The Core High Pass Filter Formula and Symbol Definitions
The standard high pass filter formula calculates the cutoff frequency ($f_c$) of a first-order passive RC (resistor-capacitor) network. At this exact frequency, the output signal's power drops by half, and the voltage amplitude falls to 70.7% ($\frac{1}{\sqrt{2}}$) of the input voltage, representing the -3dB point. According to foundational circuit theory documented by Electronics Tutorials, the formula is derived from the point where the capacitive reactance ($X_c$) equals the resistance ($R$).
The primary equation is:
$f_c = \frac{1}{2 \pi R C}$
Symbol Definition Table
| Symbol | Unit | Base SI Unit | Description |
|---|---|---|---|
| $f_c$ | Hertz (Hz) | $s^{-1}$ | Cutoff frequency (-3dB point) |
| $\pi$ | Dimensionless | N/A | Archimedes' constant ($\approx 3.14159$) |
| $R$ | Ohms ($\Omega$) | $kg \cdot m^2 \cdot s^{-3} \cdot A^{-2}$ | Resistance of the series or shunt resistor |
| $C$ | Farads (F) | $s^4 \cdot A^2 \cdot m^{-2} \cdot kg^{-1}$ | Capacitance of the coupling or shunt capacitor |
Rearranged Forms
When designing a filter, you rarely solve for $f_c$ from scratch; usually, you have a target cutoff frequency and need to select standard component values. Here are the algebraic rearrangements:
- Solving for Resistance: $R = \frac{1}{2 \pi f_c C}$
- Solving for Capacitance: $C = \frac{1}{2 \pi f_c R}$
- Angular Frequency Form (Radians/Second): $\omega_c = \frac{1}{R C}$ (where $\omega_c = 2 \pi f_c$)
Assumptions, Unit Traps, and Realistic Magnitudes
Blindly plugging numbers into the high pass filter formula without understanding its boundaries leads to broken prototypes. Here is what you must account for on the bench.
When the Formula Applies (and Its Assumptions)
This formula assumes ideal components in a sinusoidal steady-state AC analysis. It assumes the capacitor has zero Equivalent Series Resistance (ESR) and zero parasitic inductance, and that the resistor is purely resistive with no parasitic parallel capacitance. Furthermore, it describes a first-order filter, meaning the signal rolls off at -20dB per decade (or -6dB per octave) past the cutoff frequency. If you cascade two identical RC stages without buffering them via an op-amp, the loading effect shifts the actual -3dB point, and this simple formula no longer accurately predicts the system response.
Unit Mistakes That Break the Math
The most common reason hobbyists and students calculate a wildly incorrect cutoff frequency is the 'Micro and Pico Trap'. Component values are rarely printed in base Farads. A '104' ceramic capacitor is 100nF ($100 \times 10^{-9}$ F), not 100 F. If you plug '100' into the formula instead of '0.0000001', your calculated frequency will be off by a factor of one billion. Similarly, mixing up Hertz ($f_c$) with radians per second ($\omega_c$) will throw your answer off by a factor of $2\pi$ (roughly 6.28). Always convert $k\Omega$ to $\Omega$ and $\mu F$ to $F$ before calculating.
What a Realistic Answer Magnitude Looks Like
Context dictates whether your answer makes sense. As noted in filter design primers by All About Circuits, audio AC-coupling high pass filters typically target a cutoff between 10 Hz and 40 Hz to block DC offset while passing bass frequencies. RF DC-blocking filters usually sit between 1 MHz and 100 MHz. If you are designing an audio input filter and your formula spits out $f_c = 0.0004$ Hz, you forgot to convert microfarads to farads. If it spits out $f_c = 4,500,000$ Hz, you likely dropped a zero on your resistor conversion.
Worked Examples with Strict Unit Tracking
Below are two bench-realistic problems. Notice how the unit tracking explicitly proves that $\frac{1}{\Omega \cdot F}$ simplifies to Hertz ($Hz$).
Problem 1: Finding the Cutoff Frequency of an Existing Network
Scenario: You have a passive RC high pass filter on a breadboard consisting of a $4.7 k\Omega$ series resistor and a $100 nF$ coupling capacitor. What is the -3dB cutoff frequency?
- Convert to Base SI Units:
- $R = 4.7 k\Omega = 4,700 \Omega$
- $C = 100 nF = 100 \times 10^{-9} F = 0.0000001 F$
- Substitute into the Formula: $f_c = \frac{1}{2 \times \pi \times 4700 \times (100 \times 10^{-9})}$
- Calculate the Denominator: $2 \times 3.14159 \times 0.00047 = 0.002953 \text{ seconds}$
- Invert to Find Frequency: $f_c = \frac{1}{0.002953} = 338.6 Hz$
- Unit Verification: $\frac{1}{\Omega \cdot F} = \frac{1}{(V/A) \cdot (A \cdot s / V)} = \frac{1}{s} = Hz$. The answer is 338.6 Hz.
Problem 2: Designing a Subwoofer Rumble Filter
Scenario: You need to build a high pass filter to block sub-sonic rumble (below 80 Hz) from reaching a midrange amplifier. You have a $1 \mu F$ film capacitor in your parts bin. What resistor value do you need?
- Select the Rearranged Formula: $R = \frac{1}{2 \pi f_c C}$
- Convert to Base SI Units:
- $f_c = 80 Hz$
- $C = 1 \mu F = 1 \times 10^{-6} F$
- Substitute and Calculate Denominator: $2 \times \pi \times 80 \times (1 \times 10^{-6}) = 0.00050265$
- Invert to Find Resistance: $R = \frac{1}{0.00050265} = 1989.4 \Omega$
- Practical Component Selection: $1989 \Omega$ is not a standard value. Using a standard E24 resistor kit or a tool like the DigiKey Resistor Calculator, you would select the closest standard 5% value: $2.0 k\Omega$. This shifts your actual cutoff slightly down to 79.5 Hz, which is perfectly acceptable for audio rumble filtering.
Frequently Asked Questions
How does the high pass filter formula change for active op-amp circuits?
The cutoff frequency formula for the RC network at the input of an active non-inverting high pass filter remains exactly the same: $f_c = \frac{1}{2 \pi R C}$. The operational amplifier (like a TL072 or NE5532) provides voltage gain ($A_v = 1 + \frac{R_f}{R_i}$) and lowers the output impedance to drive heavier loads, but it does not alter the pole location established by the passive RC components. However, you must ensure the op-amp's Gain-Bandwidth Product (GBWP) is significantly higher than your target cutoff frequency to prevent the op-amp's internal limitations from creating an unintended low-pass roll-off.
What is the high pass filter formula in radians per second?
When working in the Laplace domain, designing state-variable filters, or plotting Bode diagrams, engineers use angular frequency ($\omega_c$) instead of Hertz. The formula simplifies to $\omega_c = \frac{1}{R C}$. The result is expressed in radians per second (rad/s). To convert this back to standard Hertz for component purchasing, you simply divide by $2\pi$ ($f_c = \frac{\omega_c}{2\pi}$).
Does the high pass filter formula apply to RL (resistor-inductor) circuits?
No, the $f_c = \frac{1}{2 \pi R C}$ formula is strictly for RC networks. While RL high pass filters exist (configured with a series resistor and a shunt inductor to ground), inductors are bulky, expensive, and prone to electromagnetic interference (EMI) pickup, making them rare in low-frequency analog signal processing. If you must calculate the cutoff for a passive RL high pass filter, the formula is $f_c = \frac{R}{2 \pi L}$, where $L$ is inductance in Henrys.
Why is my measured -3dB point different from my calculated formula result?
Real-world components introduce parasitics that the ideal formula ignores. First, ceramic capacitors (especially X7R and Y5V dielectrics) exhibit severe capacitance derating under DC bias voltage; a $1 \mu F$ cap might act like $0.4 \mu F$ in-circuit, shifting your cutoff frequency higher. Second, source and load impedance matter. If your signal source has a $600 \Omega$ output impedance and your filter uses a $1 k\Omega$ resistor, the effective resistance is actually $1.6 k\Omega$, lowering the cutoff frequency. Always buffer high-precision filters with unity-gain op-amps to isolate the RC network from source and load impedance variations.
How do I calculate the high pass filter formula for a second-order (Sallen-Key) topology?
A second-order Sallen-Key high pass filter cascades two RC networks and uses an op-amp for buffering and Q-factor control. The generalized cutoff frequency formula becomes $f_c = \frac{1}{2 \pi \sqrt{R_1 R_2 C_1 C_2}}$. If you design it as a Butterworth response (maximally flat passband) and choose equal component values ($R_1 = R_2 = R$ and $C_1 = C_2 = C$), the formula simplifies back to the familiar $f_c = \frac{1}{2 \pi R C}$, but the roll-off steepness doubles to -40dB per decade.
