The cutoff frequency of a passive first-order RC high pass filter is calculated using the formula fc = 1 / (2πRC). At this exact frequency, the output signal is attenuated by -3dB (roughly 70.7% of the input voltage amplitude), and the phase shift is exactly +45 degrees. Frequencies above fc pass through with minimal attenuation, while frequencies below fc are rolled off at a rate of -20dB per decade.
This guide breaks down the exact mathematical derivation, provides rearranged formulas for component selection, and walks through worked examples with strict unit tracking to prevent the most common bench mistakes.
The Core RC High Pass Filter Formula and Assumptions
The fundamental equation governing the -3dB cutoff frequency ($f_c$) of a first-order passive high pass filter is:
$$f_c = \frac{1}{2 \pi R C}$$
Every variable in this equation must be expressed in standard SI base units before calculating. The table below defines each symbol and its required unit.
| Symbol | Parameter | SI Base Unit | Common Bench Prefixes |
|---|---|---|---|
| fc | Cutoff Frequency (-3dB point) | Hertz (Hz) | kHz ($10^3$), MHz ($10^6$) |
| R | Resistance | Ohms (Ω) | kΩ ($10^3$), MΩ ($10^6$) |
| C | Capacitance | Farads (F) | μF ($10^{-6}$), nF ($10^{-9}$), pF ($10^{-12}$) |
| π | Archimedes' Constant | Dimensionless | ≈ 3.14159265 |
Critical Assumptions and Parasitic Limits
This formula assumes ideal components and an ideal circuit environment. In real-world applications, the formula applies perfectly only when the following conditions are met:
- Source Impedance ($Z_{source}$): Assumed to be 0Ω. If your signal source has a high output impedance (e.g., a passive guitar pickup at 10kΩ), that impedance adds directly to your filter resistor, lowering the actual cutoff frequency.
- Load Impedance ($Z_{load}$): Assumed to be infinite. If the stage following your filter has an input impedance close to the value of R, it forms a voltage divider, attenuating your passband gain and shifting the pole. As a rule of thumb, $Z_{load}$ should be at least $10 \times R$.
- Parasitics: The formula ignores the Equivalent Series Inductance (ESL) of the capacitor and the parasitic parallel capacitance of the resistor. At RF frequencies (above 10 MHz), a standard 0603 ceramic capacitor will self-resonate and begin acting like an inductor, destroying the high-pass behavior. For RF, you must use S-parameter models rather than this idealized lumped-element formula.
Rearranged Forms for Component Selection
On the bench, you rarely calculate the frequency from arbitrary parts. Usually, you have a target cutoff frequency and a limited inventory of standard E24/E96 resistors and capacitors. You need to solve for the missing component.
Solving for Resistance (R)
Use this when you have a specific capacitor on hand (or a capacitor value dictated by dielectric constraints) and need to find the required resistor:
$$R = \frac{1}{2 \pi f_c C}$$
Solving for Capacitance (C)
Use this when you have selected a standard resistor value (often chosen to set the input/output impedance of the stage) and need to find the capacitor:
$$C = \frac{1}{2 \pi f_c R}$$
Worked Examples with Strict Unit Tracking
The most common reason an online filter calculator gives you a nonsensical answer is a unit conversion failure. Below are two solved problems demonstrating explicit unit tracking.
Problem 1: Audio AC-Coupling Filter
Scenario: You are designing an AC-coupling high pass filter to block DC offset from a DAC output before it reaches an audio amplifier. You select a standard 10 kΩ resistor and a 100 nF ceramic capacitor. What is the exact -3dB cutoff frequency?
Step 1: Convert to SI Base Units
- $R = 10 \text{ k}\Omega = 10 \times 10^3 \, \Omega = 10,000 \, \Omega$
- $C = 100 \text{ nF} = 100 \times 10^{-9} \text{ F} = 0.0000001 \text{ F}$
Step 2: Substitute into the Formula
$$f_c = \frac{1}{2 \times \pi \times (10 \times 10^3) \times (100 \times 10^{-9})}$$
Step 3: Calculate the Denominator
- Multiply the powers of 10: $10^3 \times 10^{-9} = 10^{-6}$
- Multiply the coefficients: $2 \times 3.14159 \times 10 \times 100 = 6283.18$
- Denominator total: $6283.18 \times 10^{-6} = 0.00628318$
Step 4: Final Division
$$f_c = \frac{1}{0.00628318} \approx 159.15 \text{ Hz}$$
Result: The cutoff frequency is 159.15 Hz. This is suitable for passing most audio, though it will slightly attenuate the lowest bass notes (which extend down to 20 Hz).
Problem 2: Designing a Subsonic Rumble Filter
Scenario: You need a high pass filter for a turntable preamp to block subsonic rumble. Your target $f_c$ is exactly 20 Hz. To maintain a high input impedance and avoid loading the phono cartridge, you choose $R = 47 \text{ k}\Omega$. What capacitor value do you need?
Step 1: Convert to SI Base Units
- $f_c = 20 \text{ Hz}$ (Already base unit)
- $R = 47 \text{ k}\Omega = 47,000 \, \Omega$
Step 2: Use the Rearranged Formula for C
$$C = \frac{1}{2 \pi f_c R}$$ $$C = \frac{1}{2 \times \pi \times 20 \times 47,000}$$
Step 3: Calculate and Convert to Practical Units
- Denominator: $2 \times 3.14159 \times 20 \times 47,000 = 5,906,194$
- $C = 1 / 5,906,194 = 1.693 \times 10^{-7} \text{ F}$
- Convert to nanofarads: $1.693 \times 10^{-7} \times 10^9 = 169.3 \text{ nF}$
Result: You need a 169.3 nF capacitor. Since this is not a standard E12 value, you would select the closest standard value, such as 180 nF (which shifts $f_c$ slightly down to 18.7 Hz) or 150 nF (shifting $f_c$ up to 22.5 Hz).
Common Unit Mistakes and Realistic Magnitudes
If your calculator spits out a number that looks wrong, it almost certainly is. Use this magnitude sanity-check table to verify your results before ordering parts. If you are designing an audio filter and your math tells you to use a 45 Farad supercapacitor, you forgot to convert microfarads to Farads.
| Application Domain | Typical fc Range | Realistic R Values | Realistic C Values |
|---|---|---|---|
| Audio AC-Coupling | 10 Hz – 20 kHz | 1 kΩ – 100 kΩ | 1 nF – 10 μF |
| Sensor Signal Conditioning | 0.1 Hz – 100 Hz | 10 kΩ – 1 MΩ | 1 μF – 100 μF |
| RF / IF Stages | 1 MHz – 500 MHz | 50 Ω – 1 kΩ | 1 pF – 100 pF |
Decision Path: From Ideal Math to Standard E24 Parts
Math gives you an ideal, continuous number. The physical world forces you to buy discrete components from the E24 (5% tolerance) or E96 (1% tolerance) series. Furthermore, capacitor dielectrics introduce severe non-linearities that the basic formula ignores. Use the following decision tree to finalize your Bill of Materials (BOM).
| Condition / Constraint | Action / Decision | Concrete Example / Part Pick |
|---|---|---|
| IF calculated C is < 1 nF | Increase R to keep C in a manufacturable range. Stray PCB capacitance (typically 2-5 pF) will ruin filters with very small C values. | Change R from 1kΩ to 10kΩ; recalculate C. |
| IF calculated C is > 10 μF | Decrease R. Large ceramic capacitors exhibit severe DC bias derating and microphonics. Electrolytics have high ESR and poor tolerance. | Change R from 100kΩ to 10kΩ; recalculate C. |
| IF application is precision audio / low distortion | Select C0G / NP0 dielectric. Avoid X7R/Y5V which exhibit piezoelectric microphonics and voltage coefficients. | Kemet C0805C104J1GACTU (100nF C0G) |
| IF application is general DC-blocking / non-critical | Select X7R dielectric. It offers high capacitance density and is cost-effective for simple AC coupling. | Murata GRM188R71H104KA93D (100nF X7R) |
| IF calculated R is non-standard | Round to the nearest E24 value, then recalculate the exact $f_c$ using the real resistor value to verify it meets system specs. | Calculated 16.9kΩ → Pick 16.9kΩ (E96) or 16kΩ (E24). |
For automated verification of your final picked values against parasitic effects, cross-reference your manual math with the Analog Devices Filter Wizard or standard textbook references like Wikipedia's High-Pass Filter derivations. Always default to C0G/NP0 dielectrics for signal-path audio filtering, and reserve X7R strictly for power supply decoupling or non-critical DC blocking where board space is at a premium.
