Profinet Update Time / Bandwidth
Calculate network load and required update times for Profinet IO devices.
Network Parameters
Network Load Estimate
Introduction to PROFINET Network Mechanics
For industrial automation engineers, ensuring reliable deterministic communication is non-negotiable. PROFINET IO provides the determinism required for real-time control, but this performance is heavily dependent on balancing your network’s bandwidth. Failing to optimize your PROFINET bandwidth and device update times can lead to network congestion, dropped packets, and ultimately, unplanned machine downtime.
Understanding the PROFINET Update Time
The update time is the precise interval at which a PROFINET IO Controller (such as a PLC) and an IO Device exchange cyclic process data.
Send Clock vs. Update Time
Engineers must understand the relationship between the Send Clock and the Update Time. The Send Clock is the base timing established by the IO Controller (commonly 1 ms). The Update Time is calculated using a Reduction Ratio:
Update Time = Send Clock × Reduction Ratio
A common mistake is setting the fastest possible update time for all devices. While a high-speed motion controller might strictly need a 1 ms update, a simple temperature sensor updating every 1 ms is a massive waste of PROFINET bandwidth.
The “Only as Necessary” Rule
Follow the engineering best practice: Update only as necessary, not as much as possible. Optimizing the reduction ratios allows slower devices to update less frequently, drastically reducing unnecessary cyclic traffic on the network.
Calculating PROFINET Network Load
Network load represents the percentage of bandwidth consumed by PROFINET cyclic traffic relative to your total available bandwidth (typically 100 Mbps in standard PROFINET installations).
Network Load Thresholds
- < 20% Load: Optimal state. No action required.
- 20% – 50% Load: Warning zone. Review your network plan, topology, and update times.
- > 50% Load: Critical state. Immediate action is required to segment the network or increase device update times to prevent packet loss.
Factors Influencing Network Load
When calculating the load, consider these critical elements:
- Frame Overhead: Every PROFINET packet carries overhead (preamble, MAC addresses, VLAN tags, etc.). At very high update frequencies, this overhead can consume more bandwidth than the actual payload.
- Line Topology Limits: In a line topology, every passing device (or integrated switch) adds latency and increases telegram traffic. Extended lines can quickly experience high jitter and load bottlenecks.
- Non-PROFINET Traffic: Industrial networks are rarely isolated. Standard TCP/IP traffic, HMI communications, IP cameras, and engineering station access share the same physical infrastructure and contribute to the total network load.
Best Practices for Bandwidth Calculation and Optimization
Leveraging Engineering Tools
Manual calculation of network load is highly prone to errors due to complex frame overheads and topology rules. Instead, engineers should utilize specialized tools:
- Hardware Configuration Tools: Modern IDEs like Siemens TIA Portal provide built-in network load indicators during the hardware configuration phase.
- PI Load Calculator: The PROFIBUS & PROFINET International (PI) organization offers calculators to estimate load based on device count and data types.
- Simulation Software: Tools like SINETPLAN can simulate network behavior before physical deployment, identifying bottlenecks proactively.
Optimizing Watchdog Times
The watchdog time determines when a device is considered “lost” if no data is received. As a standard engineering rule of thumb, set the watchdog time to 3 times the update time. This provides a robust tolerance against minor network jitter without sacrificing safety.
Conclusion
Calculating and managing your PROFINET bandwidth is fundamental to industrial network design. By strategically assigning update times based on device requirements rather than defaults, and utilizing proper calculation tools, engineers can ensure a robust, high-performance, and resilient automation system.