Troubleshooting S7-1500H PROFINET Redundancy Loss: Send Clock Fix

1. Problem Overview

When an SIMATIC S7-1500H redundant controller (CPU 1518H-4 PN or CPU 1517H-3 PN) is configured with PROFINET IO devices inside a PROFINET ring, the station may intermittently report Loss of Redundancy even though both H-CPUs and both redundancy connections are physically intact. The error is most often observed as:

• Diagnostic buffer entries of type PROFINET IO: Redundancy loss on the IO controller (H-CPU).
• Flapping S2/R1 partner port status LEDs on the affected IO device.
• Temporary station failure of one or more PROFINET IO devices on the ring.
• Re-establishment of the AR (Application Relationship) every few minutes, even though the physical link and fiber/copper paths show no errors.

The root cause is rarely the cable; it is a mismatch between the Send Clock of the PROFINET IO system and the Update Time / Watchdog assigned to the IO devices. In a default TIA Portal configuration, TIA attempts to run the ring at a 1 ms send clock, which is faster than the redundancy update path inside the H-system can reliably service. The H-controller drops the redundancy AR, the ring partner times out, and the field device briefly disappears from the IO controller.

2. Symptoms and Error Indicators

Before changing parameters, confirm that the symptoms are consistent with a send-clock / update-time mismatch rather than a physical layer fault. Typical indicators are:

If the cable is genuinely broken, the PROFINET ring manager (typically the H-CPU on port 1/2) will report a topology change and the diagnosis will be a hard Link down. A send-clock mismatch produces transient failures with the link staying up — the key discriminator.

3. Root Cause: Send Clock and Update Time Mismatch

The PROFINET real-time channel is driven by a base Send Clock (also called the SendCycleTime) that defines the fundamental transmission period. Every IO device in the system is assigned an Update Time, expressed as an integer multiple of the send clock:

Update Time [ms] = Watchdog Factor × Send Clock [ms]

The Watchdog Time on the IO device must be larger than the worst-case round-trip time of the IO controller, which includes the H-redundancy re-send path. On an S7-1500H, every PROFINET frame on the ring crosses the H-backplane twice (primary CPU → secondary CPU → ring), adding a deterministic latency overhead of typically 1.5–3.0 ms relative to a single CPU 1500.

When the user selects the default 1.000 ms send clock and lets TIA compute the Update Time in Automatic mode, TIA frequently picks a 1× or 2× multiplier. With watchdog factors of 3 (the minimum permitted) the resulting watchdog is 3–6 ms, which is shorter than the H-system's worst-case redundancy latency. The IO controller therefore declares an AR (Application Relationship) loss before the redundancy path has a chance to respond, and the IO device enters station failure.

4. Affected Hardware and Firmware

The phenomenon is documented for the following configuration range. Confirm your hardware matches before applying the parameter changes.

Verify the exact firmware on the SIMATIC S7-1500H using Online & Diagnostics → Diagnostics → CPU Information and on the IO devices with the device's web server (https://<ip>/index.html) or TIA Online & Diagnostics → PROFINET diagnostics.

5. PROFINET Redundancy Architecture for S7-1500H

The S7-1500H supports two PROFINET redundancy modes defined by the PROFINET specification:

For the S7-1500H with PROFINET ring (MRP / MRPD), the controller typically uses S2 to maintain a logical AR with each IO device even if one ring segment is interrupted. When the send-clock / update-time mismatch occurs, the S2 partner port flips rapidly between primary and backup, producing the symptom of loss of redundancy without physical link loss.

6. S7-1500H PROFINET Ring Topology

The reference topology is a single PROFINET ring managed by the two H-CPUs (one ring manager, one ring client) plus a shared device with two ports connected to the ring. The SVG below shows the typical arrangement:

Both H-CPUs participate in the ring with port P1 and P2 of the PROFINET interface. The ET 200SP/MP IM closes the ring on port 1 and port 2. With MRP enabled, the ring manager blocks one port logically to prevent loops; when a single segment fails, MRP re-routes in < 200 ms.

7. Solution Procedure

The fix is performed in TIA Portal (V16 or newer) on the offline project of the S7-1500H station, then downloaded to both H-CPUs.

1. Open the device configuration of the H-station (Devices & Networks → S7-1500H CPU → Device view).
2. Select the PROFINET interface (X1 or X2) of the H-CPU and open Properties → PROFINET interface → Advanced options → Real time settings → PROFINET IO system.
3. Change Send Clock from 1.000 ms to 4.000 ms. The other values to consider are 0.500 ms (not recommended on H), 2.000 ms (still risky with many IO devices), and 4.000 ms (recommended).
4. Click OK to apply the send clock change. TIA will warn that the update times of the IO devices will be re-calculated. Accept the warning.
5. Select each IO device in the network view and open Properties → PROFINET interface → Advanced options → Real time settings → Update time.
6. Set Mode from Automatic to Fixed update time.
7. Set Update Time to 16.000 ms. This corresponds to a 4× multiplier on the 4 ms send clock. Acceptable values in this configuration are 8 ms (2×), 12 ms (3×), and 16 ms (4×); avoid 4 ms (1×) on H-systems.
8. Repeat steps 5–7 for every IO device in the ring. Use Ctrl+A in the network view to multi-select, then apply the change in bulk if the firmware is uniform.
9. Verify the Watchdog Time automatically derived by TIA: it should be at least 48.000 ms (3× the 16 ms update time, the minimum allowed by PROFINET spec).
10. Compile the project (Hardware) and download the configuration to both H-CPUs in RUN with the redundancy function Download to target device → Consistent download.
11. After the download, perform a redundancy test: force the primary CPU to STOP or pull the primary fiber. The IO devices should remain reachable via the backup CPU with no diagnostic interruption.

8. Parameter Reference Table

The following table summarizes the recommended parameter set for the S7-1500H PROFINET ring. Use these values as a starting baseline for H-systems with 5–64 IO devices.

The relationship between these parameters is governed by the PROFINET specification. The minimum allowed watchdog is three update cycles. For an H-system with a 4 ms send clock and a 16 ms update time, the resulting 48 ms watchdog is the smallest value that still gives the H-redundancy re-send path time to recover from a single CPU failover.

9. Verification Steps

After applying the fix, run the following verification sequence. Each check should be logged in the commissioning report.

1. No spurious diagnostic events — leave the system in RUN for at least 24 hours and verify the H-CPU diagnostic buffer shows no Loss of redundancy entries.
2. Live cycle time — in TIA Portal, open Online & Diagnostics → PROFINET diagnostics → Cycle. The actual bus cycle should be steady at ~4 ms with jitter < 1 ms.
3. Forced failover — in the SIMATIC Automation Tool or via the H-CPU display, force the primary CPU to STOP. The backup CPU takes over. The IO devices should report no station failure and process data should resume within 100 ms.
4. Cable break test — physically disconnect one fiber segment of the ring. MRP should reconfigure the ring in < 200 ms; the IO devices should not generate an AR loss.
5. Diagnostic OBs — in the user program, evaluate OB83 (pull/plug) and OB86 (rack/IO failure) call counts. They should be zero in steady state.
6. Web server check — on the IO device web page, the redundancy status should remain Primary or Backup consistently without oscillation.

10. Extended Troubleshooting Matrix

Use this matrix when the symptoms persist after the parameter change. Each row identifies a likely cause and the corresponding diagnostic step.

11. Best Practices for Stable PROFINET Redundancy on S7-1500H

• Always start from the Send Clock and work down to the Update Time. Re-compile the project after each change so TIA re-derives the watchdogs.
• Use Fixed update time on every IO device. Automatic mode assumes a single-controller path and underestimates the H-redundancy latency.
• Keep the IO device count per AR reasonable. An S7-1500H comfortably supports 64–128 PROFINET IO devices per PN interface, but heavy I/O density (e.g. many ET 200SP with high-channel-density modules) benefits from a 4 ms send clock.
• Disable MRPD unless you specifically need parallel redundancy. The default S2 over MRP is the most robust topology for S7-1500H rings.
• Place the MRP manager on the H-CPU, not on an external SCALANCE. This ensures the ring state is consistent with the H-controller state.
• Document the chosen values in the network diagram. Future service engineers must not revert to the 1 ms default without understanding the H-redundancy timing budget.
• Use S2 for the ring and R1 only for true dual-IF devices. Mixing S1, S2, and R1 devices in the same ring is supported but increases the diagnostics complexity.
• Validate the configuration with the TIA Topology Editor. The Topology Editor in TIA Portal can verify the planned ports and avoid miswiring that masquerades as a redundancy-loss event.
• Update the H-CPU firmware to V2.9 or newer. Earlier firmware versions have longer redundancy re-send paths and may require a 8 ms send clock for stability.
• Keep the SIMATIC Automation Tool handy for live diagnostic capture during commissioning. It exports PROFINET cycle time history and AR status logs.

12. Additional S7-1500H Redundancy Considerations

Beyond the send-clock and update-time parameters, the S7-1500H redundancy behavior interacts with PROFINET in two specific ways documented in the official S7-1500R/H system manual:

• Failure of both redundancy connections and the PROFINET cable in the ring — the H-system maintains the AR with the IO device through the H-backplane, but the IO device itself is no longer reachable on PROFINET. The H-CPU logs a station-failure event. This is the worst-case scenario described in the Siemens S7-1500H redundancy scenarios manual.
• Redundant PROFINET R1 configuration — when an IO device uses two interface modules, each establishes one application relationship with one H-CPU. This is described in detail in the PROFINET in SIMATIC PCS 7 – Guidelines and Blueprints document.

For general PROFINET specification details, refer to the PROFINET leading communication system document published by PI (PROFIBUS & PROFINET International). The PROFINET specification (IEC 61784-2, IEC 61158) is the normative reference for the Send Clock, Update Time, and Watchdog semantics used throughout this article.

13. Sample SCL Code for H-System Redundancy Diagnostics

Use the following Structured Control Language (SCL) snippet in the H-CPU user program to capture redundancy-loss events and forward them to the operator panel. Place the code in OB1 or in a cyclic interrupt OB (e.g. OB30).

// SCL: Capture PROFINET redundancy loss events on S7-1500H
// Variables in the H-station's global DB
VAR
    RedLossCounter   : DINT;   // Increments on every loss of redundancy event
    LastRedLossTime  : DTL;    // Timestamp of last event
    RedLossActive    : BOOL;   // TRUE while a redundancy loss is latched
END_VAR

// Inside OB1 / OB30:
IF "IO_device_1".PROFINET_IO_Device.Status.Redundancy_Loss = TRUE THEN
    RedLossCounter := RedLossCounter + 1;
    LastRedLossTime := RD_SYS_T;
    RedLossActive := TRUE;
ELSE
    RedLossActive := FALSE;
END_IF;

// Reset on operator command
IF "HMI".ResetRedLoss THEN
    RedLossCounter := 0;
END_IF;

The exact tag path for Redundancy_Loss is generated by TIA from the IO device's diagnostic model. In the default configuration the tag is <DeviceName>.PROFINET_IO_Device.Status.Redundancy_Loss.