IMDTouchKnowledge BaseShop main store

How-to

How to Show Only a PLC Connection Alarm in WinCC Flexible

HMI and touchscreens Published 6 min read

Short answer

Yes, but on WinCC flexible 2005 you do it with a dedicated application alarm driven by the connection status, not by filtering the system-alarm stream: the 2005 alarm view selects whole alarm classes, not single messages. Decide first whether the operator must see communication loss or a CPU in STOP, since a stopped PLC can still communicate, and keep the unfiltered system-alarm view on a maintenance screen.

On this page
  1. Decide what the alarm must mean
  2. Capture the real failure messages first
  3. Build the application alarm on the connection status
  4. What alarm-view filtering can do on each runtime
  5. Worked example: sizing the stale-data timeout for a 1 s heartbeat
  6. Acceptance test for the finished alarm
  7. Questions asked

Yes, you can show only the PLC-loss indication, but on WinCC flexible 2005 the right tool is a dedicated application alarm driven by the connection status, not the system-alarm view. The 2005 alarm view is selected by whole alarm classes, so it cannot isolate one system message out of the stream; that is why turning on all system alarms floods the operator. Decide first whether the alarm must report communication loss or a CPU in STOP, because those are different failures, then build the alarm and keep the unfiltered system-alarm stream on a maintenance screen. The steps below were written for an MP370 15-inch touch and hold for any WinCC flexible 2005 target.

Decide what the alarm must mean #

A system alarm is a diagnostic message generated by the HMI runtime itself; an application alarm is a message triggered by a condition you configure in the project. Your requirement is one operator-facing condition, so fix its meaning before touching any dialog and write it into the commissioning record:

Condition What it tells the operator Wording that fits
HMI cannot exchange data with the PLC Communication unavailable PLC communication lost
CPU reports STOP while communication is up Controller not executing its program PLC stopped
Heartbeat value stops changing Fresh application data not arriving PLC data stale

One alarm may cover several rows only if its wording says so. Calling every case connection lost sends maintenance to the network when the cable is fine and the CPU is simply in STOP, and a stopped PLC can still communicate.

Capture the real failure messages first #

  1. Take the machine to an approved test state before forcing anything: a running machine that depends on this link behaves as though the controller had failed.
  2. Add a temporary screen with an unfiltered system-alarm view and keep it off the operator screen. Record the complete message text, the identifier the runtime displays, and the order in which messages appear and clear.
  3. Break the path with a controlled maintenance action, such as disconnecting at a service point. Watch the process values stop updating, and separate the failure messages from the recovery messages, because the first visible message may clear long before the connection is healthy again.
  4. Restore the path and confirm live updates resume.
  5. Where the operating procedure permits, put the CPU in STOP with the cable connected, and record which messages respond to operating mode rather than transport.

If the loss is intermittent rather than a clean break, work through diagnosing intermittent HMI to PLC Ethernet comms hangs before you freeze the alarm wording. While recording the sequence, give yourself time to read it: the reasoning in why extending HMI system event display time reveals comms faults applies to any system-alarm capture.

Build the application alarm on the connection status #

Filtering is limited on this runtime, so the reliable path is your own alarm:

  1. In the communication area of the project, enable the per-connection status indication documented for your driver and connect it to an internal HMI tag. The status must be produced on the HMI side: a PLC tag cannot report the loss of its own communication path, and an unchanged received value is the last good sample, not a healthy connection.
  2. Create an application alarm triggered by the status condition that means connection down, worded from the table above.
  3. For the STOP row, use the CPU operating-mode indication documented in Siemens support entry 19687561 and treat STOP as offline for the operator. Either word one combined alarm PLC offline so it covers transport loss and STOP, or run a second alarm worded PLC stopped when maintenance needs the distinction on screen.
  4. Configure activation, clearing and acknowledgment separately. Acknowledgment records that the operator saw the alarm; it must never make an ongoing failure look healthy.
  5. If you also need the stale-data row, monitor a heartbeat the PLC toggles and word that alarm PLC data stale: a stopped CPU, a broken connection and a wedged HMI all stop the toggling, and the text must own that scope.

What alarm-view filtering can do on each runtime #

On WinCC flexible 2005 the alarm view is configured by alarm class and alarm group. You can dedicate a view to the system-alarm class, but you cannot select one message inside that class, so commit to the application alarm above instead of fighting the view.

WinCC Flexible 2008 gives the alarm view a filter tab, and the filter is a compile-time property: you set it in the engineering tool and transfer it, and the operator cannot change it at runtime. That filter matches the alarm text, the alarm number or the tag name as a case-insensitive substring, so on a 2008 project you can reduce a system-alarm view to the connection message by entering part of its captured text. On current WinCC Basic Panels the same control sits at Properties > Properties > Filter > Alarm filter in the Inspector window, with a Custom option for the substring.

Whatever the runtime, keep one unfiltered system-alarm view on a maintenance screen. That is where the rest of the stream stays useful, and it is the reference you compare against whenever a new diagnostic appears over the panel's life.

Worked example: sizing the stale-data timeout for a 1 s heartbeat #

Inputs, with sources: the PLC program toggles a heartbeat bit every 1.0 s; the WinCC flexible tag for that bit is set to acquire every 1.0 s; the design rule is to declare the data stale after three missed toggles, which rides through two missed changes and normal jitter. These are the example's configured values: substitute your own toggle period and acquisition cycle and the arithmetic holds.

The rule in words: the timeout equals three toggle periods, and the worst-case detection time equals that timeout plus one acquisition cycle, because the HMI can be up to one cycle away from seeing the last change.

Arithmetic written out:

  • Timeout: 3 × 1.0 s = 3.0 s
  • Worst-case detection: 3.0 s + 1.0 s = 4.0 s

Result: a dead connection or a stopped program is flagged 4.0 s after the last observed toggle, worst case.

Decision: against an operator response budget of 10 s for this machine, 4.0 s detection passes with room for the banner and the acknowledgment. If your budget is 5 s, halve the toggle period: 3 × 0.5 s = 1.5 s, and 1.5 s + 0.5 s = 2.0 s, which passes. If the budget is under 2 s, drop the heartbeat and rely on the connection-status alarm from the previous section, which flags the loss without waiting on data.

Acceptance test for the finished alarm #

With the machine in an approved test state:

  1. Healthy connection: process values update and no alarm is active.
  2. Break the connection at the service point: the application alarm appears within the time computed above, carries the wording from the table, and appears only on the operator screen.
  3. Restore the connection: the alarm clears or waits for acknowledgment exactly as configured.
  4. CPU to STOP with the cable connected: the offline or STOP alarm appears; with separate alarms, the communication alarm stays quiet.
  5. Trigger one unrelated diagnostic: it shows on the maintenance system-alarm view only. That last check is the acceptance that the operator sees one condition, not the stream.

Questions asked

Can WinCC flexible 2005 filter out individual system alarms?

No. The 2005 alarm view is selected by alarm class or group, so you can show or hide the system-alarm class as a whole but not one message inside it. WinCC Flexible 2008 added the per-view filter, a compile-time property that matches alarm text, alarm number or tag name as a case-insensitive substring. Current WinCC also has an integrated filtering tool in the alarm view: pick the property to filter on and enter the text to search for.

How do I tell a lost connection from a PLC in STOP?

Use your capture record to tell them apart: if the panel still shows live values while the CPU is in STOP, communication is up and the wording belongs to the STOP row. For the alarm itself, use the CPU operating-mode indication documented in Siemens support entry 19687561 and treat STOP as offline, either as one combined PLC offline alarm or a second alarm worded PLC stopped.

How quickly can the panel detect a dead connection?

A status-driven alarm appears as soon as the driver flags the connection down, within a few acquisition cycles and with no timeout to wait out. A heartbeat monitor adds its timeout on top: with the worked example settings, worst case is 4.0 s. Match the timeout to the operator response budget for the machine, not to the network speed.

Should the alarm clear itself when the connection returns?

Let the alarm go to its cleared state when the connection returns, so the banner leaves the screen on recovery. Keep acknowledgment as the operator record of having seen it, and keep the active condition independent: acknowledging never resets a failure that is still present. Configure the two behaviors in separate dialogs so they cannot be merged by accident.

Keep reading

  • Troubleshooting

    Why rebuilt HMI runtime files show #### for PLC tags

    Hashes on every tag after transferring a rebuilt WinCC Runtime Advanced project are a display symptom, not a communication code. Restore the working runtime files as a baseline, add an alarm view showing buffered System alarms, and branch on the message text: connection establishment, TLS certificates and clocks, hidden project pairing, or single-field formatting.

    HMI and touchscreens

  • Troubleshooting

    Why WinCC Flexible Shows a Blank Station Search Over MPI

    A blank station field in WinCC flexible means the S7ONLINE programming path reaches the CPU but the HMI connection has no defined route. Give every MPI node a unique address — PG 0, panel 1, CPU 2, a second PC 3 — confirm the USB adapter supports OP as well as PG connections, and expect STEP 7 symbols only through an integrated HMI station.

    HMI and touchscreens

  • How-to

    Building a Pie Chart on a Comfort Panel in TIA Portal

    A TP1200 Comfort project in TIA Portal V17 has no pie chart object, and its scripts cannot draw one. Compute each sector in a local script as 360 degrees times value divided by total, guard the zero-total and invalid-data cases before dividing, then display the result by switching pre-drawn chart images from an integer tag or scaling bar objects to the percentages.

    HMI and touchscreens