A high pass and low pass filter are frequency-selective circuits that either block signals below a specific cutoff threshold (high pass) or block signals above it (low pass), shaping the AC waveform's amplitude and phase. In a real circuit, these filters do not just erase unwanted frequencies; they introduce a frequency-dependent phase shift and attenuate the signal at a specific roll-off rate (typically -20dB per decade for a first-order passive design), fundamentally altering how downstream components like ADCs, microcontrollers, or audio amplifiers interpret the waveform.

The most common mistake hobbyists make is treating the cutoff frequency ($f_c$) as a "brick wall." At $f_c$, the signal is not eliminated—it is attenuated by exactly -3dB. This means the voltage drops to 70.7% of its original value, and the power delivered to the load is halved. If you need a signal completely gone, you must design for a cutoff frequency far beyond the target interference, or use a higher-order active filter.

The Math: Cutoff Frequency and a Worked Numeric Example

For a standard first-order passive RC (Resistor-Capacitor) filter, the -3dB cutoff frequency is determined by a single, unbreakable formula:

$f_c = \frac{1}{2 \pi R C}$
Where $f_c$ is in Hertz, $R$ is in Ohms, and $C$ is in Farads.

Let us run a worked numeric example for a common bench scenario: smoothing a noisy 5V PWM signal from an ESP32 (GPIO 25) to create a pseudo-analog DC voltage for a 0-5V panel meter. The ESP32 is outputting a 5 kHz PWM wave. To smooth this into a flat DC line, we need a low pass filter with a cutoff frequency well below 5 kHz. Let us target $f_c = 50$ Hz to ensure the ripple is virtually eliminated.

We will select a standard 10 kΩ metal film resistor for $R$ to keep the impedance high enough that it does not overload the ESP32 GPIO pin (which can safely source/sink up to 40mA, but 10kΩ limits current to a safe 0.5mA). Now, we solve for $C$:

  • $C = \frac{1}{2 \pi \times 50 \times 10,000}$
  • $C = \frac{1}{3,141,592}$
  • $C \approx 3.18 \times 10^{-7}$ Farads, or 318 nF.

The closest standard E12 capacitor value is 330 nF. Plugging 330 nF back into the formula yields an actual cutoff of 48.2 Hz, which is perfectly acceptable for this application. However, the physical type of 330 nF capacitor you choose will dictate whether this circuit actually works on the bench—a trap we will cover in the component selection section below.

Where You Meet This in Practice

Filters are not just theoretical textbook exercises; they are mandatory for signal integrity in almost every mixed-signal and power design. Here is where you will physically wire them on the jobsite or workbench:

  • Sensor Signal Conditioning (Low Pass): Thermocouples and load cells pick up high-frequency RF noise from nearby switching power supplies. A low pass filter acts as an anti-aliasing stage before the signal hits a 12-bit or 16-bit ADC, preventing high-frequency noise from folding back into your DC measurement.
  • Audio Crossovers (High Pass & Low Pass): In a 2-way speaker cabinet, a high pass filter routes high frequencies to the tweeter (preventing low-frequency bass from physically destroying the delicate tweeter dome), while a low pass filter routes bass to the woofer.
  • EMI Sniffing and Debugging (High Pass): When probing a digital rail with an oscilloscope, you might insert an AC coupling capacitor (which forms a high pass filter with the scope's 1MΩ input impedance) to block the 3.3V DC offset, allowing you to zoom in on the 50mV high-frequency switching ripple.

Decision Tree: Which Filter Topology Do You Need?

Do not default to a simple passive RC filter for every problem. Use this decision matrix to select the exact topology and component grades for your specific application.

Application Scenario Required Topology Concrete Component Pick
Smoothing MCU PWM to DC (Low frequency, low precision) 1st Order Passive RC 10kΩ Metal Film Resistor + 330nF C0G/NP0 Ceramic Capacitor
Active Audio Crossover or DAC Reconstruction 2nd Order Active Sallen-Key Low Pass OPA2134 Op-Amp + 100nF Polypropylene Film Capacitors
Removing 60Hz Mains Hum from Audio/Mic Lines Active Twin-T Notch / High Pass TL072 JFET Op-Amp + 1% Tolerance Resistors
High-Frequency EMI suppression on I2C/SPI lines Passive Ferrite + Capacitor (Pi Filter) Murata BLM18PG121SN1D Ferrite Bead + 10nF X7R Ceramic
Bench Tip: If your application involves audio or precision DC measurements, never use standard X7R or Y5V ceramic capacitors in your filter network. Always specify C0G (also known as NP0) dielectrics or film capacitors.

The Hidden Trap: Capacitor Dielectrics and Loading Effects

The biggest point of failure in DIY filter design is ignoring the physical construction of the capacitor. If you built the 50 Hz ESP32 PWM filter from our earlier example using a cheap X7R ceramic capacitor, your circuit will likely fail to smooth the signal properly.

X7R and Y5V ceramics exhibit a severe voltage coefficient. A 330 nF X7R capacitor rated for 50V might actually measure only 150 nF when 5V DC is applied across it. This capacitance drop shifts your cutoff frequency from 48 Hz up to 106 Hz, allowing more PWM ripple to pass through to your meter. Furthermore, X7R ceramics are piezoelectric; they act as microphones, picking up physical vibrations from your workbench or cooling fans and injecting them into your signal path as microphonic noise.

The Fix: For precision analog filters, spend the extra $0.15 per part and buy C0G/NP0 ceramic capacitors (like the Kemet C315C331J1G5TA) or polypropylene film capacitors. C0G dielectrics have a near-zero voltage coefficient and a temperature stability of ±30 ppm/°C, meaning your 330 nF capacitor stays exactly 330 nF regardless of the DC bias or ambient temperature.

Warning on Cascading Passive Filters: A common instinct when you need a sharper roll-off (e.g., -40dB/decade instead of -20dB/decade) is to simply wire two passive RC low pass filters in series. Do not do this. The input impedance of the second stage will load the first stage, altering the cutoff frequency and flattening the Q-factor of the filter. If you need a second-order (or higher) response, you must either use an active topology (like a Sallen-Key op-amp circuit) or insert a unity-gain op-amp buffer between the passive RC stages.

Frequently Asked Questions

Can I use an inductor instead of a resistor for a low pass filter?

Yes, an RL (Resistor-Inductor) or LC (Inductor-Capacitor) filter works on the exact same mathematical principles, but the physical realities of the workbench make them less common for low-frequency signal processing. Inductors are bulky, expensive, and prone to picking up stray magnetic fields from nearby transformers or motors. For signals under 100 kHz, active RC filters using op-amps are vastly superior in size, cost, and noise immunity. Reserve LC filters for high-frequency RF work or high-current power supply output filtering.

What happens to the phase shift at the cutoff frequency?

In a first-order RC filter, the phase shift is exactly 45 degrees at the cutoff frequency ($f_c$). As you move deeper into the stopband, the phase shift approaches 90 degrees. If you are feeding this signal into a feedback loop (like a PID controller or a switching power supply error amplifier), this phase shift can eat into your phase margin and cause the system to oscillate. Always check the Bode plot of your filter if it resides inside a control loop.

When designing your next circuit, stop treating filters as an afterthought. Calculate the exact -3dB point, select the correct dielectric material for your capacitors, and buffer your passive stages with an op-amp if you need a steep roll-off. By following these physical and mathematical rules, your signals will be clean, your ADC readings will be stable, and your audio outputs will be free of hum.