The transfer function of an op amp defines the mathematical relationship between its output and input voltages, typically expressed in the Laplace domain as $H(s) = V_{out}(s) / V_{in}(s)$. While textbooks treat this as abstract calculus, on the workbench, the transfer function is the blueprint that dictates your circuit's closed-loop gain, bandwidth limits, phase margin, and stability. If you don't understand how feedback shapes this function, your amplifier will oscillate, clip, or act as an unintentional radio receiver.
This guide strips away the pure physics and focuses on how to select, bias, design with, and test operational amplifiers in real-world analog circuits.
The Op-Amp Symbol, Pinout, and Safe Default Part Numbers
Before calculating transfer functions, you need to know the physical interfaces. The standard op-amp symbol features a non-inverting input (+), an inverting input (-), positive and negative power supply rails ($V_{CC}$ and $V_{EE}$), and a single output. In the ubiquitous 8-pin DIP or SOIC dual-op-amp package, the pinout is standardized:
- Pin 1: Output A
- Pin 2: Inverting Input A (-)
- Pin 3: Non-Inverting Input A (+)
- Pin 4: $V_{EE}$ (Negative Supply / GND)
- Pin 5: Non-Inverting Input B (+)
- Pin 6: Inverting Input B (-)
- Pin 7: Output B
- Pin 8: $V_{CC}$ (Positive Supply)
Choosing the right silicon is half the battle. Here are the safe default part numbers I keep in my bench drawer, complete with their critical ratings:
| Part Number | Best For | Supply Range | GBW (Gain-Bandwidth) | Slew Rate | Approx. Cost |
|---|---|---|---|---|---|
| LM358 | General purpose, single-supply, low-cost | 3V to 32V | 1 MHz | 0.3 V/µs | $0.15 |
| TL072 | Audio preamps, JFET inputs | ±5V to ±18V | 3 MHz | 13 V/µs | $0.40 |
| OPA2134 | High-fidelity audio, precision DC | ±2.5V to ±18V | 8 MHz | 20 V/µs | $3.50 |
| MCP6002 | Low-power 3.3V/5V MCU interfacing | 1.8V to 6.0V | 1 MHz | 0.6 V/µs | $0.25 |
Understanding the Transfer Function in Real Circuits
An op-amp's open-loop transfer function, $A(s)$, is essentially a low-pass filter with a massive DC gain ($A_{OL}$, often 100,000 or 100dB) and a dominant pole at a very low frequency (around 10 Hz to 20 Hz for the LM358). The equation is:
$A(s) = \frac{A_{OL}}{1 + s/\omega_p}$
Because $A_{OL}$ is so high and varies wildly with temperature and part tolerance, we never use op-amps open-loop for linear amplification. We apply negative feedback via a network with a transfer function $\beta(s)$. The resulting closed-loop transfer function becomes:
$H(s) = \frac{A(s)}{1 + A(s)\beta(s)}$
At frequencies where $A(s)\beta(s) \gg 1$, the equation simplifies beautifully to $H(s) \approx 1/\beta(s)$. This is the core principle of op-amp design: the circuit's behavior is dictated by the passive feedback components, not the semiconductor's internal quirks. For a deeper dive into how feedback shapes stability and phase margin, the Analog Devices Op-Amp Basics tutorial is the definitive reference.
Operation Regions and Biasing for the Job
The transfer function only holds true if the op-amp is operating in its linear region. If the output hits the supply rails, the math breaks down and the device acts as a comparator. Here is how the regions break down on the bench:
| Operation Region | Input Differential ($V_+ - V_-$) | Typical Output Voltage | Circuit State |
|---|---|---|---|
| Linear | ≈ 0V (within µV) | $V_{EE} + 1V < V_{out} < V_{CC} - 1V$ | Active negative feedback |
| Positive Saturation | > +2mV (Open Loop) | $V_{CC} - V_{drop}$ (e.g., 13.5V on 15V rail) | Comparator / Overdriven |
| Negative Saturation | < -2mV (Open Loop) | $V_{EE} + V_{drop}$ (e.g., -13.5V on -15V rail) | Comparator / Overdriven |
If you are running an LM358 off a single 9V battery, the negative rail ($V_{EE}$) is 0V. The transfer function requires the output to swing both positive and negative relative to the input. To fix this, create a 'virtual ground' at $V_{CC}/2$ (4.5V) using a voltage divider (two 10kΩ resistors) buffered by another op-amp, or at least bypassed with a 10µF capacitor to maintain a low-impedance AC ground. Reference all your input signals to this 4.5V node, not 0V.
Application Circuit: Unity-Gain Sallen-Key Low-Pass Filter
Let's apply the transfer function to a practical circuit. A Sallen-Key topology is the standard for active 2nd-order filters. Here, we will design a unity-gain low-pass filter with a cutoff frequency ($f_c$) of roughly 1.59 kHz.
Component Values:
- Op-Amp: TL072 (powered by ±12V)
- R1, R2: 10 kΩ (1% tolerance metal film)
- C1, C2: 10 nF (C0G/NP0 ceramic for low distortion)
Circuit Wiring:
- Input signal connects to R1.
- R1 connects to R2 and C1 at the non-inverting input (Pin 3).
- C1 connects from Pin 3 to Ground.
- C2 connects from Pin 3 to the Output (Pin 1).
- Tie the inverting input (Pin 2) directly to the Output (Pin 1) for unity gain.
The transfer function for this specific Sallen-Key circuit is:
$H(s) = \frac{1}{s^2 R_1 R_2 C_1 C_2 + s(R_1 C_1 + R_2 C_1) + 1}$
By selecting C0G/NP0 capacitors instead of standard X7R ceramics, you prevent the capacitance value from dropping under DC bias voltage, which would otherwise shift your cutoff frequency and alter the transfer function dynamically. For more on active filter topologies, consult the Texas Instruments Op-Amp Design Resources.
Bench Testing: How Op-Amps Fail and How to Test Them
Op-amps rarely fail gracefully. Common failure modes include output stage short circuits (from driving a capacitive load without a series isolation resistor), input stage blowout (from exceeding the differential input voltage rating), and latch-up (from input voltages exceeding the supply rails).
Here is how to test a suspected dead op-amp using a standard digital multimeter (DMM) without desoldering it, assuming you've removed power and discharged all capacitors:
- Set DMM to Diode/Continuity Mode.
- Check Supply Rails: Place the red probe on $V_{CC}$ (Pin 8) and black on $V_{EE}$ (Pin 4). It should read 'OL' (open). If it reads 0Ω or a low resistance, the internal bias network is shorted. Reverse probes; you should see a diode drop (0.6V - 1.2V) due to internal ESD protection diodes.
- Check Output to Rails: Measure from the Output pin to $V_{CC}$, then Output to $V_{EE}$. Both should read 'OL' in one direction and a diode drop in the other. A reading of 0.00V means the output push-pull transistors have melted into a short.
- Check Inputs to Rails: Measure the inverting and non-inverting pins against both supply rails. You should never see a dead short (0Ω). A short here indicates the input differential pair has been destroyed, usually by a static discharge or an overvoltage event.
Frequently Asked Questions
How does the transfer function of an op amp change with frequency?
As frequency increases, the open-loop gain $A(s)$ drops at a rate of -20dB per decade past the dominant pole. Because the closed-loop transfer function relies on the loop gain ($A(s)\beta(s)$) being much greater than 1, your circuit will only maintain its ideal transfer function up to the bandwidth limit. This limit is calculated by dividing the op-amp's Gain-Bandwidth Product (GBW) by your circuit's closed-loop gain. For example, an LM358 (1MHz GBW) configured for a gain of 100 will only maintain its transfer function up to 10 kHz.
Why is my op amp transfer function oscillating in simulation but failing on the bench?
Simulations often use idealized power supplies and ignore parasitic inductance. On the bench, if your power supply bypass capacitors are missing or placed too far from the $V_{CC}$/$V_{EE}$ pins, the transient current demands of the output stage will cause local supply voltage dips. This injects noise back into the input stage via the power supply rejection ratio (PSRR), creating a parasitic feedback loop that alters the transfer function and causes high-frequency oscillation. Always place a 100nF MLCC capacitor within 2mm of the supply pins.
How do I calculate the transfer function of an op amp with multiple feedback loops?
When dealing with multiple feedback paths (like a multiple-feedback bandpass filter), use nodal analysis at the inverting input. Assume the op-amp is in the linear region, meaning the inverting input is at a virtual ground (or virtual $V_{ref}$). Write Kirchhoff's Current Law (KCL) equations summing all currents entering and leaving that node to zero. Solve the resulting algebraic equations for $V_{out}/V_{in}$. Tools like LTspice can verify your hand-derived transfer function by running an AC analysis and plotting the Bode diagram.
What happens to the transfer function when the op amp hits its slew rate limit?
The transfer function assumes linear, time-invariant operation. Slew rate is a large-signal, non-linear limitation defined by the maximum rate of change of the output voltage ($dV/dt$), dictated by the internal compensation capacitor charging current. If your input signal demands an output swing faster than the slew rate (e.g., 0.3 V/µs for the LM358), the output becomes a linear ramp instead of a sine wave. The transfer function effectively collapses, generating massive harmonic distortion and intermodulation products that no amount of negative feedback can correct.
