Sinumerik 840D NCU 573.2 Booting Error: PLC Watchdog Recovery

Sinumerik 840D NCU 573.2 Booting Error: PLC Watchdog Recovery Procedure

The Sinumerik 840D system with NCU 573.2 occasionally refuses to complete the boot sequence, leaving the operator faceplate (OP) showing a frozen NCK identification and the seven-segment display on the NCU locked. The dominant root causes are loss of PLC communication (alarm 120202 "Waiting for connection to NCK"), PLC watchdog trips that force the PLC into STOP, and bus faults on the MPI/Profibus segment that disconnect the PLC from the NCK. This document consolidates the field-proven recovery procedure, bus diagnostics, and PLC diagnostic-buffer interpretation required to bring the NCU back online without losing the machine's PLC project.

1. Problem Profile and Failure Mode

An NCU 573.2 is the compact numerical control unit used in the Sinumerik 840D sl product line. It contains the NCK (numerical control kernel), the integrated PLC (CPU 317-2DP equivalent on later firmware), the CF card holding the PLC project and NC data, and the MPI/Profibus interfaces used to talk to the OP, the I/O slaves, and the drive bus (PROFIBUS-DP). When the NCK cannot establish the handshake with the PLC, the entire boot sequence stalls. Typical symptoms reported in the field are:

• NCU seven-segment display stuck on 6 or 2 rather than progressing through the normal 1 → 2 → 3 → 5 → 6 handshake states.
• PS LED permanently ON, PF LED permanently ON (PLC fault), or PS LED slow-flashing (PLC in STOP with pending error).
• OP displays Alarm 120202 "Wait for connection to NCK" or "No communication to PLC".
• PLC diagnostic buffer entry: "PLC watchdog error detected, PLC is in STOP mode".
• Spontaneous reset of I/O stations, intermittent Profibus slave dropouts, or field wiring short circuits on 24 V sensor/actuator rails returning through the PLC inputs.

2. NCU 573.2 Hardware Reference and LED Map

The NCU 573.2 front panel exposes a fixed set of indicators that communicate boot progress and PLC state. The first step in any recovery is to read these correctly.

The rotary switch S4 on the NCU front face is the master mode selector. It is read by the NCK firmware at boot and overrides the project setting in mmc.ini for the current power cycle only.

3. NCU Boot Stages (Seven-Segment Decoder)

The 7-segment display is the most reliable indicator of where the boot has stalled. Capture the exact digit before cycling power, because it points directly to the failing sub-system.

A stall at display 6 with the PS LED slow-flashing and PF LED ON is the canonical signature of the NCU 573.2 PLC watchdog condition described in the field report.

4. Recovery Procedure: S4 General Reset to Recover from PLC Watchdog

The following procedure is the minimum-impact path to clear a PLC watchdog condition on an NCU 573.2. It wipes the PLC work memory and forces a clean reload of the PLC project from the CF card. NC machine data and tool/zero-point data are not erased. Always perform Steps 1-2 from Section 5 (preflight) before executing this sequence.

4.1 Confirm the PLC project is intact on the CF card

1. Power down the NCU. Remove the CF card (slot on the front, label-up). Read it with a Siemens card reader or any industrial CF-to-SATA adapter in a PG running SIMATIC Manager / TIA Portal.
2. Verify the following files are present and non-zero: /siemens/sinumerik/nck/dsfb.dir/, /siemens/sinumerik/plc/, and on later builds /card/user/sinumerik/.
3. If the PLC folder is missing or unreadable, the recovery will fail and the NCU will return to display 3 indefinitely. In that case, restore the PLC project from a SIMATIC Manager backup (.s7p archive) via PG with MPI before continuing.

4.2 Power down, isolate, baseline

1. Bring the machine to a safe state — axes stopped, drives inhibited, hydraulic/pneumatic dump valves closed per the machine builder's lockout procedure.
2. Open the main disconnect and verify zero potential on the 24 V DC NCU supply with a calibrated DMM.
3. Note S4 position. Standard production state is 0.

4.3 Bussed-cable isolation test (eliminate the field as the cause)

1. With the NCU still powered down, disconnect every plug-in connector from the NCU front face: X1 (24 V), X2 (MPI/Profibus OP), X3 (Profibus-DP to drives), X4 (I/O bus), plus any encoder and Ethernet ports.
2. Power the NCU from a clean 24 V supply only. Observe the 7-segment and LEDs.
3. If the NCU completes boot to display 7 or runs normally with all external cables removed, the fault is on the field side (Profibus slave, I/O short, OP cable, or 24 V rail). Reconnect one segment at a time to isolate the offender.
4. If the NCU still fails to boot in isolation, proceed with the S4 procedure — the PLC project itself is the suspect.

4.4 S4 General Reset Sequence

1. NCU powered down. Set S4 to position 3.
2. Apply 24 V power. The NCU will perform a general PLC reset. Expected LED pattern: PF LED ON briefly, then PS LED slow-flashing. The 7-segment will cycle through 1, 2, 3, 5, 6 and may pause at 6 while the PLC reloads.
3. Wait until only the PS LED is ON (steady) — this confirms the PLC has finished loading its project from the CF card and entered RUN. Typical reload time: 30-90 seconds for a medium-sized project.
4. Power the NCU down again. Set S4 back to 0 (or 1 if you want PG-visible debug).
5. Apply 24 V power. Confirm normal boot: 7-segment reaches 7 and clears, OP shows the channel-ready screen with no active alarms.

5. Preflight Checklist and Prerequisites

Before touching S4, complete the preflight below. Skipping it is the single most common cause of a "recovery" turning into a total PLC loss.

1. PLC backup verified. A SIMATIC Manager project archive exists on a network share or USB, dated after the last commissioning change.
2. NC archive available. A series-commissioning archive (.arc) created with Commissioning → Series commissioning → Create archive is on the OP or a USB stick. This preserves options, NC machine data, and tool data if a full NCK reset is later required.
3. PG/PC with SIMATIC Manager or TIA Portal ready, with the MPI cable (6ES7901-0BF00-0AA0 or equivalent) and the correct PLC authorization/license.
4. Machine safe state. All axes referenced or known to be unreferenced; hydraulic pressure relieved; tool changer in a safe index; spindle stopped.
5. Bus topology map in hand. Profibus slave addresses, expected bus termination points, and cable routing. Without this map, a reconnection in Section 4.3 is guesswork.

6. Network and Bus Diagnostics (MPI / Profibus)

A PLC watchdog on the NCU 573.2 is often a symptom, not the root cause. The PLC goes into STOP because OB1 stopped being called, and OB1 stops being called because the Profibus-DP I/O scan failed or because a 24 V short on an input module pulled the PLC into a diagnostic interrupt loop. The two bus segments that must be validated are:

6.1 MPI bus (NCU ↔ OP ↔ PG)

The MPI port on the NCU is X2 (9-pin D-sub, female). Default addresses: NCU = 2, OP = 1, PG = 0. The bus must be terminated at both physical ends with 220 Ω, and the cable shield must be bonded to the connector shell at every node.

6.2 Profibus-DP bus (NCU ↔ I/O slaves ↔ drives)

NCU 573.2 has two Profibus ports on X3 and X4. The drive bus (typically X3) and the I/O bus (typically X4) must each be terminated exactly once at each physical end, with the terminator powered (9-pin D-sub with the switch in the ON position and the 5 V supply available — the NCU supplies this on its own connector). Slave addresses must be unique; a duplicate address forces the entire bus down.

7. PLC Diagnostic Buffer Interpretation

Once the NCU is back online enough to accept a PG connection (or the PLC can be reached via MPI at 187.5 kbit/s), open the PLC online view in SIMATIC Manager and read the diagnostic buffer. This is the definitive record of why the watchdog fired.

Save the diagnostic buffer to a plain-text file using the SIMATIC Manager "Save as text" button. That file is the single most valuable artifact for Siemens support escalation if the recovery does not hold on subsequent boot.

8. Verification Checklist After Recovery

Run the following before returning the machine to production. Any failure here indicates the underlying cause has not been resolved and the watchdog will return on the next shift.

1. NCU 7-segment reaches 7 and clears. PS LED steady, PF LED off, BF LED off.
2. OP shows channel-ready state with all axes "Available". No alarm 120202, 120201, 2001, or 380001 in the active alarm list.
3. From the OP, navigate to Diagnostics → Bus diagnostics and confirm all expected Profibus slaves are present with green status, no red "station failure" entries, and error counters at 0.
4. Reference all axes and run a controlled axis move of at least 10 mm in each direction from the OP handwheel, confirming actual position feedback tracks the commanded position within the loop tolerance.
5. Toggle a representative sample of digital inputs (limit switches, door interlocks, lube OK) and confirm the corresponding PLC input indicator in SIMATIC Manager reflects the change within one OB1 cycle.
6. Verify PLC time-of-day is correct; if not, the NCU buffer battery is exhausted and the next power-down will lose retentive flags.
7. Trigger a controlled NCK reset (via OP Reset button) and confirm clean re-initialization, no PLC re-entering STOP.

9. Prevention and Hardening Recommendations

• Schedule a periodic PLC backup. Use the HMI's Commissioning → Series commissioning function to create a dated archive on a network share after every PLC or NCK change. The archive should include the SIMATIC project, the NC archives, and the drive configuration files.
• Replace the NCU buffer battery on a calendar basis. 5-year maximum service life, replace at 4 years in high-availability cells.
• Audit Profibus terminations annually. Thermal cycling loosens D-sub screws; 0.4 Nm torque check takes ten minutes per segment.
• Install RC snubbers on every 24 V DC contactor coil driving machine safety circuits. A single un-snubbed coil is sufficient to take down the Profibus segment when it de-energizes.
• Keep the CF card write-protected in production. A failing CF card with random bit flips in the PLC project is the most insidious cause of intermittent watchdog faults that look like network issues.

10. Escalation Path

If the S4 general-reset procedure has been performed twice on a freshly restored PLC project from a known-good backup and the watchdog condition returns within one shift, the fault is environmental: a defective I/O module, a damaged Profibus cable, or a 24 V rail instability. The next steps are:

1. Capture the PLC diagnostic buffer in text form at the moment of the next watchdog event.
2. Capture the OP alarm log as a CSV via the Commissioning → Logs menu.
3. Capture the NCU LED state, 7-segment display, and the Profibus slave status from Diagnostics → Bus diagnostics.
4. Open a Siemens support request (Siemens Industry Online Support, support.industry.siemens.com) with all three artifacts attached and the NCU serial number, firmware version (read from Commissioning → NC → NCK version), and the CF card part number.