A low pass active filter is an electronic circuit that uses an operational amplifier alongside resistors and capacitors to pass low-frequency signals while attenuating high-frequency noise, simultaneously providing signal gain or buffering. While a passive filter simply bleeds off high frequencies and inevitably drops your signal voltage, an active filter uses external power to maintain or amplify the desired low-frequency signal. What it changes in a real circuit is the impedance relationship: it completely isolates your signal source from the load, preventing the amplitude droop and cutoff-frequency shifting that plagues passive designs. Hobbyists commonly confuse active filters with basic passive RC networks, assuming a resistor and capacitor are enough, or they confuse the non-inverting Sallen-Key topology with inverting multiple-feedback (MFB) designs.
The Core Mechanics of a Low Pass Active Filter
At its heart, an active filter relies on the high input impedance and low output impedance of an operational amplifier (op-amp). In a standard passive RC filter, the cutoff frequency is dictated by the resistor and capacitor values, but this assumes the load connected to the filter has infinite impedance. In reality, if you connect a passive filter to a microcontroller ADC or a downstream amplifier, the load impedance forms a parallel resistance with your filter resistor, shifting your cutoff frequency and attenuating your passband signal.
An active filter solves this by placing the op-amp either before the RC network (as a buffer) or integrating the RC network into the op-amp's feedback loop. The most popular configuration for hobbyists and bench engineers is the Sallen-Key topology. It uses two resistors, two capacitors, and a single op-amp to create a second-order (12 dB/octave) roll-off. Because the op-amp's output drives the load directly, the filter's mathematical behavior remains locked in place regardless of what you connect to the output pin.
Passive vs. Active: Why Add an Op-Amp?
Adding an op-amp requires a dual power supply (or at least a VCC and GND rail), which adds BOM cost and board space. Here is exactly when that trade-off is worth it, based on standard electronics design principles.
| Criteria | Passive RC Filter | Active Sallen-Key Filter |
|---|---|---|
| Signal Gain | Always < 1 (insertion loss) | Can be 1 (unity) or > 1 (amplified) |
| Load Isolation | Poor (load alters cutoff freq) | Excellent (op-amp buffers load) |
| Roll-off Rate (1 stage) | -20 dB/decade (1st order) | -40 dB/decade (2nd order) |
| Power Requirement | None (passive components only) | Requires VCC/GND for op-amp |
| Inductor Required? | No | No (simulates inductance via C) |
If you are simply rolling off high-frequency hiss on a line-level audio signal going into a high-impedance guitar amp input, a passive RC filter is fine. If you are conditioning a millivolt sensor signal for a microcontroller, you need an active filter.
Designing a 1 kHz Sallen-Key Filter: A Numeric Walkthrough
Let's design a unity-gain, second-order Butterworth low pass active filter with a target cutoff frequency (f_c) of roughly 1 kHz. A Butterworth response gives us a maximally flat passband with no peaking at the cutoff edge.
- Set the Resistors: For a unity-gain Sallen-Key Butterworth filter, the standard design rule is to set R1 = R2 = R, and C1 = 2 * C2. Let's choose R = 10 kΩ to keep impedance low enough to avoid thermal noise, but high enough to avoid overloading the op-amp output.
- Calculate Base Capacitance: Using the formula C = 1 / (2 * π * R * f_c), we get C = 1 / (2 * π * 10,000 * 1000) ≈ 15.9 nF.
- Assign C1 and C2: Since C1 needs to be roughly double C2, we can set C2 to the standard value of 10 nF, and C1 to the standard value of 22 nF (closest E12 series value to 20nF).
- Verify the Cutoff: Plugging these real-world values back into the master equation: f_c = 1 / (2 * π * √(R1 * R2 * C1 * C2)). This yields 1.07 kHz, which is perfectly acceptable for a 1 kHz target.
Wire the non-inverting input of the op-amp to a voltage divider for a virtual ground (if running on a single supply), tie the output directly to the inverting input for unity gain, and route the signal through the R1-R2-C1-C2 network into the non-inverting pin.
Where You Meet This in Practice (And a Bench War Story)
Theory is clean; the bench is messy. Here is a real-world scenario where a passive filter failed and an active filter saved the design.
The Setup: I was reading a 10Hz strain gauge signal using an ESP32's internal 12-bit ADC. The sensor output was conditioned by an instrumentation amp, but the board also housed a TPS5430 buck converter generating 50 kHz switching noise.
The Outcome: The ADC readings were jumping wildly. The 50kHz noise was aliasing into the sampling window, rendering the 12-bit resolution useless.
What Went Wrong (The Passive Attempt): My first fix was a simple passive RC low pass filter: a 10 kΩ resistor and a 1 µF capacitor, yielding a 15 Hz cutoff. The 50kHz noise vanished, but a new problem emerged. The ESP32's ADC input impedance is notoriously low and variable during the sampling phase (often dropping below 100 kΩ). The 10 kΩ series resistor formed a voltage divider with the ADC's internal sampling capacitor. This caused a massive 30% amplitude droop and severe non-linearity, as documented in Espressif's ADC peripheral guidelines.
The Fix (The Active Attempt): I ripped out the passive RC and built a unity-gain Sallen-Key filter using an MCP6002 rail-to-rail op-amp. I used 10 kΩ resistors and scaled the capacitors to set the cutoff at 50 Hz. Because the MCP6002's output impedance is less than 100 ohms, it easily drove the ESP32's ADC sampling capacitor without any voltage droop. The 50kHz noise was crushed by the 40 dB/decade roll-off, and the 10Hz signal arrived at the ADC perfectly intact.
Troubleshooting Active Filter Instability
Q: My op-amp output is oscillating at a high frequency, even with no input signal. Why?
A: You likely have a capacitive load on the op-amp's output (like a long coaxial cable or a large bulk capacitor). Capacitive loads introduce a pole in the feedback loop, destroying phase margin. Fix: Add a small series resistor (10 Ω to 100 Ω) directly at the op-amp output pin, before the capacitive load.
Q: The output is pinned to the positive or negative supply rail.
A: You have violated the op-amp's Input Common-Mode Voltage Range. If you are using a classic LM358 on a single 5V supply and trying to filter a signal centered at 0V (GND), the inputs cannot read down to the negative rail. Fix: Switch to a true Rail-to-Rail Input/Output (RRIO) op-amp like the MCP6002, or bias your AC signal to VCC/2 using a voltage divider.
Q: The filter is peaking (ringing) at the cutoff frequency instead of rolling off smoothly.
A: Your Q-factor is too high. In a Sallen-Key design, this happens if your capacitor ratio (C1/C2) is too small for the chosen gain. Fix: Verify your capacitor values. For a Butterworth response at unity gain, ensure C1 is roughly twice the value of C2. For deeper design math, reference the All About Circuits semiconductor textbook on active network synthesis.
