IMDTouchKnowledge BaseShop main store

How-to

Showing a PLC Connection Fault on a Comfort Panel Alarm Display

HMI and touchscreens Published 9 min read

Short answer

Enable the Comfort Panel's connection and disconnection system alarms, add an alarm buffer view, and drive a persistent HMI-local status tag from their events, because the system message shows for about 2 seconds and then leaves the screen while values freeze; a PLC clock heartbeat adds a stale-data check.

On this page
  1. How do I make a Comfort Panel show a lost PLC connection?
  2. Which system alarms are the connect and disconnect messages?
  3. Can I just extend how long the message stays on screen?
  4. How do I add an alarm buffer that keeps the disconnect history?
  5. How do I keep a status on screen after the message clears?
  6. How do I know the values on screen are still live?
  7. Doesn't the area pointer already do this?
  8. How do I verify all of it without waiting for a real fault?
  9. Questions asked

Enable the panel's connection and disconnection system alarms in the operator alarm view, add an alarm buffer that keeps their history, and drive a persistent HMI-local status tag from those alarms' events — then verify all of it with one deliberate disconnect. The system message alone is not a strategy: it displays for about 2 seconds at the default duration, then leaves the current-alarm view, while every value on screen stays frozen at its last good read. Four layers give complete visibility: the alarms themselves, the buffer, the status tag, and a heartbeat that flags stale data.

How do I make a Comfort Panel show a lost PLC connection? #

Enable the connection and disconnection system alarms in the alarm view's filter, and put an alarm buffer on a maintenance screen, because the default display time of the system message is about 2 seconds and the panel otherwise looks healthy while its values freeze.

The panel raises these messages locally, so they appear even when nothing can be read from the controller; what is missing by default is an alarm display whose filter includes them. Build visibility in layers, because each layer answers a different question:

Layer What it answers Built-in limit
System alarm in the current view Which communication event is happening now Clears after about 2 s at the default display duration
Longer display duration Keeps the transient message readable Still self-clears; no history, no live status
Alarm buffer including system alarms Did it drop, and when did it come back Finite entries; says nothing about the present state
Persistent status tag Is this connection up right now Must be initialized and re-tested after a panel restart
PLC-clock heartbeat Is the data on screen fresh Needs a continuously changing value and a timeout

Which system alarms are the connect and disconnect messages? #

The engineering project's alarm editor carries a long system alarm list with dedicated entries for connection establishment and connection loss; select the entries for the connection your operators actually use rather than the whole category.

The event number 14000 gets suggested for this job. Treat it as a candidate: open its documented entry in the system alarm list of your installed version and confirm the wording before you build anything on it. System alarm texts can also carry parameters that look cryptic — they reference the runtime's internals, and they are exactly what product support asks for, so keep the full text in the buffer unedited.

Enable the two communication entries, not the entire system event category, or the current view floods with unrelated diagnostics and buries the one message that matters. Filter at the display: give the operator screen the communication entries plus process alarms, and leave the full system list to the maintenance buffer. If the machine still runs a WinCC Flexible project, showing only a PLC connection alarm in WinCC Flexible is its own separate filter configuration.

Can I just extend how long the message stays on screen? #

Extending the display duration to something like 10 seconds buys one operator glance, but the message still self-clears, so it leaves no history and no live connection status for the screens that matter.

The display duration of system events is a project setting; raise it and the loss message holds the current view instead of flashing past. Extending HMI system event display time to reveal comms faults is worth doing in the same pass, because faults that recur every few minutes become visible as repeated entries rather than one blur.

Treat the extension as comfort, not coverage. After the extended message clears, you still cannot tell from the screen whether the link came back.

How do I add an alarm buffer that keeps the disconnect history? #

Add a second alarm view configured as a buffer, with the system alarm class included in its filter, so the connect and disconnect entries stay listed with timestamps after the transient message has left the current-alarm view.

  1. Place an alarm view on a maintenance screen or a separate operator page, and configure it as an alarm buffer rather than a current-alarms view.
  2. Include the system alarm class, or the specific connection entries, in the buffer's filter so communication events are recorded with timestamps.
  3. Leave the buffer reachable without a maintenance login if operators are expected to report when a drop happened; otherwise gate it with the rest of the maintenance screens.
  4. During the verification test at the end of this article, confirm the disconnect entry is still listed after the current view has cleared, and that the reconnect entry follows it on restore.

A buffer is finite and it is not an historian: where your maintenance process needs long-term retention, add proper logging on top. For answering whether the link dropped during the night shift, the buffer is enough.

How do I keep a status on screen after the message clears? #

Use the events of the two system alarms to set and reset an HMI-local tag — connection established sets it, connection loss resets it — and show that tag in a status field or banner on every screen where operators read values or acknowledge alarms.

  1. Confirm the runtime supports events on system alarms: Comfort Panels do, Basic Panels do not and need the color-dynamization workaround instead.
  2. Create an internal tag that lives on the panel itself, initialized to the not-connected state.
  3. On the connection-established event of the intended connection, add a tag action that sets the internal tag; on the connection-loss event, add the action that resets it.
  4. Bind a banner or status field on the operating screens to that tag, colored so the fault state survives glare and a quick glance.
  5. If the project talks to several PLCs, give each connection its own tag, so one connection's disconnect event can never make another PLC look healthy.

The tag must be HMI-local because a bit that lives in the PLC cannot deliver a new value to the panel after the link has failed — the failure you are reporting is the failure that would block the report.

How do I know the values on screen are still live? #

Watch a PLC value that changes continuously — the PLC's clock is the usual choice — and declare the data stale when the received value stops changing for longer than your timeout, with the timeout sized from the update and acquisition periods.

A status tag answers whether the connection is up; a heartbeat answers whether data is still arriving, which is what matters when someone is about to act on a number. Compare successive received values on the panel, not the absolute clock readings of two devices — clocks are offset, and a received timestamp marks the last observed update, not the exact instant the path failed. Set the heartbeat tag's acquisition so it keeps updating while other screens are open, or the check goes blind the moment the operator navigates away.

Worked example, sizing the timeout: the PLC updates a seconds counter every 1 s, and the panel's acquisition cycle for that tag is 1 s. The worst case for a healthy link to show no change is one PLC update period plus one acquisition cycle: 1 s + 1 s = 2 s. Set the stale timeout to 5 s — the 2 s worst case plus 3 s of margin for one missed poll or a retry — so the check does not false-alarm on jitter.

At runtime, the value last changed at 12:04:31 and the check runs at 12:04:37: 12:04:37 - 12:04:31 = 6 s, and 6 s > 5 s, so the data is declared stale while the last good values stay on screen. If you need faster detection, shortening the acquisition cycle to 500 ms drops the worst case to 1 s + 0.5 s = 1.5 s and supports a 4 s timeout (1.5 s + 2.5 s margin): detection about 1 s sooner, at double the polling load on that tag.

The pattern travels beyond Ethernet panels: a radio link heartbeat check between a PLC and SCADA watches clock values crossing the link in both directions for exactly this reason, and the compare-on-both-sides idea is the same one.

Doesn't the area pointer already do this? #

No — the area pointer runs in the opposite direction: the PLC uses it to supervise whether the HMI is still alive, so it cannot tell your panel that the PLC side has gone away; for that you need the system alarms and the heartbeat above.

Keep the area pointer for what it does, letting the controller react to a dead panel. If you want supervision in both directions, toggle a bit on each side and time out when it stops changing; that explicit handshake is easier for the next technician to read than a configured area pointer, and it gives both ends a failure indication.

How do I verify all of it without waiting for a real fault? #

Pull the connection deliberately during a quiet moment and watch three things — the buffer entry, the persistent tag, and the heartbeat's stale flag — then restore the link and confirm the reconnect entry, the tag returning, and the value changing again.

  1. Do it with the machine stopped or in a proven safe state: a panel that cannot see the PLC can also stop reacting to operator input, and this is not the moment to discover what the controller does on communication loss.
  2. Open the operator alarm screen, then interrupt the connection at the controller end — patch cable out or PLC powered down per your site's practice.
  3. The loss message appears, holds for its display duration, then clears; the status banner should drop and stay down; after the heartbeat timeout the stale flag should rise while the last values stay on screen.
  4. Restore the link: the reconnect entry appears in the buffer, the banner returns, and the stale flag clears on the next observed change.
  5. Restart the panel while the PLC is still unreachable: the status must come up as not connected, never as a leftover healthy state.
  6. Read the buffer timestamps back against your watch: the recorded sequence is your evidence that the configuration works.

If those timestamps show drops nobody noticed, diagnosing intermittent HMI to PLC Ethernet comms hangs is the next job, and the timestamps you just verified are its input data.

Decide by screen. An operator screen that people act on gets the persistent status tag plus the buffer, and the heartbeat wherever its numbers are trusted as live. A maintenance screen gets the buffer with the full system alarm list. Extending the message duration is worth doing everywhere, and the un-extended default message alone protects nobody.

Questions asked

Can I do the tag set-and-reset trick on a Basic Panel?

No. Basic Panels do not support events on system alarms, so the connection-established and connection-loss events cannot drive a tag. The documented workaround there is color dynamization: animate an element's background so it flashes while the connection is healthy and holds one steady color when it is not. The alarm buffer part works the same as on a Comfort Panel.

Why does my alarm buffer come up empty after a panel restart?

Comfort Panel alarm buffers coming back empty after a restart is a known issue across TIA Portal V13 SP1 through V20, and the fix documented for it is moving the alarm log to persistent storage, typically a USB stick, in the panel's logging settings. After that change, cycle power once and confirm the entries survive.

The panel lost the PLC right after a firmware update. Is my alarm configuration wrong?

Probably not. There is a documented case where Basic and Comfort panels, and Runtime Advanced, failed to establish a connection to the controller after WinCC Panel Images V17 Update 6. Check which panel image the unit runs before reworking alarms; if the timing matches an update, the image is the suspect, not the display configuration.

Why doesn't the acknowledge key clear alarms during the outage?

Acknowledgement of an alarm that lives in the PLC has to be written to the PLC, and that write cannot complete while the connection is down. Repeated presses change nothing on the controller. Restore the link first, then acknowledge; the buffer shows what arrived in the meantime.

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

    Getting Unconfirmed Alarm Status Out of a C-More Panel

    No tag on the EA9 exports alarm-confirmation state to the PLC, so the buzzer cannot follow the C-more confirm action. Latch the alarms in the PLC, acknowledge them with a panel pushbutton that writes a PLC bit, and run the buzzer while any alarm stays unacknowledged; the PLC to Panel Beep ON/OFF tag can mirror the same state onto the panel beep.

    HMI and touchscreens

  • How-to

    How to Display PLC CSV Data Logs on an HMI Screen

    A standard HMI tag driver exchanges tag values with the PLC; it cannot browse the controller memory card or parse its CSV files. For an operator-facing history, log the same tags with the HMI's own sampling and bind the screen's trend or table to that dataset. Display the exact PLC-generated CSV only after transferring it to a PC that can parse it.

    HMI and touchscreens