C135 PCS 7 (PCS7)
6C
6 Cielo Ciencia y Tecnología
·Publicada en C135 PCS 7 (PCS7)

How to Diagnose and Resolve Cycle Time Exceeded (OB80) Faults in PCS 7 S7-400H Systems

How to Diagnose and Resolve Cycle Time Exceeded (OB80) Faults in PCS 7 S7-400H Systems

To resolve recurring "Cycle time exceeded" faults (Organization Block OB80 calls) in SIMATIC PCS 7 S7-400H redundant systems, eliminate underlying fieldbus interrupt storms and rebalance program execution across slower cyclic interrupt OBs. When the controller cycle time exceeds the maximum configured scan watchdog (default 6,000 ms) or when cyclic interrupt OBs (e.g., OB35 at 100 ms) are interrupted repeatedly, the CPU calls OB80 to log a time fault. If the cycle time is exceeded twice consecutively without completing the cycle checkpoint, the CPU trips to STOP. Immediate remediation requires checking the diagnostic buffer for diagnostic interrupt storms (OB82/OB86 caused by damaged fiber optics or unstable DP slaves), instantiating the CPU_RT block for automated load shedding, and relocating slow analog control loops from OB35 to OB34 or OB32.

Table of Contents

Mechanics of Cycle Time Monitoring and OB80 in S7-400H

The SIMATIC S7-400H operating system enforces strict watchdog timers to guarantee deterministic process control:
- OB1 Cycle Watchdog: Monitored by a hardware timer configured in HW Config (default 6000 ms). If OB1 execution does not complete before the timer expires, the CPU triggers OB80 (Time Error). If OB80 is missing or if the cycle times out a second time in the same cycle, the controller enters STOP mode.
- Cyclic Interrupt Overflow: Periodic tasks execute in timed organization blocks (OB38: 10 ms, OB35: 100 ms, OB34: 200 ms, OB33: 500 ms, OB32: 1000 ms). If the execution duration of code inside an OB exceeds its configured cycle interval, a nested start attempt occurs, generating a time error.
- H-System Synchronization Overhead: In redundant S7-400H controllers, the primary and standby CPUs continuously exchange synchronization data over optical links. High interrupt rates dramatically increase synchronization overhead, compounding CPU latency.

Root Causes: Bus Disturbance Storms vs. Cyclic Code Overload

A recurring cycle time fault that suddenly manifests on a previously stable plant is almost always caused by an external physical communication fault rather than sudden code expansion:

Failure CategoryUnderlying MechanismObservable Diagnostic SignatureBus Storms (OB82 / OB86)Failing fiber-optic cables, misaligned Optical Link Modules (OLM), or loose PROFIBUS/PROFINET connectors cause devices to rapidly disconnect and reconnect.Thousands of OB86 (Rack Failure) and OB82 (Diagnostic Interrupt) events flooding the diagnostic buffer within seconds, starving cyclic execution.Cyclic Execution OverloadToo many complex technological blocks (APL motors, valves, high-order PID controllers) packed into fast interrupt OB35 (100 ms).Steady CPU utilization above 85%; net execution time of OB35 approaching or exceeding 90 ms.H-Sync Coupling DelaysMemory modifications during runtime downloads or memory-intensive background communication jobs delaying cyclic synchronization check-ins.Synchronous state delays logged in the redundant controller status window.

Method: Diagnostic Buffer Analysis and Event Identification

Before altering program code, identify the precise trigger by inspecting the controller diagnostic buffer:

Step 1: Open the Online Diagnostic Buffer

  1. In SIMATIC Manager, select the AS station and click PLC -> Diagnostic Buffer.
  2. Examine the recent event list for time error events:
    - Event ID 16#4533: Cycle time exceeded (OB80 called because OB1 exceeded maximum cycle monitoring time).
    - Event ID 16#4535: Cyclic interrupt overflow (OB80 called because a timed OB was requested before its previous execution finished).

Step 2: Correlate Interrupt Events

  1. Scroll down immediately below the 16#4533 or 16#4535 event.
  2. Check for preceding burst events:
    - Multiple instances of Event ID 16#38C4 / 16#39C4 (OB86 station failure/return on PROFIBUS or PROFINET).
    - Multiple instances of Event ID 16#3842 / 16#3942 (OB82 module diagnostic interrupt).
  3. If repetitive station disconnects appear, record the DP slave address or PROFINET device name. Inspect the physical network media, test optical power on OLM links, and replace damaged connectors.

Method: Program Rebalancing and CPU_RT Load Shedding

If physical communications are intact but computational load is excessive, rebalance the runtime groups and install automated load management:

Step 1: Relocate Technological Blocks to Slower OBs

In PCS 7, fast execution (OB35, 100 ms) is only required for fast dynamic loops (flow, pressure, motor speed control). Slower thermal processes do not require 100 ms updates:
1. Open CFC and select Edit -> Run Sequence....
2. Locate the runtime groups controlling temperature loops, slow tank levels, and auxiliary monitoring blocks.
3. Move these runtime groups from OB35 (100 ms) to OB34 (200 ms) or OB32 (1000 ms).
4. Result: OB35 execution time drops substantially, providing ample processing margin for cyclic interrupts.

Step 2: Implement the CPU_RT Block for Controlled Load Shedding

The PCS 7 standard library provides the CPU_RT block (FB70) specifically to prevent cycle watchdog CPU crashes:
1. In the Master Data Library or PCS 7 Basic Library, locate block CPU_RT (FB70).
2. Insert an instance of CPU_RT into a dedicated chart and install it in the execution sequence of OB100 (Startup) and OB1 (Cyclic background).
3. Configure CPU_RT parameters:
- Interconnect NET_TIME to monitor net OB execution.
- When CPU load exceeds configurable thresholds (e.g., 85%), CPU_RT automatically sheds non-essential tasks temporarily, preventing the cycle time from exceeding the fatal 6,000 ms threshold.

Verification

To verify that the cycle overload condition is completely resolved:
- In SIMATIC Manager, open PLC -> Module Information -> Performance Data: confirm that the current and maximum scan cycle times remain well below the watchdog limit (recommended maximum < 50% of watchdog).
- Inspect the diagnostic buffer over a 24-hour operational window: confirm zero occurrences of Event ID 16#4533 or 16#4535.
- Monitor optical communication links: verify that Optical Link Modules (OLMs) show solid green CH1/CH2 LEDs with no flickering yellow/red fault indicators.
- In CFC test mode, confirm that all PID control loops running in OB34 or OB32 maintain smooth setpoint tracking without lag.

Limitations and Operational Precautions

  • Never Arbitrarily Increase the Watchdog to Mask Faults: Increasing the Maximum Cycle Time in HW Config from 6,000 ms to 10,000 ms does not solve the root problem; it merely delays the inevitable controller shutdown and increases safety response latency.
  • F-System Safety Execution Constraints: For fail-safe (FH) systems, fail-safe blocks run in a dedicated F-runtime group (typically in OB35). The maximum execution time of safety logic is constrained by the configured Safety F-Monitoring Time. Rebalancing standard code must never compromise the strict execution cycle of the F-program.
  • Full Download Caution: Moving runtime groups across different OBs in CFC can be downloaded in RUN mode via Change Download, provided block types are not deleted.

Engineer FAQ

What happens if OB80 is not loaded in the CPU?

If a cycle time exceeded condition occurs and OB80 is not present in the controller memory, the CPU immediately switches to STOP mode on the very first occurrence.

How do I measure the exact execution time of an individual OB?

You can query the execution duration using the CPU_RT block parameters (OB35_Net, OB1_Net) or by using the SFC64 TIME_TCK system function to timestamp entry and exit points.

Why did changing an OLM and fiber cable resolve a CPU cycle time fault?

When fiber cables or OLMs degrade, high optical attenuation causes packet drops. The PROFIBUS master retries telegrams continuously and fires diagnostic interrupt OBs repeatedly, consuming up to 90% of the CPU's processing time and starving cyclic execution.

Is it safe to move APL motor and valve blocks to OB34 (200 ms)?

Yes. While digital interlocks benefit from fast response, standard industrial pumps, fans, and valves with mechanical transit times of multiple seconds operate with complete stability when executed at 200 ms intervals.

Technical References

  • Siemens Industry Online Support: SIMATIC PCS 7 CPU_RT Block Description and Application Example.
  • SIOS Entry 109742323: PCS 7 Compendium Part A - Configuration Guidelines (Section: CPU Performance and Cycle Times).
  • SIOS Entry 109809264: SIMATIC Process Control System PCS 7 Fault-Tolerant Process Control Systems (S7-400H).
  • Siemens Technical Manual: SIMATIC S7-400 Automation System Module Data (Watchdog Timers and OBs).

Related Articles

Related Tags

  • S7-400H
  • OB80
  • Cycle Time
  • CPU_RT
  • Diagnostics
  • Watchdog
  • Industrial Automation
20 respuestas