The Butler oscillator is a crystal-controlled oscillator that uses the crystal near its series resonance point. They are used where a simple low-cost circuit is needed which can oscillate at high frquencies (>50MHz^[1]) by using overtones of a crystal, and also giving low low phase noise.^[2]

It was described by Butler in 1946 as the earthed grid oscillator, a derivative of the Hartley oscillator.^[3] It is also known as the bridged-T oscillator or the grounded-base oscillator.^[4]

The classic Butler oscillator circuit is a two-stage circuit with two non-inverting stages, a grounded base stage and an emitter follower.^[5] The crystal is inserted in series in the overall feedback path.^[5]

The more common modern form of the circuit uses just the emitter follower stage.^[1][6] The circuit may be analysed by considering it as a equivalent AC circuit with three parts. The emitter follower forms an amplifier with no phase shift. The crystal and its loading capacitor then produce a phase lag network, followed by the LC network of the resonant tank circuit. This then produces a phase lead, which overall meets the Barkhausen criteria for self-oscillation.^[1]

The Butler circuit is a free-running or tuned oscillator. If the crystal is replaced temporarily with a low value resistor, the circuit will still oscillate at approximately the design frequency of the tank circuit. This allows the circuit to be set-up and adjusted initially without the crystal, and also encourages the selection of the correct crystal harmonic.^[6] To avoid the circuit oscillating at the strong resonance of the crystal's fundamental, a small inductor may be placed in parallel with the crystal.^[1][6]

Both the better-known Pierce and Colpitts oscillator circuits may be considered as derivatives of the Butler.^{[6][lower-roman 1]}