Skip to content
PLCampus
Course contents
  1. 1. What is a PLC?
  2. 2. How a PLC works inside
  3. 3. Types of PLC
  4. 4. Inputs and outputs
  5. 5. Wiring and connections
  6. 6. Ports and communications
  7. 7. Memory and addressing
  8. 8. IEC 61131-3 languages
  9. 9. Basic instructions
  10. 10. Real-world good practices
On this page

Fundamentals

How a PLC works inside

The parts of a PLC —power supply, CPU, memory and I/O modules— and the scan cycle it repeats endlessly.

5 min readReviewed: September 28, 2026

The main parts

From the outside, a PLC can be a single box or several modules mounted on a rail. On the inside, they all follow the same scheme:

Communication: PC, HMI, network
Sensors and buttons
Input module

CPU

Program memoryData memory
Output module
Actuators
Power supply

Power supply

Converts mains power (for example, 230 VAC) into the DC voltage the CPU and modules need, usually 24 VDC. Some units have it built in; in others it is a separate module or an external supply mounted in the same panel.

CPU

The PLC’s processor. It runs the program, manages memory, handles communications and monitors its own operation. Its front usually shows status indicators such as RUN, STOP and ERROR, and sometimes a physical mode switch.

Memory

Broadly speaking, it has two areas:

  • Program memory: stores the instructions you wrote. It is not erased when power is lost.
  • Data memory: stores the state of inputs and outputs, internal bits, timers, counters and other variables. Part of it can be configured as retentive (it keeps its value through a power loss) and the rest is reset on restart.

We will go deeper into this in the memory and addressing topic.

Input and output modules

They are the interface between the electrical world of the panel and the digital world of the CPU. They adapt signals (for example, from 24 VDC to logic levels) and electrically isolate them, usually with optocouplers, so that a voltage spike in the field does not damage the CPU. Each channel usually has an LED showing its state: a technician’s first diagnostic tool.

Communications

Every PLC has at least one port to connect to the programming computer. Most add network ports to talk to HMI panels, other PLCs, drives or supervisory systems.

The scan cycle

A PLC does not run its program just once: it runs it in an endless loop. Each pass is called a scan cycle (or simply a scan) and has four phases:

  1. Read inputs: copies the state of every physical input into a memory area called the process image of inputs.
  2. Execute the program: runs the program from top to bottom using that image and leaves the results in the process image of outputs.
  3. Write outputs: copies the output image to the physical outputs.
  4. Diagnostics and communication: serves communications, checks its health and starts again.

Try it yourself: pause the diagram, flip the switch in different phases and watch when the change actually reaches the output.

1 / 4

Read inputs

The PLC copies the state of every physical input into the input image. From now until the next cycle, the program works with that copy.

  1. Physical input
  2. Input imageI0.0 0
  3. ProgramQ0.0 := I0.0
  4. Output imageQ0.0 0
  5. Physical output

Flip the switch at any moment: the PLC will only notice it in the next “Read inputs” phase.

What is the process image for?

Copying the inputs at the start has a very important advantage: during the whole cycle, the program sees frozen inputs. If a sensor changes right while the program is running, the change does not affect half of the program and not the other half. The whole program works with the same “snapshot”.

Cycle time

The time a full pass takes is called the cycle time (or scan time). It depends on the CPU and the size of the program, and typically ranges from under a millisecond to a few tens of milliseconds. For a machine with buttons, sensors and motors, that is practically instantaneous.

The CPU monitors this time with a watchdog. If a cycle takes longer than the configured maximum —for example, because the program got stuck in a loop— the PLC stops and goes to STOP rather than keep controlling the machine with stale information.

Order matters

Because the program runs from top to bottom, the order of instructions changes the result. A classic case: if two different rungs write the same output, the value left at the end of the cycle is the one from the last rung.

Rung 1:   I0.0 ──┤ ├─────────( Q0.0 )
Rung 2:   I0.1 ──┤ ├─────────( Q0.0 )   ← this one decides the final value of Q0.0

Operating modes

  • RUN: the PLC executes the scan cycle and controls the outputs.
  • STOP: the program is not executed and outputs go to a safe state, usually off (on many devices this state can be configured).

Many PLCs also allow online monitoring: seeing from the computer, in real time, which contacts are closed and what values the timers hold. Some even accept program changes without stopping the machine, something that is used with great care.

What happens at power-up

When powered on, the CPU runs a self-test, checks the connected modules, restores retentive variables and resets the rest. If the mode switch is in RUN, the first scan cycle begins. Many PLCs also provide a special bit or block that is only active during that first scan, handy for initialising values.

Try it in the simulatorOpen a preloaded example and experiment yourself.

Summary

  • A PLC has a power supply, CPU, memory, I/O modules and communication ports.
  • It works in a loop: read inputs → execute program → write outputs → diagnostics.
  • The process image freezes inputs during the cycle; that is why instruction order matters.
  • A watchdog stops the PLC if a cycle takes too long.