sallen key filter
A Sallen-Key filter builds a second-order low-pass (or high-pass) response from just one op-amp, two resistors, and two capacitors, using positive feedback through the RC network itself to create a pair of complex-conjugate poles rather than needing an inductor or multiple active stages — the op-amp's own gain, set by a simple resistor divider, directly controls the resulting pole Q, which is what lets one circuit topology implement a gentle Butterworth rolloff or a peaky resonant response just by changing that one gain-setting ratio.
Positive feedback through the RC network itself, not through a separate feedback resistor, is what creates the complex pole pair. A passive RC low-pass filter made from two cascaded R-C sections only produces two real poles — a response that's always overdamped and can never peak; the Sallen-Key topology instead routes the op-amp's output back into the junction between the two RC sections, and that specific feedback path is what lets the characteristic equation produce a genuine complex-conjugate pole pair instead of two separate real ones, which is the mathematical trick that lets a single op-amp stage mimic the resonant behavior normally associated with an LC tank circuit.
The gain-setting resistor ratio is the single knob that tunes the filter's whole personality, from gentle to peaky. Because the op-amp's closed-loop gain K appears directly in the denominator of the pole-location equation, small changes in that one resistor ratio move the poles from deep in the overdamped region (gain near 1, Bessel-like gentle rolloff with no peaking) through critically damped (Butterworth, maximally flat passband) to underdamped (Chebyshev-like peaking near the corner frequency) — and as gain approaches 3, the poles move onto the imaginary axis entirely and the circuit stops being a filter and starts being a Wien-bridge-style oscillator, which is why Sallen-Key designs deliberately stay well clear of that gain value.
Component sensitivity gets worse, not better, as Q rises, which is the topology's well-known practical limitation. At low Q the Sallen-Key filter is forgiving of ordinary resistor and capacitor tolerances, but as the design pushes gain closer to the oscillation boundary to achieve a high Q, the pole locations become increasingly sensitive to small component value errors — a 1% resistor tolerance that's harmless at Q of 0.7 can produce a meaningfully wrong response at Q of 10, which is why high-Q analog filter designs often switch to multiple-feedback or state-variable topologies that distribute sensitivity more evenly, reserving Sallen-Key for the low-to-moderate Q ranges where its single-op-amp simplicity is a clear win.
Cascading identical-looking stages with different gain and component values is how real anti-aliasing and reconstruction filters reach higher order. A single Sallen-Key section only provides a second-order (12 dB per octave) rolloff, so practical filter designs needing steeper rolloff — a sharp anti-aliasing filter ahead of an ADC, or a reconstruction filter after a DAC — cascade two, three, or four Sallen-Key sections in series, each tuned to a different pole-pair location from a Butterworth, Chebyshev, or Bessel polynomial table, which is why filter design software and reference tables exist specifically to hand designers the per-stage gain and component ratios needed to assemble a target overall response from these individually simple building blocks.
Read a Sallen-Key filter through a *one-knob-many-responses* lens rather than a *fixed low-pass circuit* lens: the same two resistors and two capacitors can produce a gentle Bessel rolloff, a maximally flat Butterworth response, or a sharply peaking Chebyshev-like curve, with the entire difference living in a single gain-setting resistor ratio — right up to the exact value where that same ratio turns the filter into an oscillator instead.