An RC phase shift oscillator is an audio-frequency sine wave generator that uses a cascade of resistor-capacitor (RC) networks to shift the signal phase by 180 degrees, which, combined with an inverting amplifier's 180-degree shift, satisfies the 360-degree Barkhausen criterion for sustained oscillation. In a real circuit, it changes a steady DC power supply into a clean, low-distortion AC sine wave without requiring the bulky, expensive, and electromagnetically noisy inductors found in LC tank circuits.
The Core Mechanism: Chasing 360 Degrees
To sustain oscillation, a circuit must feed a portion of its output back to its input in-phase (a total of 360°, or 0°) with enough amplitude to overcome losses. This is known as the Barkhausen criterion. In an RC phase shift oscillator, we use an inverting operational amplifier (op-amp) as the gain element. Because an inverting amp inherently flips the signal by 180°, the feedback network must provide the remaining 180° of phase shift at the target frequency.
A single RC low-pass filter can theoretically provide a maximum of 90° of phase shift, but reaching that requires the output amplitude to drop to zero. To get a usable signal and exactly 180° of shift, we cascade three identical RC sections. Each section provides 60° of phase shift at the oscillation frequency.
The trade-off for this phase shift is severe signal attenuation. A three-section RC network attenuates the signal by a factor of 1/29. Therefore, the inverting amplifier must have a voltage gain of at least 29 to compensate for the feedback network's losses and sustain the waveform.
The Math: Sizing R and C for a 1 kHz Tone
Let's design a fixed-frequency oscillator for a 1 kHz audio test tone. The formula for the oscillation frequency of a three-section RC network is:
f = 1 / (2 * π * R * C * √6)
Required Amp Gain: ≥ 29
Phase Shift per RC Section: 60°
First, select a practical capacitor value. For audio circuits, we want low distortion and tight tolerance, so we will use 10 nF (0.01 µF) WIMA MKS film capacitors rather than sloppy ceramic or electrolytic types. Now, solve for R:
R = 1 / (2 * π * 1000 * 10e-9 * 2.449)
R ≈ 6,497 Ω
The closest standard 1% resistor value is 6.49 kΩ, or you can use a common 6.8 kΩ 5% resistor (which will shift your actual frequency down to roughly 951 Hz).
Next, set the amplifier gain. Using a standard non-inverting or inverting op-amp configuration, we need a gain of 29. If we use an inverting configuration with an input resistor ($R_i$) of 10 kΩ, the feedback resistor ($R_f$) must be 290 kΩ. Since 290 kΩ isn't a standard E24 value, use a 270 kΩ fixed resistor in series with a 50 kΩ trimpot to dial in the exact gain.
Where You Meet This in Practice
You will typically find RC phase shift oscillators in fixed-frequency audio test gear, low-end function generators, electronic organs, and DTMF tone generation circuits. They are prized for their simplicity, low part count, and lack of inductors, making them easy to integrate into silicon ICs.
What people commonly confuse it with:
- The Wien Bridge Oscillator: The Wien bridge uses a lead-lag RC network (two RC pairs in a bridge configuration). It is vastly superior for variable frequency applications because you only need to gang two potentiometers together to tune it, whereas tuning an RC phase shift oscillator requires ganging three pots simultaneously to maintain the phase relationship.
- Relaxation Oscillators (e.g., 555 Astable): Beginners often confuse sine wave oscillators with relaxation oscillators. A 555 timer astable circuit generates square or triangle waves by charging and discharging a capacitor between threshold limits, not by relying on linear phase shift and Barkhausen feedback.
Bench Walkthrough: Building a 1 kHz Test Generator
Theory is clean; the breadboard is not. Here is a real-world scenario of building this circuit using a TL072 dual op-amp on a standard solderless breadboard.
- Setup: Wire the TL072 with ±12V DC rails. Cascade three 10 nF film caps and three 6.8 kΩ resistors from the output to the inverting input. Place a 10 kΩ resistor from the inverting input to ground, and a 270 kΩ resistor + 50 kΩ trimpot from output to inverting input.
- Numbers: Set the trimpot to its midpoint (yielding roughly 295 kΩ total feedback resistance, gain ≈ 29.5). Power on the circuit and connect an oscilloscope to the output pin.
- Outcome: The scope shows a flatline at 0V. No oscillation.
- What went wrong: The theoretical gain of 29 is the absolute minimum to sustain oscillation. To start oscillation from ambient thermal noise, the loop gain must be strictly greater than 1 (meaning amp gain > 29). Furthermore, breadboard parasitic capacitance (typically 2-5 pF per node) and 5% resistor tolerances shifted the phase slightly, requiring more gain to compensate.
- The Fix: Adjust the trimpot to max out the feedback resistance to 320 kΩ (gain = 32). The scope suddenly shows a waveform, but it's a trapezoid, not a sine wave. The peaks are slamming hard into the ±12V rails.
- Final Stabilization: Because the gain is > 29, the amplitude grows until the op-amp clips, introducing massive harmonic distortion. To fix this, solder two back-to-back 5.1V Zener diodes in parallel with the feedback resistor. As the output amplitude approaches 5.1V, the Zeners begin to conduct, effectively lowering the feedback resistance, dropping the gain to exactly 29, and softly clamping the amplitude into a clean, low-distortion sine wave.
Troubleshooting and FAQ
Why is my output a square wave instead of a sine wave?
Your amplifier gain is set too high, causing the op-amp to saturate against the power supply rails. Reduce the feedback resistor value, or implement an automatic gain control (AGC) mechanism like back-to-back Zener diodes or a small incandescent lamp (e.g., #327 bulb) in the input resistor path to thermally limit the gain as amplitude rises.
Why won't the circuit start oscillating at all?
Check three things: First, ensure your gain is strictly greater than 29. Second, verify your op-amp has sufficient bandwidth and slew rate for your target frequency. Third, check your capacitor tolerances; if you used 20% tolerance electrolytic or Y5V ceramic capacitors, the phase shift across the three sections will be mismatched, preventing the exact 180° shift required. Always use C0G/NP0 ceramics or film capacitors.
Can I use this to generate a 100 kHz sine wave?
It is not recommended. At higher frequencies, stray breadboard and PCB parasitic capacitances become significant relative to your designed 'C' values, destroying the phase relationship. For frequencies above 20 kHz, LC oscillators (like the Colpitts or Hartley) or crystal oscillators are far more stable and practical.
For deeper mathematical derivations of the transfer function and Barkhausen stability criteria, refer to the comprehensive guides on Electronics Tutorials and the practical circuit analysis provided by All About Circuits. Understanding the exact interaction between your passive component tolerances and active gain limits is the difference between a textbook diagram and a reliable, low-distortion bench instrument.






