An op amp instrumentation amplifier is a closed-loop differential amplifier engineered for high input impedance and extreme Common-Mode Rejection Ratio (CMRR). Unlike a basic difference amplifier that loads down high-impedance sources, an instrumentation amp extracts microvolt-level differential signals buried in volts of common-mode noise—making it the undisputed standard for strain gauges, thermocouples, and medical ECG front-ends.
If you need a reliable starting point, here are the safe default part numbers with their core ratings:
- Texas Instruments INA128: ±2.25V to ±18V dual supply, 700µA quiescent current, 120dB CMRR. (~$4.50) - INA128 Datasheet
- Analog Devices AD620: ±2.3V to ±18V dual supply, 900µA quiescent current, 120dB CMRR. (~$6.00) - AD620 Product Page
- TI INA333: 1.8V to 5.5V single/dual supply, 50µA quiescent current, zero-drift architecture. (~$3.20)
Internal Architecture and Pinout Breakdown
Most monolithic instrumentation amplifiers rely on the classic three-op-amp topology. The input stage consists of two non-inverting operational amplifiers configured as buffers. This provides the high input impedance and allows the differential gain to be set by a single external resistor. The output stage is a precision difference amplifier that subtracts the common-mode voltage and provides a single-ended output.
Here is the standard 8-pin DIP pinout description, which applies directly to the INA128, AD620, and most industry-standard variants:
- Pin 1 & Pin 8 (RG): Gain Resistor pins. Placing a resistor across these pins sets the differential gain of the input stage.
- Pin 2 (IN-): Inverting input. Connects to the low-side of your differential sensor.
- Pin 3 (IN+): Non-inverting input. Connects to the high-side of your differential sensor.
- Pin 4 (V-): Negative supply rail. Requires a 100nF ceramic decoupling capacitor placed within 2mm of the pin.
- Pin 5 (REF): Reference pin. Used to offset the output voltage. Tie to ground for dual-supply operation, or to a mid-rail voltage for single-supply ADC interfacing.
- Pin 6 (VOUT): Amplifier output. Low impedance, capable of driving 10kΩ loads directly.
- Pin 7 (V+): Positive supply rail. Requires a 100nF ceramic decoupling capacitor.
Selecting and Biasing for the Job
Biasing an op amp instrumentation amplifier correctly is where most hobbyist designs fail. The inputs must remain within the specified common-mode voltage range relative to the supply rails. If you are running a single 5V supply, you cannot ground the REF pin and expect to measure a 10mV signal—the output cannot swing below 0V. You must bias the REF pin to a mid-rail voltage (e.g., 2.5V) so the output can swing both up and down from that reference point.
Below is an operation regions table comparing our three default ICs to help you select the right silicon for your power budget and signal chain.
| Parameter | TI INA128 | ADI AD620 | TI INA333 |
|---|---|---|---|
| Supply Voltage Range | ±2.25V to ±18V | ±2.3V to ±18V | 1.8V to 5.5V (Single/Dual) |
| Quiescent Current (Iq) | 700 µA | 900 µA | 50 µA |
| Input Bias Current | ±2 nA | ±1 nA | ±20 pA |
| Min CMRR (G=10) | 105 dB | 100 dB | 100 dB |
| Gain Bandwidth Product | 1.5 MHz | 1.2 MHz | 150 kHz |
| Best Application | General industrial | High-precision lab | Battery-powered / wearables |
Application Circuit: Precision Load Cell Interface
Let us build a complete, copy-pasteable circuit for interfacing a standard 2mV/V wheatstone bridge load cell to a 3.3V microcontroller ADC. We will use the INA128.
Design Parameters:
- Load Cell Excitation: 5V
- Load Cell Sensitivity: 2mV/V
- Full-Scale Output: 10mV (at 5V excitation)
- Target ADC Input: 0V to 3.3V
- Required Gain: 3.3V / 10mV = 330 V/V
Gain Resistor Calculation:
The INA128 gain equation is G = 1 + (50kΩ / RG).
Solving for RG: RG = 50kΩ / (G - 1)
RG = 50,000 / (330 - 1) = 151.97 Ω
Select the closest 1% standard resistor value: 150 Ω (yielding a real gain of 334 V/V, max output 3.34V).
Complete Component List & Wiring Steps:
- U1 (INA128): Place the IC across the breadboard center trench.
- RG (150Ω 1% Metal Film): Connect directly across Pin 1 and Pin 8. Keep leads under 5mm to prevent parasitic inductance.
- Decoupling: Solder a 100nF X7R ceramic capacitor from Pin 7 (V+) to Ground, and another from Pin 4 (V-) to Ground. Place them physically touching the IC pins.
- Power Supply: Apply +5V to Pin 7 and Ground to Pin 4. (Running single supply).
- Reference Biasing: Because we are on a single supply, tie Pin 5 (REF) to a stable 100mV offset. You can create this with a simple voltage divider (10kΩ and 200Ω) from the 5V rail, buffered by a unity-gain op-amp to prevent loading. This ensures the output never hits the 0V rail limit, avoiding dead-zone non-linearity.
- Inputs: Connect the load cell's Signal+ to Pin 3 and Signal- to Pin 2. Add 100Ω series resistors on both input lines to protect against ESD and fault currents.
- Output Filtering: Place a 10nF capacitor from Pin 6 (VOUT) to Ground to create a low-pass filter rolling off high-frequency VFD and switching noise.
Failure Modes and Multimeter Diagnostics
Instrumentation amplifiers rarely fail spontaneously; they are almost always killed by external abuse. The most common failure modes are input overvoltage (exceeding the supply rails by more than 0.5V, which forward-biases internal ESD diodes and melts the bond wires) and latch-up from fast transient spikes on the power supply.
If your circuit is misbehaving, do not just swap the chip. Follow this multimeter diagnostic sequence to isolate the fault:
- Verify Supply Rails: Set your multimeter to DC Volts. Probe Pin 7 and Pin 4. You must read exactly your expected supply voltage. A reading of 2V on a 5V rail indicates a shorted decoupling capacitor or a damaged IC pulling excessive current.
- Check the REF Pin: Probe Pin 5. If this pin is left floating, the output stage will behave erratically and likely saturate. It must be tied to a low-impedance DC voltage.
- The Shorted-Input Test: This is the definitive health check. Power down, and use a jumper wire to short Pin 2 (IN-) directly to Pin 3 (IN+). Power the circuit back up. Set your multimeter to DC mV and probe Pin 6 (VOUT). The output voltage must exactly match the voltage present on your REF pin (Pin 5). If VOUT is pegged to the positive rail while the inputs are shorted, the internal output op-amp is destroyed.
- Measure Input Common-Mode Voltage: Remove the short. Probe Pin 2 and Pin 3 relative to ground. Ensure both voltages sit comfortably between V- and V+. If you read 5.2V on the inputs while running a 5V single supply, your sensor excitation is exceeding the IC's absolute maximum ratings.
For deeper architectural theory on how the internal difference stage rejects common-mode noise, refer to this excellent breakdown on instrumentation amplifier fundamentals by All About Circuits.
Frequently Asked Questions
What is the difference between an op amp instrumentation amplifier and a standard difference amplifier?
A standard difference amplifier uses four resistors and a single op-amp. Its input impedance is relatively low and determined by those resistors, meaning it will load down high-impedance sensors like piezoelectric elements. An op amp instrumentation amplifier adds two buffer op-amps at the inputs, providing gigohm-level input impedance. Furthermore, the CMRR of a standard difference amplifier relies entirely on the physical matching of four discrete resistors (which drift with temperature), whereas an instrumentation amplifier's CMRR is laser-trimmed internally at the silicon level, yielding vastly superior noise rejection.
How do I calculate the gain resistor for an op amp instrumentation amplifier?
The formula depends on the specific IC's internal feedback network. For the industry-standard INA128 and AD620, the equation is G = 1 + (50kΩ / RG). To find the resistor value, rearrange it to RG = 50kΩ / (G - 1). Always use 1% tolerance metal film resistors or better for RG; a 5% carbon resistor will introduce massive gain errors and temperature drift that defeats the purpose of using a precision amplifier.
Why is my op amp instrumentation amplifier output saturated at the supply rail?
Output saturation almost always points to one of three issues: First, the input common-mode voltage exceeds the allowable range (the inputs are too close to the positive or negative supply rails). Second, the REF pin is floating or improperly biased, forcing the output stage out of its linear region. Third, the required output voltage (Input Differential × Gain + VREF) exceeds the physical output swing limits of the IC, which typically cannot reach closer than 1V to the supply rails on older bipolar designs like the INA128.
Can I build an op amp instrumentation amplifier using discrete LM358 op amps?
You can wire three LM358 op-amps to mimic the topology, but you should not do it for precision work. The CMRR of a discrete build is strictly limited by how perfectly you can match the resistors in the output difference stage. Even with 0.1% resistors, temperature gradients across your PCB will cause the resistor ratios to drift, destroying your common-mode rejection. Furthermore, the LM358 has high input offset voltage and crossover distortion. For anything beyond audio-frequency hobby experiments, a $4 monolithic IC like the INA128 will outperform a discrete LM358 build by orders of magnitude.
