A non-inverting amplifier is an operational amplifier circuit configuration where the input signal is applied to the non-inverting (+) terminal, producing an output voltage that is in phase with the input and scaled by a positive gain factor. In a real circuit or installation, this topology changes the game by boosting signal amplitude while preserving the original waveform phase and offering extremely high input impedance, which prevents sensitive signal sources from being loaded down. Because of this high impedance and phase preservation, it is frequently confused with the inverting amplifier (which flips the signal phase by 180 degrees and presents a low input impedance to the source) or the voltage follower (which is simply this exact same circuit configured with a fixed gain of exactly 1).
The Core Math: Gain Formula and Input Impedance
To design a non inverting amp, you only need to set two resistors. The closed-loop voltage gain ($A_v$) is determined by the feedback resistor ($R_f$) and the ground resistor ($R_{in}$) using this formula:
$A_v = 1 + \frac{R_f}{R_{in}}$
Notice the 1 + at the beginning. This is the defining mathematical quirk of the non-inverting topology. Because the input signal goes directly to the non-inverting pin, the output can never be less than the input. The minimum possible gain is 1 (when $R_f = 0\Omega$ and $R_{in} = \infty\Omega$), which turns the circuit into a unity-gain buffer.
The input impedance ($Z_{in}$) of this configuration is exceptionally high. While the op-amp itself has an open-loop input impedance (often in the megaohms or gigaohms), the negative feedback multiplies this effective impedance by the loop gain. For a modern CMOS op-amp, the practical input impedance is limited only by PCB leakage and the bias network, easily exceeding 10 G\Omega. This means it draws virtually zero current from your signal source.
Worked Numeric Example: Designing a 15x Gain Stage for an ESP32
Let us design a front-end amplifier for a high-impedance piezoelectric vibration sensor that outputs a 0.2V peak sine wave. We need to amplify this to roughly 3.0V to maximize the resolution of an ESP32’s 12-bit ADC (which operates on a 3.3V logic scale).
1. Calculate the Required Gain:
Target Output = 3.0V
Input Signal = 0.2V
$A_v = \frac{3.0V}{0.2V} = 15$
2. Select the Resistors:
Using $A_v = 1 + \frac{R_f}{R_{in}}$, we need $\frac{R_f}{R_{in}} = 14$.
Let us choose standard 1% metal film resistor values: $R_{in} = 10k\Omega$ and $R_f = 140k\Omega$.
Verification: $1 + (140,000 / 10,000) = 1 + 14 = 15$.
3. Choose the Right Op-Amp (The Real-World Gotcha):
A beginner might grab a legacy LM358 because it is cheap and ubiquitous. However, the LM358 has bipolar inputs that draw nanoamps of bias current, which will cause a massive DC offset voltage across the high-impedance piezo sensor. Furthermore, the LM358 output cannot swing rail-to-rail; on a 3.3V supply, its maximum output is only about 2.0V, meaning our 3.0V target would severely clip.
Instead, we select the Microchip MCP6002. It is a dual, rail-to-rail I/O CMOS op-amp that operates perfectly at 3.3V. Its input bias current is a mere 1 pA, preventing sensor loading, and its output can swing within 50mV of the 3.3V rail.
4. Verify the Bandwidth:
Every op-amp has a Gain-Bandwidth Product (GBWP). The MCP6002 has a typical GBWP of 1 MHz. Because closed-loop bandwidth is $\frac{GBWP}{A_v}$, our usable bandwidth is $\frac{1,000,000}{15} = 66.6 \text{ kHz}$. Since our piezo vibration signal peaks around 5 kHz, we have more than enough headroom to avoid phase shift and amplitude roll-off. For deeper reading on op-amp bandwidth limitations, the All About Circuits guide on op-amp configurations provides excellent foundational math.
Where You Meet This in Practice
You will find the non inverting amp topology anywhere a fragile, high-impedance signal needs to be buffered or boosted without phase inversion. Common bench and jobsite encounters include:
- Audio Pre-Amplifiers: Buffering electret microphone capsules or passive guitar pickups where phase coherence and high input impedance are critical to preserving high-frequency harmonics.
- Sensor Signal Conditioning: Amplifying millivolt-level outputs from strain gauges, thermocouples, or piezo elements before feeding them into a microcontroller ADC.
- DAC Output Buffers: Taking the weak, high-impedance output of an I2C digital-to-analog converter (like the MCP4725) and providing the low-impedance current drive needed to feed a power amplifier or motor controller.
To understand when to choose this topology over others, refer to the comparison matrix below. For more advanced architectural trade-offs, Analog Devices Tutorial MT-041 breaks down the internal transistor-level differences that dictate these external behaviors.
| Feature | Non-Inverting Amp | Inverting Amp | Voltage Follower |
|---|---|---|---|
| Gain Equation | $1 + (R_f / R_{in})$ | $- (R_f / R_{in})$ | $1$ (Fixed) |
| Minimum Gain | +1 (0 dB) | 0 (Negative Infinity dB) | +1 (0 dB) |
| Phase Shift | 0° (In-phase) | 180° (Inverted) | 0° (In-phase) |
| Input Impedance | Extremely High (G\Omega) | Low (Equal to $R_{in}$) | Extremely High (G\Omega) |
| Best Used For | High-Z sensor buffering | Summing mixers, DAC I/V | Isolation, impedance matching |
Frequently Asked Questions
Why does my non inverting amp output clip before reaching the supply rail?
This is almost always caused by the op-amp’s output voltage swing limitations. Standard op-amps like the LM741 or LM358 cannot drive their output pins all the way to the positive or negative supply rails; they typically lose 1.5V to 2V of headroom. If you are powering an LM358 with a single 5V supply, the maximum output is only about 3.0V to 3.5V. If your calculated gain demands a 4.5V output, the waveform will flat-line (clip) at ~3.5V. To fix this, either increase the supply voltage or switch to a modern “Rail-to-Rail Output” (RRO) op-amp like the MCP6002 or OPA333.
Can I use a non inverting amp to attenuate a signal (gain less than 1)?
No. The mathematical floor for a standard non-inverting configuration is a gain of exactly 1 (unity). You cannot set the resistors to achieve a gain of 0.5. If you need to attenuate a signal while maintaining high input impedance, the correct approach is to place a passive resistor voltage divider at the input to drop the voltage, and then feed that divided signal into a unity-gain voltage follower (buffer) to restore the low output impedance.
What is the difference between a non inverting amp and a voltage follower?
A voltage follower is simply a specific, edge-case implementation of the non inverting amp. In a voltage follower, the feedback resistor ($R_f$) is replaced with a direct wire (0 ohms), and the ground resistor ($R_{in}$) is removed entirely (infinite ohms). Plugging these into the gain formula yields $1 + (0 / \infty) = 1$. The voltage follower provides zero voltage gain but offers massive current gain and impedance transformation, acting as a perfect 1:1 isolation buffer.
How does input offset voltage affect my high-gain non inverting amp?
Input offset voltage ($V_{os}$) is a tiny internal mismatch inside the op-amp that makes it think there is a signal even when the inputs are shorted together. In a non-inverting configuration, this offset voltage is amplified by the exact same closed-loop gain as your actual signal. If your op-amp has a $V_{os}$ of 2mV and your circuit gain is 100, your output will have a permanent 200mV DC error added to it. For high-gain DC applications, you must select a precision “zero-drift” or chopper-stabilized op-amp like the TI OPA2188, which limits offset to microvolts.
