Colpitts Oscillator

Part of Vacuum Tubes

The Colpitts oscillator is a practical, stable LC oscillator using a capacitive voltage divider to feed back energy from output to input, producing sustained sine wave oscillation.

Why This Matters

A radio transmitter needs a stable oscillating signal at its carrier frequency. A superheterodyne receiver needs a local oscillator that tunes across a range of frequencies. Test equipment needs a signal source for alignment and calibration. The Colpitts oscillator, invented in 1918, is one of the most practical LC oscillator topologies for these applications — it works well over a wide frequency range, is easy to build with salvage components, and produces a relatively pure sine wave.

Understanding the Colpitts circuit lets you build working oscillators for HF radio (3-30 MHz), medium wave (500-1500 kHz), and low frequencies down to audio range with large inductors. The circuit is forgiving of component tolerances and tube variations, making it a good first oscillator design for builders working without precision instruments.

The Colpitts circuit appears in countless pieces of salvage radio equipment, often as the variable-frequency oscillator (VFO) that controls tuning. Recognizing it and understanding its operation allows you to repair, modify, or substitute components when originals are unavailable.

Oscillator Fundamentals

Any oscillator consists of an amplifier and a feedback network. The amplifier overcomes losses in the resonant circuit; the feedback network couples a fraction of the output back to the input with the correct phase to sustain oscillation. For sustained oscillation, two conditions must be met: the loop gain must equal exactly 1 (greater than 1 and the amplitude grows without limit; less than 1 and it decays to zero), and the total phase shift around the loop must be exactly 360 degrees (or equivalently, 0 degrees).

In a common cathode triode amplifier, the signal at the plate is inverted from the grid — 180 degrees phase shift. The feedback network must provide the remaining 180 degrees to complete the loop. An LC tank circuit at resonance can provide zero phase shift (for a parallel resonant circuit) or 180 degrees depending on the configuration and whether the signal is taken from across the capacitor or the inductor.

The Colpitts oscillator achieves the feedback phase requirement by splitting the tank circuit capacitance into two series capacitors. The voltage divider formed by these two capacitors provides the correct fraction of the output signal, with appropriate phase, back to the grid. This is the defining feature of the Colpitts circuit.

Circuit Description

The Colpitts oscillator consists of:

An amplifying tube in common cathode configuration (triode or pentode) An inductor (L) connected between the plate and one end of the capacitor chain Two series capacitors (C1 and C2) with their junction connected to the cathode (or ground) The top of C1 connects to the plate (through the inductor); the bottom of C2 connects to ground The grid connects to the junction between C1 and C2 through a small capacitor or directly

The resonant frequency is determined by L and the series combination of C1 and C2:

f = 1 / (2π × √(L × Cseries))

where Cseries = (C1 × C2) / (C1 + C2)

The feedback fraction is set by the ratio of C1 to C2. The voltage at the junction (fed to the grid) is C1/(C1+C2) of the full tank voltage. Making C1 smaller than C2 reduces the feedback fraction, which is sometimes necessary to prevent the oscillator from producing a clipped, distorted output.

For a practical design, start with equal values for C1 and C2, giving a feedback fraction of 0.5. If the oscillator clips (producing a square-wave-like output instead of a sine wave), increase C1 relative to C2 to reduce feedback. If the oscillator fails to start reliably, decrease C1 relative to C2 to increase feedback.

Component Selection and Frequency Range

For an HF oscillator covering 7-14 MHz (the amateur radio 40m and 20m bands, useful for community communication):

Inductor L: approximately 1.5 microhenry, wound as 15 turns of 1mm wire on a 25mm diameter coil form, close-wound or slightly spaced. C1: 47 pF C2: 47 pF Series capacitance: 23.5 pF Resonant frequency: 1/(2π × √(1.5µH × 23.5pF)) = approximately 24 MHz

This is too high — to reach 7 MHz, increase the inductance or capacitance. Trying L = 10 µH with C1 = C2 = 100 pF:

Cseries = 50 pF f = 1/(2π × √(10µH × 50pF)) = 7.1 MHz

For a VFO that tunes across a range, make the inductor variable (an adjustable slug-tuned coil) or make one capacitor a variable type. A variable capacitor of 10-100 pF in parallel with C1 tunes the frequency while maintaining the feedback ratio from C2.

Temperature stability is the key performance criterion for an oscillator in a real transmitter. The resonant frequency drifts as components warm up and change dimension. Use NPO or C0G type capacitors for C1 and C2 — these ceramic capacitor types have near-zero temperature coefficients. Avoid silver-mica capacitors (good stability) and disk ceramics (poor stability, large temperature coefficient).

Biasing the Oscillator Tube

Grid leak bias is often used in oscillator circuits. When the oscillation starts, the grid occasionally swings positive, allowing grid current to flow. This current charges the coupling capacitor connected to the grid, making the grid negative. The time constant of the grid capacitor and grid resistor determines the average bias. This self-biasing mechanism limits the oscillation amplitude automatically.

For a stable, low-distortion oscillator, the grid leak resistor value is critical. Too large a value causes the bias to build up excessively, cutting off the tube too strongly and producing a distorted, non-sinusoidal output. Too small a value allows the grid to swing too far positive, causing heavy grid current that loads the tank circuit and reduces Q (quality factor). A value of 22-100 kilohms is typical for most HF oscillator designs with small-signal triodes.

Cathode bias can also be used, particularly when consistent amplitude is more important than using the simplest circuit. Set the cathode resistor to bias the tube to approximately Class A operation. The oscillation builds up to the amplitude at which the tank circuit’s resistive losses equal the energy added by the tube each cycle. This produces lower distortion than grid-leak bias.

Buffer Amplifier

An oscillator must be isolated from the load it drives. Any change in load impedance — from tuning the transmitter, keying the circuit, or varying supply voltage — changes the frequency slightly if the load is coupled directly to the oscillator. In a transmitter, this frequency pulling causes the signal to drift as the transmitter is modulated, defeating any attempt at accurate frequency control.

Follow the oscillator with a buffer amplifier: a separate tube stage, often a pentode, that amplifies the oscillator signal and presents a high input impedance to the oscillator. The pentode’s high plate impedance and low grid loading minimize the energy drawn from the oscillator tank circuit. Connect the oscillator output to the buffer through a small coupling capacitor (10-22 pF) that provides weak coupling. Any variation in the load seen by the buffer stage does not affect the oscillator tank because the coupling is so loose.

A simple common cathode pentode buffer amplifier following the Colpitts oscillator completes a practical VFO suitable as the frequency-determining stage of a community radio transmitter.