancel out. In heavily damped circuits, the resonant peak disappears entirely, but the -3 dB corner frequency still exists.
2. LC Roll-off vs. RC Roll-off
An RC filter attenuates signals at -20 dB per decade. An LC filter attenuates at -40 dB per decade. Think of an RC filter's roll-off as a single lane of traffic merging, slowing high-frequency signals gradually. An LC filter's -40 dB/decade roll-off is like a concrete barricade that stops that same high-frequency traffic twice as fast. This steeper slope is exactly why LC filters are mandatory for switching power supplies, where a -20 dB drop isn't nearly aggressive enough to hide 500 kHz noise from a 3.3V logic rail.
3. The -3 dB Point vs. The Passband Edge
Amateurs often assume the corner frequency is where the signal "stops." In reality, at the corner frequency, the signal is still passing through at 70.7% of its original voltage amplitude (-3 dB). If your design requires a signal to be attenuated by 99%, you must calculate how many decades past the corner frequency your target noise sits.
Where You Meet This in Practice
You will encounter LC filter corner frequency calculations across several distinct domains of electrical engineering. The table below outlines typical target ranges and the primary design constraint for each application.
| Application | Typical f_c Range | Primary Design Constraint |
|---|---|---|
| SMPS Output Filtering (Buck/Boost) | 10 kHz – 100 kHz | Must be low enough to attenuate switching ripple, but high enough to avoid control loop phase margin issues. |
| Class-D Audio Amplifiers | 40 kHz – 80 kHz | Must pass 20 kHz audio cleanly while aggressively attenuating the 300+ kHz PWM carrier. |
| EMI/RFI Input Filters | 100 kHz – 1 MHz | Tuned specifically to block conducted emissions measured during FCC/CE compliance testing. |
| Passive Audio Crossovers | 200 Hz – 5 kHz | Must match the specific impedance curve of the speaker driver, not just its nominal 8-ohm rating. |
For deeper mathematical modeling of these passive networks, resources like Electronics Tutorials provide excellent baseline transfer functions, while All About Circuits offers great visual breakdowns of Bode plots showing exactly how the phase shifts as you approach and pass the corner frequency.
LC Filter Corner Frequency FAQ
How does load resistance affect the LC filter corner frequency?
Load resistance does not significantly change the mathematical corner frequency (the -3 dB point), but it drastically changes the filter's Q-factor (quality factor) and peaking behavior. If the load resistance is very high (a light load), the LC filter will exhibit a massive resonant peak right at the corner frequency, potentially amplifying noise instead of attenuating it. If the load resistance is low (a heavy load), it dampens the circuit, flattening the peak but potentially softening the -40 dB/decade roll-off into something that looks more like a -20 dB slope. To fix light-load peaking, designers often add a small RC snubber network in parallel with the output capacitor to provide high-frequency damping without wasting DC power.
Why is my LC filter ringing at the corner frequency?
Ringing occurs when a fast transient step (like a microcontroller waking up and drawing 500mA in a microsecond) excites the LC tank at its resonant frequency. Because the corner frequency and resonant frequency are closely tied in low-damping circuits, the energy bounces back and forth between the inductor's magnetic field and the capacitor's electric field. You can measure this on an oscilloscope by probing the output with a ground-spring (not a long pigtail ground clip, which adds parasitic inductance and causes false ringing readings). To stop it, you must increase the ESR of the circuit intentionally, either by choosing an electrolytic or polymer capacitor instead of a pure ceramic, or by adding a dedicated damping resistor.
Can I use the LC filter corner frequency formula for high-pass filters?
Yes. The formula f_c = 1 / (2π√LC) calculates the -3 dB corner frequency for both LC low-pass and LC high-pass topologies. The difference lies entirely in how the components are arranged. In a low-pass filter, the inductor is in series with the signal and the capacitor is in parallel to ground. In a high-pass filter, the capacitor is in series and the inductor is in parallel to ground. The math remains identical, but the physical behavior reverses: the high-pass configuration blocks DC and low frequencies while passing everything above the calculated corner frequency.






