--- title: How to Design a Low‑Noise Voltage Feedback Amplifier for Precision Sensors siteUrl: https://logzly.com/feedbackamps author: feedbackamps (Feedback Amplifier Hub) date: 2026-06-22T00:05:38.521326 tags: [lownoise, opamp, circuits] url: https://logzly.com/feedbackamps/how-to-design-a-lownoise-voltage-feedback-amplifier-for-precision-sensors --- When a tiny voltage from a pressure sensor or a photodiode has to be turned into a usable signal, the noise added by the amplifier can drown out the very thing you are trying to measure. That’s why a low‑noise voltage feedback amplifier (VFA) is the unsung hero of any precision measurement system. In this post I’ll walk you through a practical design flow that keeps the noise low, the gain stable, and the layout tidy—exactly the kind of recipe I love to share on Feedback Amplifier Hub. ## Why Noise Matters More Than You Think Imagine you are trying to hear a whisper in a crowded room. If your ears (the sensor) are perfect but the microphone (the amplifier) adds a hiss, you’ll never catch the whisper. In many industrial and scientific applications the signal we care about is only a few microvolts. Adding even a few nanovolts of extra noise can shift the reading by a noticeable amount. That is why we start with noise budgeting right at the top of the design. ## Step 1: Choose the Right Op‑Amp ### Look for Low Input‑Referred Voltage Noise The first place to cut noise is the op‑amp itself. Input‑referred voltage noise (often quoted as nV/√Hz) tells you how much random voltage the device injects at its inputs. For precision sensors, aim for an op‑amp with a noise density below 5 nV/√Hz at 1 kHz. Devices like the OPA211 or AD8628 are popular choices because they combine low noise with good bandwidth. ### Consider Current Noise Too If your sensor source impedance is high (tens of kilo‑ohms or more), the op‑amp’s input current noise can become a bigger problem than voltage noise. Current noise (pA/√Hz) multiplied by the source resistance creates extra voltage noise. In those cases pick an op‑amp with low current noise, such as a JFET‑input device. ### Power Supply Rejection Ratio (PSRR) A high PSRR helps keep power‑rail ripple from leaking into the output. Look for a PSRR of at least 80 dB at the frequencies of interest. This is especially important when you are powering the circuit from a switching regulator. ## Step 2: Set the Gain Wisely ### Use a Modest Closed‑Loop Gain The closed‑loop gain (the ratio of output voltage to input voltage) determines how much the op‑amp’s own noise is amplified. A gain of 10 to 100 is typical for sensor front ends. Higher gain pushes more of the op‑amp’s noise to the output and can also make the circuit more prone to instability, a problem detailed in our [op‑amp stability mistakes](/feedbackamps/5-common-opamp-stability-mistakes-and-how-to-fix-them-in-analog-signal-chains) article. ### Calculate the Feedback Resistor Values The gain is set by the ratio of the feedback resistor (Rf) to the ground resistor (Rg). Choose values that keep the thermal (Johnson) noise low. Thermal noise voltage is √(4kTRΔf), where k is Boltzmann’s constant, T is temperature, R is resistance, and Δf is bandwidth. Using resistors in the 1 kΩ to 10 kΩ range usually gives a good balance between low thermal noise and manageable current draw. ### Keep Resistor Tolerance Tight A 0.1 % metal‑film resistor pair will keep gain error under control and also reduce the chance of gain drift with temperature. I still keep a small stash of these in my lab drawer for exactly this reason. ## Step 3: Bandwidth and Filtering ### Bandwidth Must Match the Sensor A wide bandwidth lets more noise in. Use the gain‑bandwidth product (GBW) of the op‑amp to set a realistic upper limit. For a sensor that only needs to respond up to 10 kHz, a low‑pass filter at 12 kHz is sufficient. A simple RC filter at the output or a passive RC network in the feedback loop can do the job. ### Add a Small Capacitor Across Rf Placing a few picofarads across the feedback resistor creates a pole that rolls off gain at higher frequencies, reducing high‑frequency noise. Be careful not to add too much capacitance, or you may introduce phase lag that leads to oscillation. ## Step 4: Power Supply Decoupling ### Bypass Capacitors Close to the Op‑Amp A 0.1 µF ceramic capacitor right at the power pins, followed by a 10 µF electrolytic a few millimeters away, forms a low‑impedance path for high‑frequency supply noise. I always solder the 0.1 µF first, then the larger one, to avoid solder bridges. ### Use a Clean Supply If you can, power the amplifier from a low‑noise linear regulator rather than a switching supply. If a switching regulator is unavoidable, add an LC filter on its output before it reaches the op‑amp. ## Step 5: PCB Layout Tips ### Short, Direct Signal Paths Keep the input traces from the sensor to the op‑amp as short as possible. Long traces act like antennas, picking up EMI. A 2‑inch trace is usually fine; anything longer should be shielded or routed on an internal layer. ### Ground Plane Strategy A solid ground plane under the analog section reduces loop area and provides a low‑impedance return path. Separate the analog ground from digital ground if your board also contains a microcontroller. Connect the two grounds at a single point near the power entry. ### Star Ground for Feedback Network The junction of Rf and Rg is a good place to star‑ground the feedback network. This prevents ground currents from one part of the circuit from modulating the feedback voltage. ## Step 6: Verify the Design ### Simulate Noise with SPICE Run a transient noise analysis in your SPICE tool. Look at the output noise spectral density and compare it to the sensor’s own noise. If the amplifier’s contribution is more than half of the total, go back and tweak resistor values or choose a quieter op‑amp. ### Build a Test Board Even the best simulation can miss layout‑induced noise. Build a quick prototype on a small FR‑4 board, measure the output with a low‑noise spectrum analyzer, and compare it to the simulated result. I once found that a stray capacitance of just 2 pF between the input and ground added a noticeable bump at 50 kHz—something the simulation missed. ## A Personal Note When I first taught a lab on sensor front ends, a student tried to use a high‑gain, high‑bandwidth op‑amp without any filtering. The oscilloscope showed a wild, noisy trace that looked like a sea of static. After we added a modest gain of 20, a 10 kHz low‑pass filter, and swapped the op‑amp for a low‑noise JFET type, the signal cleared up like a fog lifting. That moment reminded me why I love sharing these practical tips on Feedback Amplifier Hub—seeing the “aha” in a student’s eyes is worth every ounce of circuit‑board sweat. ## Quick Checklist - Pick an op‑amp with < 5 nV/√Hz voltage noise and low current noise if source impedance is high. - Set gain between 10 and 100 using 1 kΩ–10 kΩ resistors, keep tolerance at 0.1 %. - Add a small capacitor (2–5 pF) across Rf to roll off high‑frequency gain. - Decouple power pins with 0.1 µF ceramic + 10 µF electrolytic close to the device. - Keep signal traces short, use a solid analog ground plane, and star‑ground the feedback network. - Simulate noise, then verify on a real board. Designing a low‑noise VFA is a blend of careful component choice, thoughtful gain setting, and disciplined layout. Follow these steps, and your precision sensor will speak clearly, without the unwanted hiss. For an exhaustive, step‑by‑step walkthrough, see our [step‑by‑step design guide](/feedbackamps/designing-a-lownoise-voltagefeedback-amplifier-for-precision-sensors-a-stepbystep-guide).