Glossary · PLC and control programming
Watchdog timer
Also known as: Watchdog, Cycle time monitoring
German: Watchdog-Zeitgeber
In control systems, a watchdog timer is a supervision function that must be retriggered within a set time by the monitored program or device; if it is not, the watchdog detects a fault and initiates a defined reaction, such as stopping the controller or switching outputs off.
- PLC programming
- Functional safety
In one sentence
A watchdog timer detects a stalled or overrun program when it is not retriggered in time and triggers a defined fault reaction.
Example
The controller's cycle watchdog is set to 150 ms; when an endless loop blocks the program, the watchdog puts the controller into STOP and the outputs switch off.
How it applies
- Engineering: PLC operating systems monitor task and cycle times with watchdogs. Communication watchdogs detect lost connections to remote I/O or drives. The reaction (stop, alarm, substitute values) is configured and must match the intended Fault reaction.
- Functional safety: In safety controllers, watchdogs are part of the diagnostic measures that detect faults in the logic and processing, and contribute to the Diagnostic coverage (DC). Their contribution is established in the safety system's design, not assumed from the presence of a watchdog.
- Commissioning: Set watchdog times with margin above the measured maximum cycle time, but short enough to detect real faults quickly.
- Documentation: Record watchdog settings and the resulting behavior. Service documentation should explain what a watchdog stop looks like and how to diagnose its cause.
Watchdog vs. timeout
A Timeout limits how long one operation may take. A watchdog checks continuously that a program or device is still alive. Both detect missing activity; the watchdog supervises the whole system.
In risk assessment, PL analysis and security analysis
- Risk assessment (CE): The watchdog decides what happens when the logic hangs or overruns its cycle. Document the reaction (outputs de-energized, STOP, error routine) and check that it leads to the defined safe state for every safety function.
- PL analysis: Annex E of ISO 13849-1 lists monitoring of the logic among the measures for diagnostic coverage: simple temporal monitoring (a timer as watchdog) counts low, temporal and logical monitoring of the program sequence counts higher. Claim only the measure that is actually implemented, and note that the watchdog must act independently of the monitored processor (own time base, own switch-off path).
- Security analysis: An attacker who overloads the controller or the network can trigger the watchdog on purpose – a denial of service that stops the machine. That is safe but not available; describe it in the security risk assessment (IEC 62443 FR 7) and in the restart procedure.
Naming this term in a document never shows conformity; the risk assessment, the validation of the safety functions and the security risk assessment do. EN 50742 is a CENELEC draft as of the review date: it is neither law nor a source of presumption of conformity.