Detecting PROFINET Device Errors on S7-1200 in TIA Portal V15

1. Overview

Detecting a peripheral fault on a Siemens SIMATIC S7-1200 connected to a PROFINET IO device (for example, a Festo valve manifold or I/O module referenced in the SIMATIC compatibility list) requires more than simply waiting for a watchdog LED. TIA Portal V15 provides three complementary mechanisms:

• OB82 – Diagnostic interrupt OB, triggered by a device-originated diagnostic event.
• OB86 – Rack or PROFINET IO system failure OB, triggered when an IO device goes offline, returns, or fails to come online.
• GET_DIAG – A user-program instruction that reads structured diagnostic data from a PROFINET IO device, including channel status, port status, and module status.

Many users start with OB82 and find it silent when a device is physically disconnected. This is expected: OB82 fires for a diagnostic interrupt condition reported by the device, not for a link-down condition. For a cable pull or power-loss on a PROFINET slave, OB86 is the correct OB. GET_DIAG complements both by providing a polled, structured view of the device's current health.

This article covers the architecture of PROFINET diagnostics on S7-1200, the role of each OB, the configuration of OB86, the parameter set of GET_DIAG, and a startup-time strategy that handles the 30–45 second PROFINET bring-up window typical for cold-start configurations.

2. Prerequisites

3. PROFINET Diagnostic Architecture on S7-1200

PROFINET diagnostics flow upward through three layers:

1. Channel diagnostics – A specific channel inside a slot reports a problem (e.g., wire break on input 3 of a digital input module). Encoded as a channel-related diagnostic record (alarm type 0x0002 / 0x0003).
2. Module diagnostics – An entire module is flagged. Encoded as alarm type 0x0004 / 0x0005.
3. Device / station diagnostics – The entire PROFINET IO device becomes unavailable (link loss, power loss, watchdog timeout). Encoded as a PNIO alarm with ChannelProperties set to the device (slot 0).

The S7-1200 firmware translates these alarms into OB calls. The mapping is as follows:

For a Festo CTEU or CPX manifold pulled from the network while the CPU is in RUN, OB86 is the only OB that fires reliably. OB82 will not fire because the device no longer has power to send a diagnostic alarm. The user's symptom – "OB82 without success" – is therefore not a configuration error; it is a normal behavior of the OB82 contract.

4. Step-by-Step: Configure OB86 for PROFINET Failure Detection

4.1 Create the OB

1. In the TIA Portal project tree, right-click Program blocks > Add new block > Organization block.
2. Select OB86 (Rack or station failure). The S7-1200 firmware automatically populates the OB's local variables with the failure event and the affected logical address.
3. Choose a language (LAD, FBD, or SCL). For pure detection, FBD is clearest.

4.2 Read OB86 local variables

OB86 exposes the following input parameters in its temporary variables (visible in the OB's interface):

4.3 SCL example – fault flag and diagnostic buffer entry

// OB86 - PROFINET station failure handling
#iEventClass := OB86_EV_CLASS;       // B#16#38 leaving, B#16#39 entering
#wFaultIO    := OB86_MDL_ADDR;       // Logical base address of failed device
#bEventID    := OB86_EVENT;          // Detailed event code

IF #iEventClass = B#16#39 THEN
    // Coming event: device is back
    "DB_Diag".DeviceFault[#wFaultIO] := FALSE;
    "DB_Diag".LastRecoveredIO := #wFaultIO;
ELSIF #iEventClass = B#16#38 THEN
    // Going event: device is gone
    "DB_Diag".DeviceFault[#wFaultIO] := TRUE;
    "DB_Diag".LastFaultedIO := #wFaultIO;
END_IF;

5. Step-by-Step: Use GET_DIAG for Structured Diagnostics

OB86 is event-driven and binary: you know that a device has gone away, but you do not know why. GET_DIAG reads the current state of a PROFINET device, port, or module, including error codes, on demand.

5.1 Insert the instruction

1. Open the OB where you want to perform the diagnostic poll (typically a cyclic OB1 segment or a watchdog OB35).
2. From the Instructions pane, expand Diagnostics > GET_DIAG and drag it into the segment.
3. On the instruction, click the MODE input to expose additional outputs.

5.2 Pinout of GET_DIAG

For details, see the SIMATIC S7-1200 / S7-1500 GET_DIAG reference (entry ID 57132240).

5.3 SCL example – poll all configured PROFINET devices every 2 s

// Call once per scan in OB1 or on a 2-second timer
FOR #i := 0 TO "DB_Devices".Count DO
    #iStatus := GET_DIAG(
        REQ     := #xTrig,             // pulse from timer
        MODE    := 0,                  // device-level status
        LADDR   := "DB_Devices".Addr[#i],
        RET_VAL := #iRetVal,
        BUSY    := #xBusy,
        STATUS  => "DB_Diag".Status[#i]
    );
    IF #iRetVal = 0 AND #xBusy = FALSE THEN
        "DB_Diag".Present[#i] := ("DB_Diag".Status[#i] AND 16#0000_0001) <> 0;
    END_IF;
END_FOR;

6. Handling PLC Startup State

After a STOP-to-RUN transition, all PROFINET devices must be parameterized and AR-established. The S7-1200 user report quoted a 45-second window before all devices are responsive. This is realistic for an installation with multiple slaves and I&M data exchange. The recommended strategy is to suppress fault evaluation until the system is in steady state.

6.1 Use a startup-elapsed timer

// In OB1 or a watchdog OB
IF "DB_Init".StartTime = 0 THEN
    "DB_Init".StartTime := TIME_TCK();
END_IF;
IF (TIME_TCK() - "DB_Init".StartTime) > T#45s THEN
    "DB_Init".Ready := TRUE;       // diagnostics are now valid
END_IF;

6.2 Use the standard OB100 / startup OBs

Add code in OB100 (warm restart) to clear your fault flags and the diagnostic DB at the moment of transition. This ensures stale entries from the previous session are not latched into the new RUN cycle. On the S7-1200, OB100 is automatically called once after STOP-to-RUN; it is the correct place to initialize state.

7. Verification and Online Testing

1. Build and download the project to the CPU.
2. Go online in TIA Portal. Open Online & Diagnostics > Diagnostics buffer.
3. Power the CPU and all PN devices. Wait for the AR to be established (the I/O area of the device turns green in Online & Diagnostics > PROFINET IO).
4. Force a fault: pull the PROFINET cable from the Festo device or cycle its 24 V supply.
5. Watch the diagnostic buffer. You should see entries of class "IO device failure" with the LADDR matching the device.
6. Confirm that DB_Diag.DeviceFault[LADDR] flips to TRUE in the online watch table.
7. Reconnect the device. The flag should return to FALSE on the "coming" event of OB86.

If the buffer shows "IO device failure" but OB86 does not execute, the most common cause is that OB86 was not downloaded or its priority was overridden. Confirm by right-clicking the project > Compile > Software (rebuild all blocks).

8. Edge Cases and Field Notes

8.1 Partial station failure

For a Festo CPX manifold with multiple sub-modules, an individual sub-module (e.g., one valve slice) may fail while the rest of the station remains online. In this case OB86 fires with EVENT = B#16#35 (partial failure) rather than B#16#38. The ID parameter identifies the slot. Your handler should distinguish these cases if you need per-module granularity.

8.2 Tool changer and topology reconfiguration

If your Festo device is part of a tool-changing PROFINET topology, the device may deliberately come and go. Do not treat every OB86 event as a fault – instead, use it as a state change notification and combine it with a permissive from your tool-change handshake.

8.3 Diagnostic alarm from a powered device

If a powered Festo device reports a wire break or overload via a diagnostic alarm, OB86 will not fire. You need OB82 for that. The fix is to keep both OBs in the program: OB82 for "device says something is wrong", OB86 for "device is unreachable".

8.4 Acyclic parameter read for non-listed devices

If the Festo slave was added with its GSDML but is not in the device catalog, you can still read acyclic parameters using the RDREC instruction (index 0x8000 series). This returns the IM0/IM1/IM2 records, which include the OrderId and serial number – useful for confirming that the device responding is the device you expect.

8.5 Watchdog time

For a station that may be temporarily offline (e.g., on a flexible cable carrier), raise the PROFINET watchdog from the default 3 to 30 cycles (in Devices & Networks > Properties > PROFINET interface > Watchdog). This prevents OB86 from firing during a normal 1–2 s plug event.

9. Troubleshooting Matrix

10. Reference Links

• SIMATIC S7-1200 Programmable Controller – System Manual (entry ID 109751587)
• STEP 7 V15.1 / TIA Portal V15.1 Update (entry ID 109751559)
• GET_DIAG / RDREC / WRREC instructions (entry ID 57132240)
• PROFINET diagnostics with SIMATIC S7 (entry ID 49948856)
• OB82, OB83, OB86, OB122 on S7-1200/S7-1500 (entry ID 77430411)

Frequently Asked Questions