A high pass filter (HPF) is an electronic circuit that allows signals with frequencies above a specific cutoff point to pass through while attenuating lower frequencies.
In a real circuit or installation, inserting an HPF changes the signal path by stripping away low-frequency noise, DC offsets, or mechanical rumble, while simultaneously introducing a frequency-dependent phase shift that causes the output waveform to lead the input. People commonly confuse the HPF with its inverse (the low pass filter), or mistakenly believe the cutoff frequency acts as an absolute 'brick wall' that instantly drops all lower frequencies to zero, rather than a gradual attenuation slope.
The Core Mechanics and the 'Brick Wall' Myth
The simplest high pass filter is a first-order passive RC network: a capacitor in series with the signal path, followed by a resistor to ground. The capacitor's reactance ($X_c$) is inversely proportional to frequency. At DC (0 Hz), reactance is infinite, blocking the signal entirely. As frequency rises, reactance drops, allowing more voltage to reach the load.
The most common misconception among hobbyists is treating the -3dB cutoff frequency ($f_c$) as a hard boundary. It is not. A first-order RC filter rolls off at -20 dB/decade (or -6 dB/octave). At exactly $f_c$, the signal power is halved, and the voltage amplitude is 70.7% of the passband. One decade below $f_c$, the signal is attenuated to roughly 10% of its original amplitude. If you need a sharper drop-off, you must cascade stages or use an active topology.
Worked Numeric Example: Designing a 1kHz Audio HPF
Let's design a passive high pass filter for an audio preamplifier to block sub-bass rumble and DC offset, targeting a cutoff frequency of exactly 1,000 Hz.
The governing formula for the -3dB cutoff frequency is:
f_c = 1 / (2 * π * R * C)
Step 1: Select the Capacitor
In audio signal paths, capacitor dielectric choice is critical. We will avoid high-K ceramics (like X7R) due to piezoelectric microphonics and voltage coefficient non-linearities. We select a 10 nF (0.01 µF) C0G/NP0 ceramic capacitor for its extreme linearity and temperature stability.
Step 2: Calculate the Resistor
Rearranging the formula to solve for R:
R = 1 / (2 * π * f_c * C)
R = 1 / (2 * 3.14159 * 1000 * 10e-9)
R = 15,915.5 Ω
Step 3: Pick a Standard Value and Verify
The closest standard 1% E96 series resistor is 15.8 kΩ. Let's recalculate the actual cutoff with this real-world part:
f_c = 1 / (2 * π * 15800 * 10e-9) = 1,007.3 Hz
- C1: 10nF 50V C0G/NP0 Ceramic (e.g., Kemet C315C103J1G5TA)
- R1: 15.8kΩ 1/4W 1% Metal Film (e.g., Vishay MRS25000C1582FCT00)
- Phase Shift at Cutoff: +45° (Output leads input)
Where You Meet a Hi Pass Filter in Practice
You will encounter HPFs across almost every domain of electrical engineering and DIY electronics:
- Oscilloscope AC Coupling: When you press the 'AC' button on a scope channel, you are inserting a hardware high pass filter (typically with a 10 Hz cutoff) in series with the BNC input. This blocks the DC bias, allowing you to increase the volts/division sensitivity to inspect tiny AC ripple on a 12V DC rail.
- Audio Loudspeaker Crossovers: Tweeters are physically incapable of reproducing low frequencies and will destroy themselves if fed high-power bass. A second-order (or higher) passive HPF is placed in series with the tweeter to safely route only high frequencies to the driver.
- Biomedical Sensor Conditioning: ECG and EEG amplifiers deal with microvolt-level signals riding on top of massive DC skin-electrode offsets. Active high pass filters (often set to 0.05 Hz) are mandatory to strip the DC drift before the signal hits the high-gain amplifier stages, preventing the op-amps from saturating against the supply rails.
Decision Tree: Choosing Topology and Components
Selecting the right filter topology depends entirely on your impedance constraints, noise tolerance, and frequency targets. Use this decision matrix to lock in your design approach.
| Application Scenario | Topology Choice | Concrete Default Pick |
|---|---|---|
| High-Fidelity Audio Signal Path | Passive RC or Active Sallen-Key | C0G Caps + TI TL072 Op-Amp |
| Sub-1Hz Precision Sensor (ECG/Strain) | Active Switched-Capacitor | Analog Devices LTC1068-200 |
| Power Supply Ripple / DC Blocking | Passive LC (Pi or T network) | Ferrite Bead + X7R Ceramic Caps |
| Microcontroller ADC Anti-Alias + DC Block | Combined Passive RC | 100nF X7R + 10kΩ 0.1% Film Resistor |
Real-World Edge Cases: Dielectrics and Parasitics
Theoretical math assumes perfect components. In practice, parasitics will ruin your filter response if you ignore them.
The Dielectric Trap: Never use X7R, Y5V, or Z5U ceramic capacitors in the signal path of a high-fidelity audio or precision measurement HPF. These 'Class II' dielectrics exhibit severe capacitance loss under applied DC bias (a 10µF X7R cap might act like 2µF at 10V) and generate piezoelectric noise when subjected to mechanical vibration. Always specify C0G (NP0) for signal-path capacitors under 1µF. For values above 1µF where C0G becomes physically massive and expensive, switch to film capacitors (polypropylene or polyester).
Op-Amp Gain-Bandwidth Product (GBWP): If you move from a passive RC filter to an active Sallen-Key high pass filter to achieve a steeper -40dB/decade roll-off without loading the signal source, you must check your op-amp's GBWP. If your HPF cutoff is 100 kHz and you require a passband gain of 10 (20 dB), your op-amp must have a minimum GBWP of 1 MHz. Using a generic LM358 (GBWP ~700 kHz) for this task will result in severe high-frequency roll-off and phase margin degradation. For active filters above 50 kHz, default to a high-speed op-amp like the ADA4891.
Frequently Asked Questions
Can I just use a capacitor without a resistor for a high pass filter?
No. A capacitor alone only forms a high pass filter if the load it is driving has a known, resistive input impedance. If you connect a series capacitor to a high-impedance op-amp buffer (e.g., 10 MΩ input), the resulting RC cutoff frequency will be fractions of a Hertz, effectively passing everything. You must provide a deliberate resistor to ground to set a predictable $f_c$.
Why does my active high pass filter oscillate or ring?
This is almost always caused by driving a capacitive load directly from the op-amp's output, or using an op-amp with insufficient phase margin for the chosen gain. Isolate the op-amp output from heavy capacitive loads (like long coaxial cables) by placing a small series resistor (e.g., 47 Ω) directly at the op-amp's output pin.
For further reading on active topologies, refer to the Analog Devices MT-202 Tutorial on Filters and the foundational All About Circuits chapter on reactive networks.
The Default Recommendation: If you are prototyping a general-purpose audio or sensor HPF and lack the time to run a full SPICE simulation, default to a first-order passive RC network using a 100nF C0G capacitor and a 10kΩ metal film resistor. This yields a reliable ~160Hz cutoff, requires no power supply, introduces zero active op-amp noise, and is virtually indestructible on the bench.
