An op-amp voltage follower is a unity-gain buffer circuit where the output is wired directly to the inverting input, forcing the output voltage to exactly match the non-inverting input voltage while providing massive impedance transformation. It doesn't change your signal's voltage amplitude; it changes your signal's stiffness (current-delivering capability), isolating fragile high-impedance sources from heavy low-impedance loads. Beginners commonly confuse it with a simple piece of wire or a BJT emitter follower, but unlike a wire, it prevents load-induced voltage sag, and unlike an emitter follower, it has virtually zero DC voltage offset.
The Math and the Magic: A Worked Numeric Example
To understand what this circuit actually changes in a real installation, we have to look at loading effects. Suppose you need a precise 2.5V reference from a 5.0V rail using a voltage divider made of two 100kΩ resistors. Unloaded, the midpoint is exactly 2.5V.
Now, connect a 10kΩ load (like a microcontroller ADC or a multimeter) directly to that midpoint. The 10kΩ load is in parallel with the bottom 100kΩ resistor, creating an equivalent resistance of 9.09kΩ. Your new output voltage drops to 5.0V × (9.09kΩ / (100kΩ + 9.09kΩ)) = 0.416V. You lost over 80% of your signal to loading effects because the source impedance was too high for the load.
Insert an op-amp configured as a voltage follower between the divider and the load. The op-amp's non-inverting input draws less than 1 picoamp (input impedance >1012 Ω). The voltage divider remains perfectly unloaded at 2.5V. The op-amp's output, capable of sourcing up to 23mA, easily drives the 10kΩ load (which only demands 0.25mA) while holding the output rock-solid at 2.500V.
Where You Meet This in Practice
You will rarely see a voltage follower drawn as a standalone block on a high-level schematic, but it is hiding inside almost every mixed-signal measurement system. Here is where it does the heavy lifting:
- SAR ADC Driving: Successive Approximation Register ADCs (like those inside the ATmega328P or ESP32) use an internal sample-and-hold capacitor. If your source impedance is >10kΩ, the internal capacitor won't charge fully during the nanosecond sampling window, causing erratic, low-reading conversions. A follower provides the <100Ω source impedance the ADC demands.
- High-Impedance Sensor Buffering: Piezoelectric vibration sensors, glass pH probes, and thermocouples generate signals with source impedances in the megaohm range. A wire would act as an antenna and a low-impedance load would short the signal to ground. A follower with a FET-input stage (like the TL072 or LMC6062) bridges this gap without loading the sensor.
- R-2R DAC Buffers: Resistor-ladder digital-to-analog converters have an output impedance that varies depending on the digital code being output. A voltage follower locks the output impedance to a near-zero constant, ensuring your analog output doesn't sag when connected to an amplifier.
Decision Tree: Choosing the Right Op-Amp for Your Follower
Not all op-amps are stable at unity gain, and not all are suited for your specific supply rails. Use this decision matrix to select your part.
| Application Need | Key Parameter | Recommended Part | Approx. Cost (2026) |
|---|---|---|---|
| 3.3V / 5V Microcontroller Interfacing | Rail-to-Rail I/O, Unity-Gain Stable | Microchip MCP6002 | $0.85 (DIP-8) |
| Audio Signal Buffering (±12V to ±15V) | Low Noise (nV/√Hz), High Slew Rate | TI OPA1612 or TL072 | $3.50 / $0.60 |
| High-Speed Video or Fast DAC (>10MHz) | High Bandwidth, Fast Settling Time | Analog Devices AD8055 | $4.20 |
| Precision DC Measurement (pH, Strain Gauges) | Ultra-Low Input Bias Current (<1pA) | TI LMC6062 or ADA4522 | $2.10 / $5.50 |
Stability Pitfalls: Why Your Follower Might Oscillate
The most common real-world failure mode of a voltage follower is high-frequency oscillation caused by capacitive loading. When you use a follower to drive a long coaxial cable (e.g., RG-58 at ~100pF/meter) or a large electrolytic decoupling capacitor, that capacitance interacts with the op-amp's internal open-loop output resistance.
This interaction adds a secondary pole to the feedback loop, eroding the phase margin. According to Analog Devices application notes on op-amp stability, once the phase margin drops below 45 degrees, the follower will ring on step transients or break into continuous oscillation at several megahertz, turning your buffer into a space heater and noise generator.
The Fix: Add a 22Ω to 47Ω isolation resistor in series with the op-amp's output pin, before the capacitive load. If DC accuracy at the load is critical, route the feedback trace from the load side of the resistor; if high-frequency stability is more important than microvolt DC accuracy, route the feedback directly from the op-amp output pin.
Frequently Asked Questions
What exactly does a voltage follower change in a real circuit?
It changes the source impedance of your signal. It takes a signal that might only be able to supply microamps of current (high impedance) and recreates it with the ability to supply tens of milliamps (low impedance). The voltage stays the same; the current capacity increases.
Why can't I just use a piece of wire instead of an op-amp?
A wire has zero gain, but it also has zero isolation. If you connect a heavy load to a weak source via a wire, Ohm's law dictates that the voltage will sag across the source's internal resistance. An op-amp follower draws virtually zero current from the source, preventing that sag, while using its own power supply rails to drive the heavy load.
Is a voltage follower the same as an emitter follower?
No. An emitter follower (a common-collector BJT amplifier) also provides current gain and high input impedance, but it suffers from a base-emitter voltage drop (typically ~0.6V to 0.7V). An op-amp voltage follower uses negative feedback to eliminate this offset, achieving a gain of exactly 1.000 with microvolt-level error.
Do I need dual power supplies (±5V) to use a voltage follower?
Not if you choose a modern single-supply, rail-to-rail op-amp like the MCP6002 or TLV2372. Older op-amps like the LM741 require dual supplies and cannot swing their outputs closer than ~2V to the supply rails, making them useless for 3.3V or 5V single-supply buffering. Always check the 'Common-Mode Input Voltage Range' and 'Output Voltage Swing' in the datasheet.
