WinCC V7 Redundant S7-300 Communication: CP343-1 Lean Setup

Problem Overview

When a redundant S7-300 station built around the CPU 315-2DP is monitored by a WinCC V7.0 SCADA station over Ethernet, a class of faults appears at the moment the master/reserve role flips. Typical symptoms include:

• The SCADA connection remains registered on the previous master even after the standby CPU has been promoted, producing stale tag values.
• Forced outputs on Station A cannot be released from Station B once Station B becomes the active master.
• The HMI shows the wrong CPU as the live partner, even though the underlying S7 connection terminates correctly at the TCP layer.
• Operator commands to release FORCE jobs are acknowledged but never reach the new master because the WinCC connection state machine has not been refreshed.

The root cause is not in the CP 343-1 Lean hardware and not in the S7-300 software redundancy library itself. The root cause is in the WinCC V7.0 redundancy glue: the redundant connection is created, but the C script that maps the S7 status word (SWR.Standby) to the WinCC @CPU_x@ForceConnectionState tag is missing or mis-wired. Without that script, WinCC keeps the original TCP/ISO-on-TCP session active against the now-standby CPU and ignores the standby bit from FB101 "SWR_ZYK".

Root Cause Analysis

Three layers must agree on which CPU is currently master:

1. Layer 1 – S7-300 software redundancy: FB101 "SWR_ZYK" runs in a cyclic interrupt OB (typically OB35) and writes the status of each CPU into its instance DB. Bit DBX9.1 of the instance DB indicates the Reserve (standby) state.
2. Layer 2 – WinCC connection driver: WinCC opens two S7 connections, one to each CPU. The connection state is exposed through the internal tag @CPU_x@ConnectionState. The driver accepts an external override through @CPU_x@ForceConnectionState.
3. Layer 3 – WinCC script glue: A C action monitors the standby bit from Layer 1 and toggles the override from Layer 2 so that WinCC terminates the session against the failed master and starts using the new master.

If Layer 3 is missing, Layer 1 correctly reports "Station B is now master," but Layer 2 keeps the original TCP socket bound to Station A. The FORCE/UNFORCE request is sent on the wrong connection, the request either times out or is rejected by the standby CPU, and the operator sees the symptom described in the problem statement.

S7-300 Software Redundancy Architecture

The standard library block set used in this topology is the "Software Redundancy" library delivered with STEP 7 V5.x. The relevant blocks are:

The instance DB of FB101 "SWR_ZYK" exposes the redundancy state of the CPU. The most relevant offsets for WinCC integration are:

WinCC cannot read these bits directly through its standard S7 driver unless they are mapped to a tag. The recommended approach is to expose DBX9.1 through a global bit tag named SWR.Standby in the WinCC tag manager. The exact name is user-defined; what matters is that the bit tag in WinCC is wired to DBX9.1 of the instance DB of FB101 "SWR_ZYK".

CP 343-1 Lean Communication Processor Role

The CP 343-1 Lean (catalog family 6GK7 343-1CX10-0XE0) is the entry-level Ethernet CP for the S7-300 family. It provides a 2-port real-time switch on the CPU side and exposes the S7 CPU to the Ethernet network through the following services:

• S7 communication (PG/OP, S7 functions)
• ISO-on-TCP (RFC1006) – the transport used by default for WinCC S7 connections
• TCP and UDP via the SEND/RECEIVE interface
• Limited PROFINET IO controller capability (not used in this redundancy scenario)

Limitations relevant to a WinCC redundant topology:

Because the CP has no native awareness of S7 software redundancy, the CP cannot decide which CPU WinCC should talk to. That decision must be made by WinCC based on the standby bit from FB101.

WinCC V7.0 Redundant Connection Configuration

Two S7 connections must be created in the WinCC project: one pointing to the master (CPU A), one pointing to the reserve (CPU B). The CP that carries the WinCC traffic on each CPU side is the CP 343-1 Lean. Both CPUs must reside in the same IP subnet so that WinCC can reach either partner without re-routing.

Recommended IP layout

The "Main connection" parameter in the WinCC connection dialog accepts the address of CPU A (192.168.0.1), and the "Reserve connection" parameter accepts the address of CPU B (192.168.0.2). Both fields are independent; do not put the same address in both.

Connection dialog parameters (WinCC V7.0)

Mapping the Standby Bit into WinCC

Create a binary WinCC tag named SWR.Standby with the following properties:

The "Reserve / Standby" indicator is at DBX9.1 of the instance DB of FB101 "SWR_ZYK". If you renamed the FB call or assigned a different instance DB number, adjust the DB number accordingly. The bit semantics are: TRUE on the CPU that is currently the reserve, FALSE on the CPU that is currently the master.

C Script for Master-Reserve Changeover

The script below is the canonical solution referenced in the Siemens application documentation for S7-300 software redundancy. It must be scheduled on a WinCC global action with a cycle of 1 second.

// WinCC global C action, cycle 1 s
// Adjust the connection name to match your project.

if (GetTagDWordWait("@CPU_3@ConnectionState") == 0)
{
    // No live S7 connection to the main CPU.
    // Force WinCC to switch to the reserve connection.
    SetTagDWord("@CPU_3@ForceConnectionState", 1);
}
else
{
    // Main CPU is reachable. If it has become the reserve, release the override.
    if (GetTagBitWait("SWR.Standby") == TRUE)
    {
        SetTagDWord("@CPU_3@ForceConnectionState", 0);
    }
}

Mechanics of the script

• @CPU_3@ConnectionState is an internal DWord WinCC tag. 0 means the S7 connection to the partner currently marked as "Main" is down or not established.
• @CPU_3@ForceConnectionState is an internal DWord WinCC tag. 1 tells WinCC to ignore the failed main connection and use the reserve connection instead.
• SWR.Standby is the user-defined binary tag wired to DBX9.1 of the instance DB of FB101 "SWR_ZYK". TRUE indicates that the CPU we are currently talking to is the standby.

The combined logic guarantees that:

1. If the main CPU dies, the S7 connection collapses to ConnectionState == 0, the override flag is set to 1, and WinCC switches to the reserve CPU.
2. If the main CPU comes back and is now the standby (because the other CPU was promoted), the override flag is reset to 0 and WinCC returns to the original main once it recovers.

Where to place the script

Open the WinCC Explorer, right-click "Global Scripts" → "C-Actions", create a new action of type "Project function" or "Global action" depending on the preferred reuse scope, paste the code above, and set the trigger to "500 ms" or "1 s". The action must be enabled; verify that the small green traffic-light icon shows green at runtime in WinCC Explorer.

Dynamic Wizard: Create Redundant Connection

If the WinCC side has not been wired through the Dynamic Wizard, perform the following sequence:

1. In WinCC Explorer, start the Dynamic Wizard "Create redundant connection".
2. Select the existing main connection (for example CPU_A_Main).
3. Define a new partner connection pointing to the reserve CPU (for example CPU_B_Reserve).
4. Confirm that both connections are now flagged as "redundant pair" in the WinCC Explorer connection list.
5. Insert the C action above into the global action list.

If the wizard is not run, the two connections exist as independent logical units and WinCC will never automatically hand over traffic between them. The wizard is what tells the WinCC runtime that the two connections are a pair.

Network and IP Addressing Rules

The original symptom "force not working after switchover" can also be triggered by a network misconfiguration. Validate the following before commissioning the script:

• Both CP 343-1 Lean modules must be on the same IP subnet. If you used 192.168.0.1 and 192.168.9.2 with a typo (192.168.9 instead of 192.168.0), WinCC will silently fail to connect to one of the partners and the redundancy state will be undefined.
• The WinCC station must be on the same subnet. A router between WinCC and the S7-300 station is supported, but the script assumes that any lost TCP/ISO-on-TCP session surfaces as ConnectionState == 0 within a few seconds; long paths through routers can delay that detection.
• The CP 343-1 Lean must allow OP connections on the configured TSAP. By default the CP opens TSAP 03.01 on the S7-300 side. If you changed the TSAP, the WinCC connection parameters must match.
• If you use a managed switch, disable IGMP snooping or pin the multicast used by S7 redundancy diagnostics to a known port, otherwise the standby detection can be delayed by 30 s or more after a fail-over.

Force / Unforce Output Behavior Across Switchover

The S7 CPU retains force jobs in its retentive memory and re-applies them after a restart. When the master role flips from CPU A to CPU B, the forced outputs on the I/O side are no longer driven by CPU A. CPU B does not inherit the force table from CPU A automatically.

For a force/unforce operation to succeed after a switchover, the WinCC client must issue the command on the active master. The C script in the previous section guarantees that the WinCC connection state tracks the S7 redundancy state, so the operator client always issues the unforce on the right partner.

Step-by-Step Commissioning Procedure

1. Compile and download the STEP 7 project with FB101 "SWR_ZYK" in OB35 and FB100 "SWR_START" in OB100. Verify that the instance DB of FB101 is generated and that the DB number is recorded.
2. Open the WinCC project, open "Tag Management", add the binary tag SWR.Standby pointing at DBX9.1 of the recorded DB number on the main connection.
3. Open "Tag Management" again and verify that the internal tags @CPU_3@ConnectionState and @CPU_3@ForceConnectionState exist. These are added automatically by the Dynamic Wizard "Create redundant connection". If absent, run the wizard.
4. Create the C action shown in the previous section, paste the script, schedule it on a 1 s trigger.
5. Download the WinCC project and start Runtime.
6. From the WinCC station, ping both CP 343-1 Lean addresses. Confirm 192.168.0.1 and 192.168.0.2 both respond.
7. Open the online diagnostics of the S7 connection in WinCC. Verify that the main connection shows "Established" and that the standby CPU shows "Standby" in the S7 redundancy status.
8. Force an output from WinCC on the active master. Confirm that the output is forced.
9. Stop the active master CPU (power off or stop the CPU). Wait 3 s.
10. Verify that WinCC switches the active connection to the reserve CPU and that the previously forced output is no longer driven to the forced value.
11. Release the force from WinCC on the new master. Confirm that the output returns to the program value.

Verification Checklist

Troubleshooting Matrix

Performance and Timing Considerations

The 1 s cycle of the C action is a reasonable default. Shorter cycles (250 ms, 500 ms) reduce the visible switchover delay but increase the load on the WinCC dispatcher. Longer cycles (2 s, 5 s) reduce load but extend the window during which the operator sees stale values after a fail-over.

Migration to WinCC 7.4

WinCC 7.4 inherits the redundancy architecture from WinCC 7.0 with two notable differences:

• The internal tag namespace uses @ConnectionState and @ForceConnectionState prefixed with the connection name as in WinCC 7.0, but the C action may require updated signatures if you migrate an existing project.
• The short-term user archive is configured under "Tag Logging" → "Archives" → "New archive" with the type "Short-term"; the duration is set in the archive properties. This is the location referenced in the original query.

The redundancy C action shown above works unchanged in WinCC 7.4 against CPU 315-2DP stations with CP 343-1 Lean, provided the same Dynamic Wizard has been run.

Documentation References

For deeper coverage, refer to the official Siemens documentation:

• SIMATIC WinCC V7.0 Communication Manual – section "Configuring redundant S7 connections"
• SIMATIC S7-300 Software Redundancy – application description for FB100/FB101 and the partner-handshake blocks
• SIMATIC NET CP 343-1 Lean – Operating Instructions, configuration of S7 connections and ISO-on-TCP TSAPs
• SIMATIC Programming with STEP 7 V5.x – description of OB35 cyclic interrupt and instance DB offsets

All of the above are available through the Siemens Industry Online Support portal at support.industry.siemens.com under the SIMATIC product family.