A negative feedback amplifier is a circuit that routes a portion of its output signal back to its input out of phase, deliberately reducing overall gain to trade raw amplification for stability, wider bandwidth, and lower distortion. If you have ever wired an operational amplifier (op-amp) and wondered why we bother with external resistors when the chip itself boasts a built-in gain of 100,000 or more, this concept is the answer. Without feedback, an op-amp acts as a comparator that slams its output to the positive or negative supply rail at the slightest microvolt difference. Negative feedback tames this wild, temperature-dependent open-loop behavior into a predictable, precise closed-loop system.
People commonly confuse negative feedback with positive feedback. Positive feedback routes the output back in phase, reinforcing the input and driving the circuit into saturation or oscillation. That is highly useful for Schmitt triggers or relaxation oscillators, but fatal for linear amplification. Negative feedback, by contrast, subtracts from the input error, forcing the output to settle at the exact voltage needed to make the differential input virtually zero.
The Core Mechanism: Trading Gain for Control
What negative feedback actually changes in a real circuit is the relationship between the amplifier's internal limitations and its external performance. Every physical op-amp suffers from limited bandwidth, thermal drift, and non-linear distortion at high amplitudes. By feeding a fraction of the output back to the inverting (-) input, the circuit continuously compares the actual output to the desired output and corrects its own errors in real-time.
This self-correction mechanism yields three massive benefits:
- Desensitized Gain: The closed-loop gain depends almost entirely on the external passive components (resistors), not the op-amp's internal silicon.
- Extended Bandwidth: The gain-bandwidth product is constant. By forcing the gain down, the usable frequency range pushes higher.
- Lower Impedance: Negative feedback reduces output impedance (making it better at driving loads) and increases input impedance (preventing it from loading down the source).
The Math in Action: A Worked Numeric Example
Let us build a standard non-inverting amplifier using a ubiquitous TL072 dual op-amp. The closed-loop voltage gain ($A_{CL}$) for a non-inverting topology is dictated entirely by the feedback network:
$A_{CL} = 1 + (R_f / R_{in})$
We will use $R_f = 10k\Omega$ and $R_{in} = 1k\Omega$.
Input Signal: 0.5V DC
Expected Output: 0.5V × 11 = 5.5V DC
Here is where the magic of negative feedback becomes visible. The TL072 has an open-loop gain ($A_{OL}$) of roughly 200,000 at DC. However, as frequency increases, $A_{OL}$ drops off at 20dB/decade. At 10 kHz, the internal gain might have fallen to just 2,000. If we relied on the chip's raw gain, our amplification would change wildly with frequency and temperature.
Because our feedback network forces the closed-loop gain to 11, and 11 is vastly smaller than the open-loop gain of 2,000 at 10 kHz, the op-amp still has massive 'reserve gain' to correct errors. The actual closed-loop gain at 10 kHz will remain virtually locked at 11. According to Texas Instruments' Precision Op-Amp Lab, the error introduced by finite open-loop gain is roughly $A_{CL} / A_{OL}$. In our 10 kHz scenario, the error is 11 / 2000, or a mere 0.55%.
Where You Meet Negative Feedback in Practice
You will rarely see an op-amp operating in open-loop mode outside of specialized comparator circuits. In practical electronics, negative feedback is everywhere:
- Audio Preamplifiers: Moving coil phono stages and microphone preamps use heavy negative feedback to flatten the frequency response and push Total Harmonic Distortion (THD) below 0.001%.
- Sensor Signal Conditioning: Strain gauges and load cells output tiny millivolt signals. Instrumentation amplifiers use precision feedback networks to boost these signals without amplifying thermal noise.
- Power Supply Regulation: The error amplifier inside a linear voltage regulator (like an LM317) uses negative feedback to compare a divided sample of the output voltage against an internal 1.25V reference, adjusting the pass transistor to maintain a rock-solid output.
- PID Motor Controllers: The 'Proportional' and 'Integral' terms in analog servo loops are implemented using op-amps with resistive and capacitive feedback networks.
Design Decision Tree: Picking Your Op-Amp and Feedback Network
Choosing the right amplifier and feedback components depends entirely on what signal you are processing. Use this decision matrix to lock in your bill of materials.
| Application Need | Primary Constraint | Recommended Op-Amp | Feedback Network Specs |
|---|---|---|---|
| High-Fidelity Audio | Low noise, high slew rate, good drive capability | NE5532 (or OPA1612 for premium) | Metal film resistors, keep $R_f \le 10k\Omega$ to minimize Johnson-Nyquist thermal noise. |
| Precision DC / Strain Gauges | Zero DC offset, zero drift over temperature | OPA2188 (Chopper-stabilized) | 0.1% tolerance thin-film resistors; match thermal coefficients (TCR) between $R_f$ and $R_{in}$. |
| High-Speed / Video | Wide bandwidth, fast settling time | LMH6702 (Current Feedback Amp) | Must use the exact $R_f$ value specified in the datasheet (usually ~1k\Omega). CFA stability depends on absolute resistance, not just the ratio. |
| High-Impedance Piezo Sensors | Ultra-low input bias current | OPA140 (JFET input) | Use high-value feedback resistors (1M\Omega+) but guard the PCB traces to prevent leakage currents from ruining the feedback ratio. |
Stability Pitfalls: When Negative Feedback Goes Wrong
Negative feedback assumes the signal returning to the inverting input is exactly 180 degrees out of phase with the input. However, every op-amp introduces internal phase shift at high frequencies. If the total phase shift reaches 360 degrees while the loop gain is still greater than 1, your negative feedback turns into positive feedback, and the circuit becomes an oscillator.
The most common way hobbyists and junior engineers trigger this is by driving a heavy capacitive load directly from the op-amp output. A long coaxial cable or a large decoupling capacitor interacts with the op-amp's internal output resistance, creating a low-pass filter that injects an additional 90 degrees of phase lag into the feedback loop.
The Fix: Never connect a large capacitor directly from the output pin to ground. Instead, isolate the capacitive load by placing a small series resistor (typically 22\Omega to 47\Omega) between the op-amp output pin and the load. Take your feedback signal from the op-amp side of this resistor, not the load side. This technique, detailed extensively in the Analog Devices Linear Circuit Design Handbook, restores the phase margin and guarantees stability.
Frequently Asked Questions
Can I use 1% carbon composition resistors for my feedback network?
You can, but you shouldn't for precision circuits. Carbon composition resistors generate excess 'flicker noise' (1/f noise) and have poor temperature coefficients. Always use 1% metal film or 0.1% thin-film resistors for feedback networks to keep noise low and gain stable across temperature variations.
Why do some schematics show a capacitor in parallel with the feedback resistor?
That is a compensation capacitor. Placing a small capacitor (often 10pF to 100pF) in parallel with $R_f$ creates a low-pass filter in the feedback path. This intentionally rolls off high-frequency gain to improve phase margin and prevent high-frequency oscillation, or to filter out RF interference before it reaches the amplifier stage.
What happens if my feedback resistor is too large?
If $R_f$ is in the megaohm range, the thermal noise generated by the resistor itself ($e_n = \sqrt{4kTR\Delta f}$) will overwhelm the op-amp's internal voltage noise. Furthermore, the stray parasitic capacitance of the PCB pads and the op-amp's input capacitance will form an unintended low-pass filter, severely limiting your bandwidth and potentially causing phase-shift oscillations.
