Diagnosing Siemens ET200 V90 Software Stops on Profinet IRT

1. Problem Overview

A Siemens ET 200 station with PN interface coupled to a SINAMICS V90 PN servo drive exhibits random software stops in TIA Portal V17 Update 6. The PLC does not transition to STOP, no diagnostic buffer entry is generated, and the S7 diagnostic LEDs stay green. Yet the controlled process halts, often when the drive operates above 90% of rated speed. Engineers typically suspect Profinet IRT timing, OB priority inversion, or V90 telegram loss, but the actual root cause in this case class is an ungrounded, undetected short circuit on a digital input (DI) that returns the input to a logic level the program interprets as a controlled stop request.

This article documents the diagnostic path, the electrical mechanism, the recommended hardware fix (channel-diagnostic DI modules), and the commissioning checks that prevent recurrence.

2. Affected System Configuration

Reference the Siemens Industry Online Support portal for current firmware, manual, and entry-ID updates for ET 200SP, ET 200pro, and V90 PN hardware revisions.

3. Root Cause: Digital Input Short to Ground

The confirmed root cause in this fault class is a short circuit between the 24 V sensor supply (L+) feeding a digital input and the bonded 0 V / PE reference. In a system where:

• The 24 V DC 0 V terminal is bonded to protective earth (grounded DC system, IEC 61131-2 Type 1 wiring, sometimes called "PNP/positive logic" or "sourcing");
• The DI module reads the input as ON when the input terminal is pulled to L+ (24 V), and OFF when the input terminal is at 0 V (grounded);
• A field cable chafes, a proximity switch fails, or a terminal block is miswired so that L+ is shorted to the grounded 0 V return,

the DI channel no longer sees the field device's switching contact. The input is clamped to the same potential as the OFF state by the short itself. The user program reads the contact as "open" or "request = false", which it interprets as a legitimate stop, pause, or permissive-loss condition, and the software implements a controlled ramp-down of the V90 through the Profinet telegram.

4. Why a Short to Ground Stops the Software (Not the PLC)

The mechanism is independent of the drive speed, but it appears correlated to high speed because:

1. At higher mechanical speeds, cable vibration and flex energy increase, accelerating insulation fatigue at the chafe point.
2. The process is more likely to be in the high-velocity state when the program is actively checking permissive logic for the next cycle.
3. Higher current draw on the same 24 V rail produces a small but measurable voltage drop, lowering the threshold margin between a valid low and a marginal high.

5. Profinet IRT and Update Rate Considerations

Profinet IRT (Isochronous Real Time) on the ET 200 PN controller and the V90 PN operates with synchronized, reserved time slots. Typical V90 PN update times range from 1 ms to 4 ms depending on the configured send clock. The IRT configuration has no direct causal link to a DI short, but two indirect effects matter:

• Diagnostic propagation: Because IRT is reserved-bandwidth, a Profinet device that loses its isochronous partner raises alarm slots quickly, but a masked DI short never enters the Profinet diagnostic stream — it is a digital bit in the process image, not a Profinet alarm.
• Cycle synchronization: If the V90 is configured as an isochronous slave, OB61 (synchronization to the bus) or OB6x isochronous OBs run aligned with the IRT cycle. A user stop request issued in OB1 between two IRT cycles will be applied at the next sync boundary, giving a deterministic but delayed response that can look like a Profinet timing problem.

To verify IRT topology and slot timing, use the TIA Portal Topology & Network View > PROFINET IO > Real Time Settings and confirm:

• Send clock matches the V90 rated send clock (V90 PN: typically 1.0 ms minimum, 4.0 ms typical).
• Synchronization role: Sync Master is assigned to a single device (usually the controller), Sync Slaves are checked.
• Reduction ratio is the same for all IRT nodes.

For additional detail on Profinet IRT conformance, refer to the PROFIBUS & PROFINET International (PI) documentation set and the IEC 61158 / IEC 61784 family of standards governing Profinet real-time classes.

6. OB Priority and Cycle Time Implications

OB priority is often the first suspect in a Siemens stop trace because OB1 (priority 1) is the lowest-priority cyclic OB, and any OB with a higher priority can preempt it. However, OB priority does not cause a software stop; it determines execution order. To eliminate OB priority as the cause, audit the project:

Cycle time considerations: with the V90 PN and ET 200 station in IRT, OB1 cycle time on a CPU 1510/1515 typically sits between 2 ms and 8 ms depending on the user program. If OB1 exceeds the configured maximum cycle time (default 150 ms on S7-1500), OB80 fires and the CPU can STOP unless an OB80 is loaded. Confirm the configured Maximum cycle time under PLC properties > Cycle in TIA Portal, and watch the Cycle time (longest) statistic in the Online & Diagnostics view.

7. Diagnostic Strategy

Use this sequence before replacing the DI module. It isolates the short without disrupting production longer than necessary.

1. Capture the diagnostic buffer. In TIA Portal, go Online & Diagnostics > Diagnostic buffer. A truly software-driven stop with no buffer entry confirms the stop is application-issued, not a CPU fault.
2. Enable the user-defined fault buffer. In the user program, add a global data block with timestamps, the last 32 process image states, and the active stop source. Latch it on any rising edge of a stop request. This is the single most useful diagnostic tool when no system-level fault is generated.
3. Trace the V90 telegram. Open Project tree > Devices & Networks > V90 PN > Telegram configuration, install the S7-1500 Motion Control trace, and record STW1 (control word 1), STW2, ZSW1, ZSW2, NSOLL, NIST, and the actual speed. The trace will show exactly which bit flipped to issue the stop.
4. Force DI inputs and observe. In Monitor & Force, force a known DI to TRUE and FALSE. If forcing cannot be performed because the field short clamps the input, the DI module's input LED will stay dark and the force will be ignored by the module.
5. Measure at the terminal block. With a multimeter on V DC at the DI terminal, measure from the input pin to the 0 V rail. A reading of 0.0 V (not 24 V) when the field contact is supposed to be closed confirms a stuck-low condition. Wiggle the cable to localize the chafe point.
6. Insulation test the field cable. With the cable disconnected from the ET 200 module, megger-test conductor-to-conductor and conductor-to-ground at 250 V DC. Resistance below 1 MΩ on a 24 V DI cable indicates degraded insulation.

8. Hardware Solution: DI Modules with Channel Diagnostics

Replace each standard DI module with a channel-diagnostic HF variant. The recommended Siemens module families for ET 200SP and ET 200pro that support channel-granular diagnostics, short-circuit to L+ detection, and short-circuit to ground detection include:

Channel-diagnostic DI modules raise a ChannelError (PROFINET alarm 0x8000 series, channel diagnostics) and a process alarm in OB55 (Status alarm) or OB56 (Update alarm) when they detect a short to L+ or a short to M. The user program can then latch a distinct fault reason: "DI short to ground on I 0.3" — not the generic "equipment stop". With the V90 PN's STW1 trace overlaid with the diagnostic alarm time, the offending input is identified to a single cycle.

For projects where budget is constrained, a hybrid fix is to keep the existing DI module but add a fused distribution block on the sensor 24 V branch (see Section 9) and a software-implemented latching fault buffer.

9. Wiring, Fusing, and Grounding Best Practices

Reference IEC 61131-2 for digital input signal levels and DIN VDE 0100-540 for PE bonding of DC systems.

• DC system grounding: Per IEC 61131-2, the 0 V rail of a positively-switched (PNP) 24 V system may be bonded to PE for EMC reasons. Confirm the bonding point and document it in the circuit diagram. A change in bonding can turn a benign 0 V into a fault path.
• Sensor supply fusing: Insert a 24 V DC fast-acting fuse (e.g., 1 A to 3 A, characteristic gG or quick-acting) on the L+ feed to the sensor distribution block, downstream of the main power supply. This converts a hard short into a controlled voltage collapse, which a channel-diagnostic DI module can detect and which a non-diagnostic module cannot.
• Branch monitoring: Wire the fused 24 V output through an auxiliary contact of the fuse holder back to a DI on the controller. A blown fuse then reports directly into the process image as a "sensor supply lost" alarm, separate from any field-device state.
• Shielded field cabling: Use shielded cable with the shield bonded to PE at the cabinet entry and at the sensor end (or per the sensor manufacturer). For servo axes on the V90 PN, segregate encoder / signal cables from power cables by at least 200 mm or use a grounded metal divider.
• Strain relief and flex rating: At the moving axis, use a flexible cable chain (cable carrier) rated for at least 10 million cycles, with a minimum bend radius of 10× the cable outer diameter. Most short-to-ground events on servo axes are found within 300 mm of the moving connector.

10. Verification and Commissioning

1. Power-cycle test: With the new DI module installed, energize and de-energize the system three times. Confirm that all diagnostic bits return to a defined state and no latched error remains in the V90 or ET 200 module.
2. Controlled short-circuit test: Only if a documented safety case exists: apply a deliberate short between L+ and the 0 V rail at a single field terminal, with the system in a non-production state. Verify that OB55 (or the configured diagnostic OB) fires, the user fault buffer latches the channel number, and the V90 does not transition to an unintended state.
3. High-speed run-up: Run the V90 from 0 to 100% speed across the full mechanical travel for at least 30 minutes. With channel diagnostics enabled, no DI short-to-ground alarm should be raised. Inspect the diagnostic buffer and the V90 message log for any of the following V90 fault codes that can also be triggered by a wiring fault:

Cross-check the V90 fault log against the ET 200 diagnostic buffer time stamps. A consistent correlation between a DI diagnostic event and a V90 motion stop is the smoking gun for the masked DI short.

11. Related Edge Cases and Platform Notes

• Negative-logic (NPN) systems: If the design is NPN, a short between L+ and the input does not mask as a normal OFF; instead, the DI module may see a valid ON regardless of contact state. Investigate the cabinet drawings for the Sourcing/Sinking switch on the DI module, or for the relevant BaseUnit type.
• S7-1200 alternative: If the application runs on an S7-1200 with CM 1243-2 PN Master or with a PN-enabled CPU, the same masking behavior applies. Use the S7-1200 System Manual, available on Siemens Industry Online Support, for the equivalent diagnostic OB (OB82 for diagnostic interrupt) and project it identically.
• V90 with PROFINET telegram 3 or 5: Telegram 3 (speed control, 16-bit) and telegram 5 (DSR with DSC) carry the position controller inside the drive. A masked DI stop issued while the position controller is active can leave the axis in a position-tracking state, so always drive the stop through the same STW1 / STW2 bits the application uses to disable the controller.
• Long-term fleet mitigation: Roll the channel-diagnostic DI module replacement out across all machines that share the same firmware (TIA V17 Update 6 in this case) and same V90 firmware. A version skew between machines is the most common reason one machine faults and an identical twin does not.

12. Quick-Reference Troubleshooting Matrix