S7-400H PROFINET Ring Topology: Resolving Dual-Ring Switch Conflicts

The S7-400H fault-tolerant PLC supports a PROFINET system redundancy scheme in which each CPU is equipped with a single PROFINET interface that integrates an internal 2-port switch. This 2-port switch lets the H-station itself act as one ring node, which dramatically reduces external switch count but also imposes a strict rule: any given PROFINET switch (including the integrated 2-port switch of an S7-400H CPU) can participate in only one HRP (High-availability Redundancy Protocol) or MRP (Media Redundancy Protocol) ring at a time. A common engineering mistake is to draw the H-station in two parallel rings — a configuration that "short-circuits" the redundancy manager's logical view of the ring and prevents the network from reaching redundancy.

This reference explains the root cause of the dual-ring conflict, the architecture of the integrated 2-port switch on the S7-400H PROFINET interface, and the correct topology alternatives for redundant plant networks.

1. Problem Summary

Symptom: a STEP 7 / TIA Portal network configuration shows two colored rings (Ring 1 and Ring 2). Ring 1 is built from two external SCALANCE switches and is correctly closed. Ring 2 is also closed, but it is looped through a switch that already belongs to Ring 1. During commissioning, the redundancy manager does not transition the ring ports into the "forwarding" state, the diagnostic buffer reports an MRP/HRP topology error, and PROFINET devices downstream of the affected switch stay in cyclic-data exchange but the redundancy manager logs MRP: Ring open / HRP: Duplicate ring events.

Root cause: a single PROFINET switch (whether an external SCALANCE or the integrated 2-port switch of an S7-400H CPU 410-5H / 412-5H / 414-5H / 416-5H) is a member of two closed ring topologies simultaneously. PROFINET redundancy protocols allow one ring membership per switch — never two.

2. S7-400H PROFINET Interface Architecture

The S7-400H CPUs use one of the two PROFINET interface variants described in the SIMATIC S7-400H Fault-Tolerant Systems manual:

Each H-CPU exposes a single RJ45 PROFINET interface (X1) that is internally wired as a 2-port managed switch. The two physical ports are labeled X1P1 and X1P2 (or P1R/P2R on the CPU 410-5H). Internally, the switch fabric forwards frames between these two ports plus the CPU's own protocol stack.

Why a 2-port switch on the CPU? In a fault-tolerant H-station, the two CPUs synchronize via fiber-optic sync modules (F0..F4) and exchange data over the backplane. The PROFINET side of each CPU must still connect to the field — but a redundant ring requires a switch with two free ring ports. By integrating a 2-port switch into the CPU, Siemens lets the S7-400H close a ring with a single cable per neighbor: X1P1 connects to the previous ring node, X1P2 connects to the next ring node. No external switch is required at the controller rack to make the H-station a ring participant.

2.1 Port role assignment

For HRP/MRP ring operation, the two integrated ports must be assigned in STEP 7 (V5.5) or TIA Portal (V15 and later) as ring ports:

• Port X1P1 (P1R) — Ring port 1
• Port X1P2 (P2R) — Ring port 2

STEP 7 / TIA Portal exposes these as Ring port 1 / Ring port 2 properties under PROFINET interface → Port → Ring. When the ports are configured as ring ports, the integrated switch forwards PROFINET frames like any ring member and cooperates with the redundancy manager (RM).

3. HRP and MRP Ring Redundancy Fundamentals

Two ring-redundancy protocols are relevant on the S7-400H PROFINET side:

Both protocols share the same physical rule: every switch in the ring has exactly one ring port 1 and one ring port 2. The ring is logically a single closed loop in which the RM blocks one of its two ring ports under normal conditions, then unblocks it on link loss so traffic re-routes through the remaining path. The PROFINET ring topology is described in detail in the Siemens ring topology planning manual (A6V11940231).

3.1 Why a switch cannot be in two rings

An MRP client or HRP client has only two ring ports. If you attempt to close two independent rings through the same physical switch, you have three ports participating in ring logic — but only two are defined as ring ports. The "extra" port either:

1. Is left as a normal port. It forwards frames normally, which means it closes a third, hidden loop. The RM sees test frames on both ring ports coming from a topology that does not match the configured ring, and refuses to bring the ring into redundancy.
2. Is also configured as a ring port. The switch now has three ring ports, but the protocol only handles pairs. Test frames on ring port 3 are interpreted as foreign ring traffic, and the redundancy manager logs MRP: Multiple managers or remains permanently in RM: passive.

The fundamental constraint is mathematical: a ring is a single closed loop with a 1-of-N blocked port. Two rings share at most one node, but only by means of a special protocol extension (e.g., MRP interconnection / MRPD — see IEC 62439-3) that introduces explicit coupling nodes. The S7-400H integrated 2-port switch and standard SCALANCE switches do not implement MRPD coupling, so the only correct way to build two independent rings is to keep them physically disjoint.

Schneider Electric's industrial network blog describes the same rule from a general ring-redundancy perspective: "Redundancy puts backbone into the ring topology" — a ring lends itself to redundancy, but the redundancy controller must see a single logical loop, not two.

4. Root Cause: Single-Ring Membership Per Switch

The configuration described in the source problem shows:

• Ring 1 (green): two SCALANCE switches and the S7-400H CPU's integrated 2-port switch form a closed ring. This is valid.
• Ring 2 (red): the same two SCALANCE switches and the S7-400H CPU are wired into a second closed ring that includes an additional switch at the bottom of Ring 1.

The bottom switch of Ring 1 is now a member of both rings. From the perspective of HRP/MRP, Ring 1 + Ring 2 form a single closed loop with one extra branch. When the RM sends its test frames, they traverse the branch and return on a non-ring port. The RM never sees a clean loop on its blocked port, so it never transitions the ring to redundancy.

5. Network Topology Analysis

The following SVG diagram shows the failing configuration on the left and the corrected configuration on the right.

The diagram on the left shows the typical failure pattern: a switch sits on the boundary between two closed loops. The diagram on the right shows the typical fix: a single ring carrying the S7-400H and its IO devices, with all redundant IO off one switch port pair.

6. Correct S7-400H Ring Configurations

There are three topologies that are correct for an S7-400H with a PROFINET ring.

6.1 Topology A: Single ring, S7-400H as one ring node

SCALANCE switches SW1..SWn are arranged in a ring. The S7-400H CPU's PROFINET ports X1P1 and X1P2 are configured as ring ports and close the ring through the CPU. The redundancy manager is typically a SCALANCE switch that supports the RM role (e.g., SCALANCE XC-200, SCALANCE XR-500, SCALANCE XC-300 with firmware V2.x and above). Maximum ring size: 50 nodes including the H-CPU.

6.2 Topology B: Redundant ring with HSR / MRP coupling

If the field requires two physically independent rings, the rings must be coupled through an MRP interconnection (MRPD) or PRP/HSR gateway. The S7-400H CPU's integrated 2-port switch does not implement MRPD; therefore the only way to get a second independent ring is to use a SCALANCE switch in Redundant Ring Port (RRP) mode with two ring ports and a third inter-ring port, configured as an MRP manager/client pair. The two rings never share a switch — they share a single dedicated coupling device.

6.3 Topology C: Linear PROFINET with RSTP

If ring redundancy is not required, a linear PROFINET with RSTP (Rapid Spanning Tree Protocol, IEEE 802.1w) is acceptable. RSTP also enforces a tree topology, so the S7-400H 2-port switch can be in the tree but no loop is allowed. Reconfiguration is 1–10 s for RSTP, which is generally too slow for PROFINET IO — prefer HRP/MRP for the S7-400H side.

7. Configuration in STEP 7 and TIA Portal

7.1 STEP 7 V5.5 (classic)

1. Open HW Config and select the S7-400H CPU.
2. Open the PROFINET interface X1 properties and select Port X1P1.
3. Set Port type to Ring port 1.
4. Repeat for Port X1P2 → Ring port 2.
5. Open Edit → PROFINET IO → Redundancy → MRP Domain and create a new MRP domain.
6. Assign each ring participant to the domain. The S7-400H CPU is added with Role = Client.
7. Assign the RM role to a single SCALANCE switch (must be a switch that supports MRP manager in its firmware).
8. Compile and download. The ring is closed automatically once the wiring matches.

7.2 TIA Portal V15 and later

1. In the device configuration, open Properties → PROFINET interface [X1] → Ports.
2. For X1P1 and X1P2, select Ring in the Connection column.
3. Open Topology view in the network editor.
4. Drag the ring ports together to close the ring. The TIA Portal will warn if a port is wired to two ring ports simultaneously.
5. Open Properties → MRP domain on the RM-capable switch. Add all ring members.
6. Compile the project and download to all devices.

8. Verification and Diagnostics

After commissioning, perform the following checks:

1. Ring status in WBM. On the SCALANCE RM switch, open Information → Redundancy → MRP. The status should read MRP: Ring closed, RM active and the blocked port is identified.
2. Test-frame rate. Wireshark on any ring port should show MRP_Test frames every 50 ms with the RM's MAC as source and the multicast destination 01:15:4E:00:00:01. Frame interval drift indicates a node that is not properly participating in the ring.
3. Forced link-down test. Disconnect a single ring cable. The RM should reconfigure within 200 ms and the WBM should log Topology change. PROFINET IO devices must not lose cyclic data exchange.
4. STEP 7 diagnostics. In the online view of the H-CPU, the diagnostic buffer should show no PROFINET IO: Bus fault entries after a forced reconfiguration.
5. H-system redundancy test. Trigger an H-CPU failover (HBY switch on the backplane or via the S7-400H system function SSS_RTM). The PROFINET ring must remain closed; the new master CPU must take over the ring within the S7-400H failover time (typically < 100 ms for the bus, plus the reconfiguration time of the ring).

9. Troubleshooting Matrix

10. Field-Commissioning Notes

• One RM per ring. Siemens strongly recommends one manager and N-1 clients per ring. Multiple managers are allowed for fast-reconfiguration fallback, but only one is active at a time.
• H-station failover time. The S7-400H CPU failover (HBY) does not interrupt PROFINET ring reconfiguration, because the new master takes over the same PROFINET MAC on the same physical ports within the H-system's <100 ms failover time. The ring sees no MAC change.
• Layer 3 boundaries. A PROFINET ring must not cross a routed boundary. Place the IRT/RT-aware switches inside the same subnet and the router (e.g., SCALANCE S615) outside the ring.
• Documentation references. Always cross-check the S7-400H manual sections on PROFINET ring topology and the SCALANCE manual sections on HRP/MRP manager role before assigning roles. The S7-400H manual explicitly states that the integrated 2-port switch is to be used as ring ports when ring topology is required.
• Firmware pairing. The CPU 410-5H process automation firmware releases track MRP enhancements. For example, firmware V8.2 added support for MRP client with up to 50 ring nodes; later versions added MRP diagnostics. Always run the H-system on a firmware release that matches the S7-400H system manual entry for PROFINET ring.

11. FAQ