Third-Order Active Butterworth Filters Compared in Low-Pass and High-Pass Form

September 8, 2026

Third-Order Active Butterworth Filters Compared in Low-Pass and High-Pass Form

Image: ChatGPT

EDN presents a side-by-side look at third-order active Butterworth filters in both low-pass and high-pass form, comparing idealized transfer functions with practical circuit realizations. The article moves from a classic equal-value R/C implementation to Sallen-Key Type I and Type IIion implant (also I2)Process or concept related to introducing controlled dopant concentrations into semiconductors. variants, an emitter-follower version, and a DCdirect currentElectric current that flows in one direction.-accurate topology.

The core polynomial, S^3 + 2S^2 + 2S + 1, is factored into a first-order and a second-order section, which explains why the original Type I approach can be built with two or three unity-gain op-amps depending on section order. Swapping resistor and capacitor positions converts the network between low-pass and high-pass responses while keeping the 3 dB corner at 1/(2πRC).

A notable variant is the Sallen-Key Type IIion implant (also I2)Process or concept related to introducing controlled dopant concentrations into semiconductors. form, which implements the factored Butterworth polynomial directly with a single op-amp and equal-value components, but with a passband gain of 2. The emitter-follower implementation replaces the unity-gain op-amps in the Type I structure and can use complementary transistors; that helps reduce DCdirect currentElectric current that flows in one direction. offset because the VBE drops cancel, although the slight gain shortfall shifts the magnitude response away from the ideal Butterworth curve.

The article also revisits a DCdirect currentElectric current that flows in one direction.-accurate low-pass filter originally used in early-1970s Fluke DMMdigital multimeters as a voltage-reference noise filter. In this topology, shunt capacitors work against an input series resistance, so the op-amp is isolated from direct resistive coupling to input and output and does not add offset or bias-current error in the same way as the other versions.

Simulation and bench measurements with LTspice, a DSO, and an AWG show the practical limits of finite op-amp output impedance, especially in stop-band rejection. The lab builds used 1% resistors and 10% film capacitors on plug-in protoboards. When the circuits are converted to high-pass form by swapping R and C, the DCdirect currentElectric current that flows in one direction.-accurate version is the only one that changes behavior in a noticeable way, because its shunt capacitors become shunt resistors and couple offset through to the output.

Source: EDN

More background: Analog Design

← All posts