Resolving CPU 416-5H Download Error 13:5472 in PCS7 v8.0 SP2

David Krause14 min read
PLC HardwareSiemensTroubleshooting
Licensed PE Working through this on a live machine? A Maine-licensed engineer can take it from here — included with IMD hardware, by the hour for everything else. Book an engineer

1. Problem Overview

When attempting to download hardware configuration changes to a Siemens CPU 416-5H running under SIMATIC PCS 7 V8.0 SP2, the engineering station presents the dialog:

"There are other online connections to this CPU. Close all connections before loading" — Message number 13:5472.

The dialog blocks the download. The message carries no Siemens help-ID cross-reference, no online documentation link, and no specific recommendation, which is what makes it field-frustrating. The 13:xxxx prefix is a Step 7 internal event/error block; 5472 maps to the resource-conflict family of warnings related to active online sessions on the target CPU.

The error blocks any download attempt to a CPU that is currently in RUN, which is exactly when most PCS 7 operators need to apply incremental changes (for example, adding a Simeas P parameterization device to a new PI address). On an H-system, the situation is compounded because there are two redundant CPUs that must each be updated without process interruption.

2. Affected Systems, Firmware, and Software Versions

Component Affected Versions Notes
CPU type 6ES7 416-5HR06-0AB0 (CPU 416-5H) Used in S7-400H redundant configurations
Engineering software SIMATIC PCS 7 V8.0 SP2 (Step 7 V5.5 + CFC/V9) HW Config, NetPro, CFC online
Firmware range V6.x firmware on CPU 416-5H Check via online → accessible nodes
New device being added SIMEAS P digital/analog input station (e.g., 7KG775x) PROFIBUS DP slave on H-system
Plant state at time of fault Both CPUs in RUN, redundant link active No process interruption allowed

The 13:5472 message is independent of the new device type. Whether the engineer is adding a Simeas P, a new ET 200M station, an additional PROFIBUS slave, or a new I/O module under an existing station, the connection-conflict check on the CPU runs the same way and will block the download if any other online session is registered.

3. Root Cause Analysis

The CPU 416-5H allows exactly one online engineering session that has permission to change configuration. If the engineering software (Step 7 / PCS 7) detects that any other application, service, or remote PG/PC holds an active online handle to the same CPU, it refuses to start a download that modifies flash-resident configuration (SDB container, system data blocks).

Common root causes observed in PCS 7 V8.0 SP2 environments:

  1. External data-acquisition service holding an OPC / S7 connection. Tools such as IBA Analyser, OPC Scout, or a WinCC archive server occasionally open a long-lived S7 connection that the engineering station registers as "occupied." This is the most frequently missed cause because the service is invisible from the engineering UI.
  2. A second programming device online simultaneously. Even a single extra PG with HW Config, NetPro, CFC, or just an open Accessible Nodes view in the same project on the same CPU holds the S7 connection and triggers 13:5472.
  3. An open NetPro or HW Config on the same project. Because PCS 7's component view is split across multiple editors, opening NetPro while HW Config is foreground can produce a hidden handle from the same ES.
  4. Stale connection on the standby CPU. The H-system duplicates the online state; if the active CPU's S7 resources are busy, the standby's matching connection table may also show "established," and Step 7 refuses to commit the change.
  5. A CFC online session left in monitor mode. CFC test-mode (Watch On) does not itself block, but CFC Test → Commissioning → Online in the wrong phase can leave a write-protect connection registered.

The CPU does not produce a fault LED, does not enter STOP, and does not write a diagnostic buffer entry for this event — the rejection is purely a PG-side / connection-arbitration refusal, not a CPU fault.

4. Pre-Download Diagnostic Procedure

Before any solo-mode escalation, run the following checks in order. Each check resolves the most common causes without operator intervention.

4.1 Quick Wins (no process impact)

  1. Re-launch the Engineering Station. Reboot the ES, wait for the S7DOS service and the S7ONLINE access point to settle, and retry the download. Many stale connection tables clear at service restart.
  2. Close all editors except HW Config. Open HW Config only, with no NetPro, CFC, SFC, or LAD/FBD editor visible. Each editor leaves a passive online handle to the project's online view.
  3. Confirm no other PG/PC is online to the same CPU. In the project, right-click the CPU → Target System → Accessible Nodes. If more than one entry is listed under the same CPU, the second one must be closed.
  4. Check PG/PC interface assignment. Verify the active S7ONLINE access point in Set PG/PC Interface points to the correct TCP/IP or PROFIBUS adapter, not a virtual loopback that could open a second logical connection.

4.2 NetPro Active Connection Audit

NetPro is the canonical source of truth for live S7 connections in PCS 7. Open NetPro → right-click the CPU → Object Properties → Connections or simply use the online status icons on the connection table.

Connection Status (NetPro) Meaning Action
Established (green) Active online S7 connection present Identify the partner and close it before download
Established by Partner (yellow) Remote end holds the connection, local CPU is the resource Stop the partner service; cannot be released from NetPro
Not Connected (gray) Configured but no live handle Safe to proceed
Disabled (red strikethrough) Manually disabled in NetPro Re-enable after download if required

Use NetPro → PLC → Connection Status (or Target System → Connections in newer Step 7 builds) to list every active handle. The column Active Connection reveals which remote TSAP / IP is currently talking to the CPU. If a service such as IBA Analyser is identified, stop that service from the Windows service console, wait 30 seconds, and retry the download.

4.3 Identifying Hidden Services

For 13:5472 the partner is often not in the PCS 7 project at all — it is a third-party data collector. To find it:

  1. On the Engineering Station and on any adjacent historian or DAQ server, run netstat -anob with administrator rights. Sort the output for the CPU's IP and look at the owning process.
  2. Open SIMATIC Manager → Options → Set PG/PC Interface → Diagnostics and trigger a connection test to the CPU. The diagnostics pane lists the active TSAP and remote port.
  3. On the CPU itself, use Online → Accessible Nodes → Online & Diagnostics to read the communication statistics. A non-zero number of active S7 connections after closing all known clients is a strong indicator of a hidden service.

5. Resolution: Solo-Mode Download on an H-System

When the engineering station cannot release the connection in the time available (typical for operating plants that cannot tolerate a full ES restart mid-shift), the field-proven workaround is to download in Solo mode on each redundant CPU. The H-system is designed to permit this without a full process interruption.

5.1 Solo-Mode Concept

In an S7-400H pair, Solo mode is a maintenance state in which one CPU is logically disconnected from the redundancy link and operated as if it were a stand-alone CPU. The remaining CPU carries the full process, while the solo CPU is taken out of the redundancy, downloaded, verified, and then re-attached. PCS 7 handles the failover transparently as long as the process is not running at the redundancy limit (CPU > 90% load, network link at full bandwidth, etc.).

5.2 Step-by-Step Solo Download Procedure

  1. Identify the standby CPU. In SIMATIC Manager → Accessible Nodes or in WinCC diagnostics, note which CPU is the active master (RUN with redundancy role = Master) and which is the standby (RUN with redundancy role = Standby/Reserve).
  2. Switch the standby CPU to Solo mode:
    Right-click the CPU in HW Config → Target System → Operating Mode → Switch to Solo Mode. Confirm the prompt. The standby CPU leaves the redundancy pair; the active CPU remains master and continues to control the plant without interruption.
  3. Download the HW Config change:
    With HW Config open, select the now-solo CPU and click Download to Target System. The 13:5472 dialog is no longer raised because the solo CPU is the only CPU in the project tree and there is no redundant partner to coordinate with. The Simeas P (or other added device) downloads cleanly.
  4. Verify the download on the solo CPU:
    Open Online → Online & Diagnostics → Diagnostics Buffer on the solo CPU. Look for the entry "System data download OK". Cross-check the new module appears in the online view of HW Config with the configured PI address.
  5. Re-attach the solo CPU to the redundancy:
    Target System → Operating Mode → Switch to Redundant Mode. The CPU re-synchronises with the active master. Allow 30–120 seconds for link-up, depending on the number of OB1 cycles buffered. PCS 7 will not allow the switch back if the system data blocks on the two CPUs are inconsistent — Step 7 enforces this with a comparison before the link command.
  6. Switch the active CPU to Solo:
    Repeat the same procedure on the master. The roles reverse: the previously-solo CPU becomes master, the new solo CPU is the one to be downloaded. Perform the HW Config download. Verify the diagnostics buffer. Re-attach to redundancy.
  7. Final consistency check:
    With both CPUs back in redundant mode, run HW Config → Station → Consistency Check against the project and against the online view. The two must match. Open CFC → Chart → Check Consistency to ensure the Simeas P signal is wired and compiled correctly in the CFCs.
Operational caveat: During the brief window when only one CPU is in the pair, the system has zero redundancy. Any fault on the active CPU during this window is a process trip. Schedule solo operations during a low-risk maintenance window, and verify that the second CPU has not accumulated un-acknowledged diagnostics that would cause an automatic failover attempt.

6. Alternative: Downloading via CFC Test Mode in RUN

If the hardware change is purely a parameterization update (for example, a new measurement range on a Simeas P that is already configured in the project), PCS 7 V8.0 SP2 supports a CFC online change download in RUN without touching HW Config. The path is:

  1. Edit the CFC chart that owns the new signal.
  2. Insert the new block or change the existing block's parameters.
  3. Click Chart → Download → Changes. The download goes via the operating-mode service in OB 1, not the system-data path that triggers 13:5472.

CFC changes are permitted alongside other online connections because they do not modify the SDB container. If a 13:5472 blocks HW Config but the change is achievable via CFC parameter modification, prefer the CFC download — it leaves the S7 connection unblocked for other services and does not require a solo-mode switch.

7. Operating-Mode Transitions and Timing

The H-system operating-mode transitions are non-trivial. Use the following timing table as a field reference for procedure planning and operator communication.

Transition Typical Duration Process Impact
Standby CPU → Solo mode 5–15 s None (active CPU continues)
HW Config download on solo CPU 20 s – 3 min None (solo CPU not in process control)
Diagnostics buffer verification 10 s None
Solo CPU → Redundant mode 30–120 s None (link-up, no role change)
Active CPU → Solo mode 5–15 s Failover; OB1 / OB40 may be triggered if a process is in active step
HW Config download on new solo CPU 20 s – 3 min None (new solo CPU not in process control)
Second CPU → Redundant mode 30–120 s None (link-up)
Total plant exposure to single CPU Approx. 1–4 min Zero redundancy; brief trip risk

8. Verification Checklist

After the download, confirm the following before signing off the work order:

  • Diagnostics buffer of both CPUs shows "System data download OK" with the same timestamp and revision counter.
  • HW Config online view matches the offline project: the Simeas P appears at the new PI address with no yellow or red status icons.
  • NetPro active connections list is identical to the pre-change snapshot, with no orphaned connections left behind.
  • CFC compile produces no warnings; chart-inconsistency list is empty.
  • WinCC / OS server compiles the project, and the new tag (e.g., analog input from the Simeas P) is visible in the picture and updates at the configured acquisition cycle.
  • Redundancy status in Online → Accessible Nodes on both CPUs shows RUN-REDUN (not RUN-SOLO) and the sync link LED is solid green on both.
  • Operator round: confirm no fault messages in the OS alarm log during the procedure window.

9. Prevention and Best Practices

  1. Single-online rule: enforce a project policy that only one engineering station may be online to a given CPU at a time. Use the SIMATIC Logon / password-protected connection rights in PCS 7 V8.0+ to prevent inadvertent second sessions.
  2. Close all editors before HW Config: the project tree should show only HW Config open during a download. CFC, SFC, NetPro, and LAD/FBD all leave passive handles.
  3. Audit third-party data services: any tool that opens an S7 connection (IBA, OPC DA, custom .NET S7 libraries) should be inventoried in the plant IT/OT document and excluded from the same subnet as the engineering station during change windows.
  4. Pre-stage changes in a test project: validate the HW Config change on an AS replica before applying to the live pair. PCS 7's AS-OS Engineering workflow supports this with branch projects.
  5. Use CFC online change when possible: it bypasses the SDB write path and therefore does not trigger 13:5472.
  6. Schedule solo operations during planned maintenance or low-load windows; never during a step change in the process.
  7. Keep the project backup current: before any solo-mode operation, run Station → Save As with a timestamped archive so a roll-back path is always available.

10. Related Siemens Documentation

The following Siemens resources are useful background when escalating 13:5472 or planning a solo download:

11. Troubleshooting Matrix

Symptom Most Likely Cause First Action
13:5472 raised immediately, no other PG visible Third-party S7 client (IBA, OPC) holding connection netstat -anob on ES, stop the offending service
13:5472 raised only during HW Config download NetPro or CFC open in background Close all editors except HW Config
13:5472 raised on both H CPUs H-system sees partner-CPU connection conflict Use solo-mode download procedure (Section 5)
Download OK on solo CPU, re-attach fails SDB mismatch between pair Re-download to both CPUs in solo sequence
WinCC loses tag after download OS compile not re-run OS server compile and download
Simeas P shows red in online HW Config PROFIBUS address conflict or terminated slave Verify PI address within CPU limits, check bus termination

What does Siemens error 13:5472 mean on a CPU 416-5H?

Error 13:5472 is a Step 7 / PCS 7 engineering-side message stating "There are other online connections to this CPU. Close all connections before loading." The CPU is reporting that another online S7 session (PG, OPC client, IBA Analyser, etc.) is holding a connection that would be invalidated by an SDB-system-data download. The CPU itself is not at fault; the engineering station refuses to proceed.

How do I download HW Config to a redundant CPU 416-5H pair in RUN without a process trip?

Use the Solo-mode procedure: switch the standby CPU to Solo, download the HW Config to it, verify, re-attach to redundancy; then switch the active CPU to Solo, download to it, verify, and re-attach. Total single-CPU exposure is typically 1–4 minutes per side, with no process interruption if the active CPU remains healthy.

Can I download hardware changes to a PCS 7 V8.0 SP2 CPU 416-5H without going through Solo mode?

Yes, if the change is achievable as a CFC parameter modification rather than a system-data change. CFC online download in RUN writes to the working data blocks only and does not modify the SDB container, so it does not trigger 13:5472. For true hardware additions (new modules, new PROFIBUS slaves, new Simeas P devices), Solo mode is the field-proven path.

How do I find which third-party application is holding the S7 connection that triggers 13:5472?

On the Engineering Station and any adjacent historian / DAQ servers, run netstat -anob with administrator rights and filter for the CPU's IP. Identify the owning process. Common culprits are IBA Analyser, OPC Scout / DA servers, and custom .NET S7 libraries. Stop the service, wait 30 seconds, and retry the HW Config download.

Why is the Simeas P PI address being rejected during the consistency check?

CPU 416-5H has a hard limit on the number of PROFIBUS slaves and the range of permissible PI addresses. Verify the new address is inside the CPU's address range, not already assigned to another slave in HW Config, and within the configured PROFIBUS segment's baud-rate capability. The Step 7 consistency check is a static pass/fail and will reject any out-of-range assignment before the download is even attempted.

Back to blog