If you are staring at a breadboard wondering why your output is pegged to the rail, you are not alone. The operational amplifier is the workhorse of analog electronics, but a bad op amp setup will quietly ruin your day. The direct answer for most makers in 2026: if you need a default part for a 3.3V/5V microcontroller project, grab the MCP6002. For dual-supply audio or precision bench work, use the TL072. For gritty, single-supply 12V industrial sensing where pennies matter, the LM358 is still king.
This guide skips the abstract textbook math and goes straight to the bench. We will cover pinouts, operation limits, a complete working circuit, and a real-world war story about how a common setup mistake leads to silent failures.
The Bench Reality: Safe Defaults and Pinouts
Selecting an op amp is about matching the silicon to your power supply and signal speed. You do not need a $4.00 precision chopper-stabilized amplifier to read a slow-moving temperature sensor. Here are the three safe defaults you should keep in your parts bin, complete with their critical ratings.
| Part Number | VCC Range | GBP (Gain-Bandwidth) | Slew Rate | Approx. Price | Best Application |
|---|---|---|---|---|---|
| MCP6002 | 1.8V to 6.0V | 1 MHz | 0.6 V/µs | $0.45 | 3.3V/5V MCU ADC buffering, true rail-to-rail I/O |
| LM358 | 3V to 32V (Single) | 1 MHz | 0.3 V/µs | $0.15 | 12V/24V industrial sensing, low-side current shunts |
| TL072 | ±5V to ±15V (Dual) | 3 MHz | 13 V/µs | $0.35 | Dual-supply audio, active filters, lab equipment |
Standard 8-Pin DIP/SOIC Pinout
Most dual op amps (including the three listed above) share the industry-standard 8-pin layout. When reading the dot or notch on the IC, Pin 1 is top-left:
- Pin 1: Output A
- Pin 2: Inverting Input A (IN-)
- Pin 3: Non-Inverting Input A (IN+)
- Pin 4: V- (Ground for single-supply, or negative rail for dual-supply)
- Pin 5: Non-Inverting Input B (IN+)
- Pin 6: Inverting Input B (IN-)
- Pin 7: Output B
- Pin 8: V+ (Positive power rail)
Operation Regions and the "Rail-to-Rail" Myth
An op amp setup only behaves predictably in its linear region. Push it outside those boundaries, and the internal transistors saturate, clip, or suffer phase inversion. Understanding these regions is how you debug a circuit that "should work according to the math."
| Operation Region | Condition | Typical Vout | Typical Iout | What You See on the Scope |
|---|---|---|---|---|
| Linear (Active) | V- < Vin < V+ (within CMV limits) | VCC/2 (or dictated by feedback) | < 20mA (sourcing/sinking) | Clean amplification, Vout = Vin × Gain |
| Positive Saturation | IN+ significantly > IN- | VCC - 1.5V (LM358) or VCC - 0.05V (MCP6002) | Max rated (e.g., 30mA short-circuit limit) | Flatlined at the top rail; clipping |
| Negative Saturation | IN- significantly > IN+ | GND + 0.8V (LM358) or GND + 0.02V (MCP6002) | Max rated sinking current | Flatlined at the bottom rail |
| CMV Violation | Inputs exceed VCC - 1.5V (on non-RRIO parts) | Unpredictable (often snaps to 0V) | Varies wildly | Phase inversion or sudden drop to zero |
Application Circuit: 10A Shunt Amplifier for ESP32
Let us build a complete, working circuit. The ESP32 ADC expects a 0V to 3.3V signal. We want to measure up to 10A using a 0.1Ω low-side current shunt. At 10A, the shunt drops 1.0V. We need a non-inverting amplifier with a gain of 3.3 to map 0-1.0V to 0-3.3V.
Component Values
- U1: MCP6002 (Single supply 3.3V, rail-to-rail I/O)
- R_shunt: 0.1Ω, 2W power resistor
- Rg (Gain set to GND): 10kΩ (1% tolerance)
- Rf (Feedback resistor): 23.2kΩ (1% tolerance). Gain = 1 + (23.2k / 10k) = 3.32
- C_decouple: 100nF (0.1µF) ceramic capacitor
- C_filter: 10nF across Rf (limits bandwidth to kill high-frequency noise)
Build Steps
- Power the IC: Connect 3.3V to Pin 8 (V+) and GND to Pin 4 (V-). Place the 100nF decoupling capacitor physically within 2mm of these pins. Skip this, and your ESP32 will read garbage due to digital noise coupling into the analog stage.
- Wire the Shunt: Place the 0.1Ω shunt on the low side (between the load's ground and the system ground). Connect the high side of the shunt to Pin 3 (IN+).
- Set the Bias/Gain: Connect Pin 2 (IN-) to GND through the 10kΩ Rg resistor. Connect Pin 2 (IN-) to Pin 1 (Output) through the 23.2kΩ Rf resistor.
- Add the Filter: Solder the 10nF capacitor in parallel with Rf. This creates a low-pass filter, rolling off noise above ~680 Hz.
- Route the Output: Connect Pin 1 (Output) to your ESP32 GPIO (e.g., GPIO 34) configured as an ADC input.
War Story: The High-Side Sense Disaster
Theory is clean; the workbench is messy. A few years ago, I was designing a battery monitor for a 12V lead-acid bank. To avoid ground loops, I decided to measure current on the high side using an LM358. Here is exactly how the setup went wrong.
The Setup
The battery was at 12.6V. The 0.05Ω shunt was placed on the positive terminal. The load was drawing 5A, meaning the voltage across the shunt was 250mV. The top of the shunt was at 12.6V, and the bottom was at 12.35V. I powered the LM358 directly from the 12.6V battery (VCC = 12.6V, GND = 0V) and fed the shunt voltages into the IN+ and IN- pins.
The Numbers and Outcome
According to the differential amplifier math, the output should have been a beautifully scaled voltage representing 5A. Instead, the output pin read 0.0V. I swapped the IC. Still 0.0V. I checked the wiring three times. The math was perfect.
What Went Wrong
I had violated the Input Common-Mode Voltage (CMV) range. The LM358 datasheet states the inputs must be at least 1.5V below VCC. My VCC was 12.6V, meaning the maximum allowable input voltage was 11.1V. My inputs were sitting at 12.35V and 12.6V. The internal PNP input stage completely starved of bias current, and the output stage collapsed to ground.
When Silicon Dies: Multimeter Testing and Failure Modes
Op amps rarely die of old age. They die from abuse. The most common failure modes are latch-up (exceeding the absolute maximum supply voltage, causing internal parasitic thyristors to short the rails), ESD strikes on the input pins, and output short-circuits that melt the internal output transistors.
If you suspect a dead IC in your op amp setup, do not just swap it blindly. Find out why it died, and test the suspect part with a digital multimeter (DMM).
How to Test an Op Amp with a DMM
Remove the IC from the circuit (or lift the power pins) to avoid parallel resistance paths. Set your DMM to Diode Test Mode.
- Input Protection Diodes: Place the red probe on Pin 4 (V-) and the black probe on Pin 2 (IN-). You should read a forward voltage drop of roughly 0.6V to 0.8V. Reverse the probes; it should read "OL" (Open Loop). Repeat for Pin 3 (IN+). If you read 0.0V (short) or OL in both directions, the input ESD diodes are blown.
- Output Stage Shorts: Switch the DMM to Resistance (Ohms) mode. Measure between Pin 1 (Output) and Pin 8 (V+), then Pin 1 and Pin 4 (V-). A healthy IC will show high resistance (typically >10kΩ). If you read less than 5Ω, the internal output push-pull transistors have melted together from a short-circuit fault.
- Rail-to-Rail Shorts (Latch-up): Measure resistance between Pin 8 (V+) and Pin 4 (V-). It should not be a dead short. If it reads near 0Ω, the silicon has suffered catastrophic latch-up or thermal runaway.
For deeper theory on internal op amp architectures and how input stage topologies dictate these failure modes, the All About Circuits semiconductor textbook provides an excellent, free breakdown of the internal differential pairs.
Mastering the op amp setup is less about memorizing formulas and more about respecting the physical limits of the silicon. Pick the right default part for your supply rails, decouple your power pins, respect the common-mode input range, and your analog front-end will work the first time you apply power.
