The short answer
In electronics, an embedded system is the combination of a processor (usually a microcontroller), its supporting circuitry and the firmware it runs, built into an electronic product to control it. It’s the part of the circuit that makes decisions. It reads inputs such as buttons and sensors, runs a program, and drives outputs such as LEDs, displays, motors, relays and radios.
From an electronics point of view, the embedded system isn’t a separate box. It’s a section of the product’s circuit board, often a single chip and a handful of components around it, designed alongside the power supply, sensors and output circuits it works with. For the broader definition, see what an embedded system is.
Where the embedded system sits in an electronic product
Most electronic products can be divided into a few blocks:
- Power: a battery, mains adapter or other source, plus regulators that provide stable voltages.
- Inputs: buttons, switches, sensors, microphones, antennas.
- Processing: the embedded system, which interprets inputs and decides what to do.
- Outputs: LEDs, displays, speakers, motors, heaters, relays, radios.
The embedded system is the processing block, connecting everything else. In a digital kitchen scale, for example, a load cell produces a tiny voltage, an amplifier and analog-to-digital converter turn it into a number, the microcontroller converts that number to grams, and it drives the display. The microcontroller also reads the tare button and turns the scale off after a period of inactivity to save the battery.
What an embedded system looks like on a circuit board
If you open an electronic product, the embedded system is usually easy to find: a square chip with many pins, surrounded by small components.
The microcontroller
The microcontroller (MCU) is the heart of most embedded systems. It’s a single chip containing a processor core, flash memory for the program, RAM for working data, and peripherals such as timers, analog-to-digital converters and communication interfaces. More demanding products may use a microprocessor or system-on-chip with external memory instead. Our guide to microcontrollers vs microprocessors explains the difference.
Clock source
The processor needs a clock to step through instructions. Many microcontrollers have a built-in oscillator, but designs that need accurate timing, such as USB, precise communication or timekeeping, often add an external crystal or resonator next to the chip. A 32.768 kHz crystal is common for real-time clocks, because it divides evenly down to one second.
Power supply and regulation
Microcontrollers typically run at low voltages such as 3.3 V or 1.8 V, while the product might be powered from a battery, a USB port or a higher-voltage supply. Voltage regulators provide a stable supply, and small decoupling capacitors placed close to the chip’s power pins smooth out rapid changes in current. Poor power design is a common cause of unreliable embedded systems.
Reset and supervision
A reset circuit makes sure the processor starts cleanly when power is applied. Many designs also use a brown-out detector (often built into the MCU) that holds the chip in reset if the voltage drops too low, and a watchdog timer that restarts the system if the firmware stops responding.
Inputs: sensors, buttons and signals
Inputs connect the embedded system to the outside world. Buttons and switches are read through digital input pins, often with pull-up or pull-down resistors and some form of debouncing. Sensors may produce analog voltages, digital pulses, or communicate over a bus like I2C or SPI. Inputs exposed to the outside may need protection components against static discharge or voltage spikes.
Outputs and driver circuits
A microcontroller pin can only supply a small current, so it usually can’t drive a motor, relay, heater or bright LED directly. Driver circuits handle the heavier load: transistors or MOSFETs for switching, H-bridges for motors, LED drivers, and relay drivers. The firmware controls these drivers through its output pins.
Communication interfaces
Embedded systems talk to other chips and devices using standard interfaces:
- UART for simple serial links, often used for debugging or talking to modules
- I2C and SPI for connecting sensors, memory chips and displays on the same board
- CAN for automotive and industrial networks
- USB and Ethernet for connections to computers and networks
- Wireless radios such as Bluetooth, Wi-Fi or sub-GHz links, sometimes built into the microcontroller itself
Programming and debug connections
Most boards include a small header or test pads for programming and debugging the chip, commonly using SWD (on Arm Cortex-M parts) or JTAG. Engineers connect a debug probe here to load firmware, set breakpoints and inspect memory. On finished products, these connections are often left unpopulated or disabled for security.
Crossing the analog-digital boundary
The real world is mostly analog: temperatures, light levels, sound and motor speeds vary continuously. Processors work with numbers. Embedded systems bridge the two:
- An analog-to-digital converter (ADC) measures a voltage and turns it into a number the firmware can use. A 12-bit ADC, for example, divides its input range into 4,096 steps.
- A digital-to-analog converter (DAC) does the reverse, producing a voltage from a number, for example to generate audio or a reference level.
- Pulse-width modulation (PWM) switches an output on and off rapidly. By varying the proportion of time it’s on, the firmware can control LED brightness, motor speed or heater power without a true analog output.
This is one of the main reasons embedded systems are so useful in electronics. A single chip can measure, decide and control, replacing what would otherwise be a complex analog circuit.
Why electronics moved from hardwired logic to embedded systems
Before cheap microcontrollers, electronic products used fixed circuits to implement their behavior: discrete logic chips, timers, analog comparators and mechanical switches. Changing how a product worked meant redesigning the circuit.
Embedded systems changed that:
- Behavior is defined in firmware. A new feature or a bug fix can often be made in software without changing the board.
- Fewer parts. One microcontroller can replace many logic chips, timers and analog stages, reducing size and cost.
- More capability. Displays, menus, communication, error detection and self-calibration become practical.
- One board, many products. Manufacturers can use the same hardware for several models and change only the firmware.
The trade-off is that the product now depends on software, which brings its own challenges in testing, security and updates. Our article on the advantages, disadvantages and challenges of embedded systems covers these.
Examples in consumer electronics
- TV remote: a tiny MCU scans the button matrix and drives an infrared LED with coded pulses.
- Digital thermometer: an ADC reads a temperature sensor, and the MCU calculates and displays the reading.
- Wireless earbuds: a Bluetooth system-on-chip decodes audio, manages the battery and handles touch controls.
- Phone charger with fast charging: a small controller negotiates the charging voltage with the phone.
- Electric toothbrush: an MCU runs the motor, times brushing intervals and manages the battery.
- Smart TV: a powerful system-on-chip decodes video and runs the user interface, with smaller controllers handling the power supply and remote receiver.
For many more examples across industries, see our guide to embedded system examples and applications.
Which electronic circuits are not embedded systems
Not every electronic circuit is an embedded system. A circuit without a programmable processor, such as a simple flashlight, a passive audio filter, a basic power supply or a blinking-LED circuit built around a 555 timer chip, works purely through its components. There’s no firmware making decisions.
The dividing line is the presence of a processor running a program that defines the product’s behavior. Once you add a microcontroller to read inputs and control outputs, the circuit becomes an embedded system.
Getting started with embedded electronics
Development boards make it easy to experiment without designing a circuit board first. An Arduino Uno, a Raspberry Pi Pico or an ESP32 board provides a microcontroller, power regulation, a clock and a USB programming connection on one board, so you can connect sensors and outputs on a breadboard. Our articles on whether Arduino is an embedded system and whether a Raspberry Pi is an embedded system explain how those platforms fit in.
Frequently asked questions
Is an embedded system hardware or software?
Both. An embedded system is the combination of hardware (the processor, memory and circuitry) and the firmware that runs on it. Neither does anything useful without the other. For more on the software side, see firmware vs software.
Is a microcontroller an embedded system?
A microcontroller is the main component of most embedded systems, but on its own it’s just a chip. It becomes an embedded system when it’s built into a circuit with inputs, outputs and power, and programmed to do a specific job.
Is embedded systems part of electronics or computer science?
It sits between the two. Embedded engineers need to understand circuits, signals and hardware interfaces from electronics, as well as programming, data structures and software design from computer science. University courses in embedded systems are offered in electrical, electronic and computer engineering departments.