An amplifier is an electronic circuit that uses a small input signal to control a larger power source, outputting a magnified replica of the original waveform. In a real circuit, it changes the signal's amplitude—voltage, current, or power—without altering the fundamental frequency or data payload. Beginners commonly confuse amplifiers with transformers; while a transformer can step up AC voltage via magnetic induction, it cannot amplify DC signals, add net power to the system, or actively drive a low-impedance load from a high-impedance source.
The Core Mechanics of Signal Amplification
To understand how amplification actually works on the bench, you have to look at the active device at the heart of the circuit—usually a Bipolar Junction Transistor (BJT), a MOSFET, or an integrated operational amplifier (op-amp). These components do not create energy. Instead, they act as electrically controlled valves.
Think of the active device as a water valve: the input signal is the hand turning the handle, and the DC power supply is the pressurized water main. The valve doesn't create water; it just modulates the massive flow from the main based on the tiny movements of your hand. In electronics, a millivolt-level input signal modulates the current flowing from a 12V or 24V DC rail to the load.
We measure this magnification as gain. Voltage gain ($A_v$) is the ratio of output voltage to input voltage, often expressed in decibels (dB) using the formula $A_v(dB) = 20 \log_{10}(V_{out} / V_{in})$. Typical small-signal voltage gains in sensor conditioning range from 10 to 100,000 (20 dB to 100 dB), while power amplifiers prioritize current delivery over raw voltage multiplication.
Comparing the Main Kinds of Amplifier Classes
When engineers talk about the "kinds of amplifier" they are usually referring to the amplifier's class, which defines how the active devices are biased and how much of the input waveform cycle they conduct current. This choice dictates the trade-off between linearity (audio fidelity/signal accuracy) and efficiency (heat generation).
| Class | Conduction Angle | Max Theoretical Efficiency | Crossover Distortion | Primary Application |
|---|---|---|---|---|
| Class A | 360° (Full cycle) | 25% - 50% | None | High-end audio, ultra-low noise RF front-ends |
| Class B | 180° (Half cycle) | 78.5% | Severe | Rarely used alone; mostly in push-pull pairs |
| Class AB | 180° to 360° | 50% - 78.5% | Low (biased past cutoff) | Studio monitors, headphone amps, op-amp output stages |
| Class D | N/A (PWM switching) | > 90% | N/A (Switching noise instead) | Bluetooth speakers, subwoofers, motor drives |
Where You Meet This in Practice
You will encounter different amplifier topologies depending on whether you are moving watts or moving data.
- Audio Power Systems: If you are building a DIY smart speaker or a garage sound system, you will almost exclusively use Class D amplifier modules. Boards based on the TI TPA3116D2 chip can deliver 50W per channel into 4 ohms from a 24V laptop brick, barely needing a heatsink. The efficiency is so high that the heat generated at idle is negligible.
- Sensor Signal Conditioning: When reading a thermocouple or a strain gauge, you need an Instrumentation Amplifier (a specialized configuration of precision op-amps). Chips like the INA128 reject common-mode noise (like 60Hz mains hum) while amplifying microvolt-level differential signals. These operate in Class A or AB internally to maintain strict linearity.
- RF Transceivers: In software-defined radio (SDR) or ESP32 WiFi front-ends, you use Low Noise Amplifiers (LNAs). These are optimized to add as little internal thermal noise as possible to the tiny microvolt signals picked up by an antenna, prioritizing a low Noise Figure (NF) over raw power output.
Bench Walkthrough: When Gain Calculations Go Wrong
Theory is clean; the workbench is not. Let’s walk through a real-world scenario where ignoring amplifier headroom leads to a clipped, useless signal.
The Setup: You are building a digital scale using a 5kg load cell and an Arduino Uno. The load cell has a sensitivity of 2mV/V. You excite it with a 5V supply, meaning at maximum load (5kg), the differential output signal is exactly 10mV. You choose an INA128 instrumentation amplifier to boost this signal to the Arduino’s 0-5V ADC range.
The Numbers: To map 10mV to 5V, you need a voltage gain of 500. According to the TI INA128 datasheet, the gain equation is $G = 1 + (50k\Omega / R_g)$. To get a gain of 500, you calculate the gain resistor: $R_g = 50,000 / 499 \approx 100.2\Omega$. You solder a 100-ohm resistor across pins 1 and 8.
The Outcome: You power the INA128 from the Arduino’s single 5V rail (V+ to 5V, V- to GND). You apply 5kg to the load cell. The Arduino serial monitor reads a maximum value of 780 out of 1023 (about 3.8V), no matter how hard you press. The signal is clipping.
What Went Wrong: The INA128 is not a rail-to-rail output amplifier. It requires "headroom"—a buffer zone between its output voltage and its power supply rails. On a single 5V supply, the INA128's output can typically only swing from about +1.2V to +3.8V. By aiming for a 5V output, you drove the internal transistors into saturation. The amplifier clipped the top off your waveform.
- Fix Option A (Hardware): Provide a dual power supply. Feed the INA128 V+ with +9V and V- with -5V (or GND with a virtual ground offset). This gives the output stage the headroom it needs to swing past 5V, which you can then scale down with a voltage divider before the Arduino ADC.
- Fix Option B (Component Swap): Replace the INA128 with a modern, true rail-to-rail instrumentation amplifier like the INA333 or AD8237, which can swing within millivolts of a single 5V supply rail.
- Fix Option C (Gain Reduction): Lower the gain to 350 (using a ~143 ohm $R_g$). This maps the 10mV signal to 3.5V, safely within the INA128's linear output range on a 5V supply, though you lose some ADC resolution.
Frequently Asked Questions
Can I use a Class D audio amplifier to amplify a DC sensor signal?
No. Class D amplifiers rely on pulse-width modulation (PWM) and require an AC-coupled LC output filter designed specifically for audio frequencies (20Hz - 20kHz). They cannot pass DC or very low-frequency signals. If you try to feed a load cell or thermocouple into a Class D amp, the output filter will block the DC component entirely, and the switching noise will ruin your ADC readings. Always use a linear amplifier (Class A, AB, or a dedicated instrumentation amp) for DC sensor signals.
Why do op-amps have a "slew rate" specification?
Slew rate defines how fast the amplifier's output voltage can change, measured in Volts per microsecond (V/µs). Even if an op-amp has a massive gain-bandwidth product, a low slew rate will cause high-frequency, high-amplitude signals to distort into triangular waves because the internal compensation capacitor cannot charge fast enough. For audio, a slew rate above 5 V/µs is generally adequate; for high-speed data acquisition, you may need current-feedback amplifiers with slew rates exceeding 1000 V/µs.
What is the difference between voltage gain and power gain?
Voltage gain ($A_v$) only looks at the ratio of output voltage to input voltage, regardless of the current. An op-amp might have a voltage gain of 100, but if it can only source 10mA, it cannot drive a 4-ohm speaker. Power gain ($A_p$) accounts for both voltage and current delivery ($P = V \times I$). Power amplifiers (like those driving speakers or RF antennas) are designed with robust output stages capable of sourcing amps of current, whereas voltage amplifiers (like sensor pre-amps) prioritize high impedance and low noise over current delivery.
