A high pass filter (HPF) is an electronic circuit that allows signals with a frequency higher than a specific cutoff point to pass through while attenuating (blocking) lower frequencies. In a real circuit, it changes the frequency spectrum of a signal by stripping out low-frequency rumble, DC offset, or AC mains hum before that signal reaches sensitive amplifier stages, microcontroller ADC inputs, or audio transducers. If you are designing an audio crossover, conditioning a sensor signal, or just trying to eliminate a 60Hz ground loop hum, understanding how to manipulate this cutoff point is mandatory bench knowledge.

The One-Sentence Definition and Core Concept

At its core, a first-order passive high pass filter consists of a capacitor in series with the signal path and a resistor to ground. The physics governing this relies entirely on capacitive reactance ($X_c$). A capacitor resists changes in voltage; at DC (0 Hz), its reactance is theoretically infinite, acting as an open circuit. As the frequency of the AC signal increases, the reactance drops, allowing the signal to pass through to the load resistor.

The critical metric for any filter is the -3dB cutoff frequency ($f_c$). This is the exact frequency where the output signal power drops to half its maximum value, and the voltage amplitude drops to 70.7% ($1/\sqrt{2}$) of the input. Frequencies well above this point pass with minimal loss, while frequencies below it are progressively choked off.

Bench Rule of Thumb: Never assume a filter completely "blocks" a frequency below the cutoff. A first-order RC filter rolls off at -6dB per octave. If your cutoff is 100Hz, a 50Hz signal isn't gone; it is merely attenuated by 6dB (half the voltage). If you need hard blocking, you need an active second-order topology.

The Math: Calculating Cutoff Frequency with Real Components

To design a filter, you need to select standard resistor and capacitor values that yield your target $f_c$. The governing equation for a simple RC high pass filter is:

$f_c = \frac{1}{2 \pi R C}$

Let us walk through a real-world numeric example. Suppose you are building an AC-coupling stage for an ESP32 ADC to read a 1kHz audio signal from an electret microphone, but you need to block the DC bias voltage and any sub-10Hz mechanical drift.

  1. Define Target: We want a cutoff frequency of roughly 15 Hz to safely pass the 20Hz-20kHz audio band while blocking DC.
  2. Select Resistor: The ESP32 ADC input impedance is roughly 1MΩ, but we want a stable, low-impedance source. We choose a standard 10kΩ resistor ($R$) to ground.
  3. Calculate Capacitor: Rearranging the formula to solve for $C$: $C = \frac{1}{2 \pi R f_c}$.
  4. Crunch Numbers: $C = \frac{1}{2 \times 3.14159 \times 10000 \times 15} = \frac{1}{942477} \approx 1.06 \mu F$.
  5. Select Component: The closest standard E12 capacitor value is 1µF.

Plugging 1µF and 10kΩ back into the original formula gives an actual cutoff of 15.9 Hz. This is perfect for audio AC coupling.

Target ApplicationResistor (R)Capacitor (C)Resulting Cutoff ($f_c$)
MCU ADC DC Blocking10 kΩ1 µF15.9 Hz
Audio Tweeter Crossover8 Ω (Speaker)2.2 µF9.0 kHz
Subwoofer Rumble Filter100 kΩ10 nF159.1 Hz
RF IF Stage Coupling50 Ω100 pF31.8 MHz

Where You Meet This in Practice

You will encounter high pass filters in almost every mixed-signal and audio design. Here is where they do the heavy lifting on the bench:

  • Oscilloscope AC Coupling: When you press the "AC" button on your scope's channel menu, you are physically switching in a high pass filter (usually with a cutoff around 10 Hz) to block the DC offset so you can zoom in on small AC ripple on a 12V power rail.
  • Audio Speaker Crossovers: Tweeters will physically tear themselves apart if fed low-frequency bass energy. A series capacitor acts as a first-order high pass filter to protect the voice coil.
  • Telecommunications: In DSL internet over copper phone lines, high pass filters at the modem separate the high-frequency data (above 25 kHz) from the low-frequency POTS voice audio (below 4 kHz).
  • Sensor Signal Conditioning: Piezoelectric vibration sensors generate massive low-frequency thermal drift. An HPF strips this drift so the ADC can focus on the high-frequency machine vibration harmonics.

Bench Scenario: Debugging a Noisy Subwoofer Crossover

Theory is clean; the workbench is not. Here is a real-world scenario demonstrating what happens when component parasitics and filter order are ignored.

The Setup: A DIY audio builder was designing an active crossover to feed a subwoofer and main bookshelf speakers. The goal was to use an op-amp (TL072) to buffer the signal, followed by a passive RC high pass filter on the main speakers to prevent them from distorting on deep bass. The target cutoff was 80 Hz.

The Numbers: Using the standard formula, the builder selected $R = 20\text{k}\Omega$ and calculated $C = 100\text{nF}$, which yields a theoretical $f_c$ of 79.5 Hz. They populated the breadboard with cheap, leaded Y5V dielectric ceramic capacitors to save money.

The Outcome: Upon testing, the main speakers still sounded muddy, the bass lacked punch, and a distinct 60Hz AC mains hum was clearly audible through the mains speakers, defeating the purpose of the filter.

What Went Wrong: Two distinct engineering failures occurred here. First, a single RC stage is a first-order filter, rolling off at only -6dB per octave. The 60Hz hum was only one-third of an octave below the 80Hz cutoff, meaning it was attenuated by a mere ~4dB—barely noticeable to the human ear. Second, and more insidious, was the capacitor dielectric physics. Y5V ceramic capacitors exhibit a massive voltage coefficient; at just 2V RMS of audio signal, a 100nF Y5V cap can lose up to 50% of its actual capacitance. This shifted the real-world cutoff frequency up to roughly 160 Hz, severely thinning out the mid-bass response of the main speakers.

The Fix: The builder replaced the passive RC stage with an active Sallen-Key second-order Butterworth topology using a dual op-amp. This provided a -12dB/octave roll-off, crushing the 60Hz hum. Crucially, they swapped the Y5V ceramics for C0G/NP0 dielectric capacitors (or WIMA film caps), which maintain stable capacitance regardless of voltage bias, locking the cutoff exactly at 80 Hz.

For active filter design topologies like the Sallen-Key, the Analog Devices Sallen-Key tutorial provides the exact transfer functions and component ratio tables needed to dial in the Q-factor without causing peaking at the cutoff frequency.

Common Confusions: High Pass vs. Low Pass vs. AC Coupling

When reading schematics or datasheets, terminology can blur. Here is what people commonly confuse with a high pass filter:

  • High Pass vs. Low Pass: They are exact inverses. A low pass filter (LPF) puts the resistor in series and the capacitor to ground, passing DC and low frequencies while shunting high frequencies to ground. If you swap the R and C positions in an RC network, you flip the filter type.
  • High Pass vs. AC Coupling: Electrically, they are often the exact same RC circuit. The difference is intent. "AC Coupling" implies the cutoff frequency is set extremely low (e.g., 1 Hz) purely to block a DC bias voltage while passing the entire AC signal of interest intact. "High Pass Filter" implies the cutoff is set intentionally within the signal band to shape the frequency response (like an audio crossover).
  • High Pass vs. Band Pass: A band pass filter is simply a high pass filter and a low pass filter cascaded in series. It blocks both the extreme lows and the extreme highs, allowing only a specific "band" of frequencies through.

Frequently Asked Questions

Does a high pass filter affect the phase of the signal?
Yes. In a first-order RC high pass filter, the phase shift is +45 degrees exactly at the cutoff frequency. As the frequency drops further below the cutoff, the phase shift approaches +90 degrees. In audio applications, this phase shift near the crossover point can cause destructive interference if not matched properly with the low pass filter feeding the subwoofer.

Can I build a high pass filter using an inductor instead of a capacitor?
Yes, this is called an RL (Resistor-Inductor) high pass filter. The inductor is placed in series (blocking high frequencies via inductive reactance) and the resistor to ground. However, RL filters are rarely used in low-frequency audio or signal conditioning because inductors large enough to handle sub-1kHz frequencies are physically massive, expensive, and prone to picking up electromagnetic interference (EMI). They are mostly reserved for high-frequency RF applications.

Why does my ESP32 ADC reading drift even with a high pass filter?
If your cutoff frequency is too low (e.g., 0.1 Hz), the filter will not attenuate low-frequency thermal drift or 50/60Hz mains hum. Furthermore, the Espressif hardware design guidelines note that the ESP32 ADC has inherent non-linearity and noise. Ensure your high pass filter is followed by a low-pass anti-aliasing filter to create a band-pass, and keep the source impedance below 10kΩ to allow the internal sample-and-hold capacitor to charge fully.