SINUMERIK 840D Boot Failure: Waiting for Connection Diagnosis

1. Problem Statement and Field Context

On a Spinner VC610 (and the broader family of SINUMERIK 840D / 840D sl controlled machine tools, including the 810D, 840Di and FM-NC variants), the operator or service engineer switches on the main disconnect, the SINUMERIK operator panel (OP / PCU 50.3 / PCU 50.5) begins to start, the splash screen of the HMI appears briefly, and the system then stops at the message Waiting for connection to NCK or simply Waiting for connection on a black or grey background. The status bar of the HMI no longer advances, the machine control panel (MCP) LEDs begin to flash in unison, and the machine can no longer be referenced, jogged or otherwise operated.

This particular symptom is one of the most common cold-start failures reported on retrofitted and aged 840D systems. It is not a single fault but a class of faults that share a common observable: the HMI successfully starts the operating system layer of the PCU / TCU, but the HMI software on the PCU cannot establish the TCP/UDP or MPI handshake with the NCK (Numerical Control Kernel) and the PLC inside the NCU / CCU. Until that handshake completes, the HMI cannot load the channel menu, alarm log, axis views, tool list or any of the user-interface pages that the operator depends on.

2. SINUMERIK 840D System Architecture Relevant to the Fault

To diagnose a boot hang it is essential to understand the bus topology that the HMI is trying to connect to. A representative SINUMERIK 840D sl installation has the following logical layers:

3. NCU / CCU Boot Sequence and Where It Hangs

When 24 V is applied to a healthy NCU, the following sequence executes. The operator sees progress in three discrete steps on the front-panel LED row:

1. Hardware power-up (0–2 s): +5 V regulator starts, +5 V LED goes green. Reset is asserted to the PLC and NCK CPUs. PR (Processor) LED is OFF.
2. Firmware / BIOS hand-off (2–10 s): Boot loader reads from CFast or PC card. NF (NCK Fail) and SF (System Fault) LEDs briefly go red and then clear. PR LED goes green.
3. NCK + PLC initialization (10–30 s): NCK reads the machine data, axis configuration, drive assignments; PLC executes OB 100 cold-restart and OB 1 cycle. PS (PLC Status) and PF (Peripheral Fault) LEDs indicate PLC state. CB (Communication Bus) LED comes on as PROFIBUS is token-passed.
4. Wait for HMI handshake (30 s – 5 min): NCK opens its listening port (default 192.168.214.1 on the X120 system network, or MPI address 0/14) and waits for a client. The PCU HMI tries to register. If no client connects within the configured timeout, the HMI displays Waiting for connection to NCK.

When the HMI stops at step 4, the NCU has either:

• Failed to reach step 4 at all (NCK not running, SF / NF stuck red), or
• Reached step 4 but never received the HMI registration packet (network or address mismatch), or
• Reached step 4 but lost PROFIBUS communication with the drives or the ET200, which causes the NCK to refuse HMI hand-shake completion.

4. Decoding "Waiting for Connection" — HMI Timeout vs. NCU Hang

Before opening the cabinet, classify which side of the connection is the offender. The combination of PCU display, MCP LED behavior, and NCU 7-segment code is the fastest triage:

5. NCU / CCU LED Indicator Diagnostics

The eight LED block on the rear of every NCU 710–760 (and on the front of legacy CCU 1 / CCU 3.4) is the single most informative component. The reference LED set is:

The exact LED reported on the Spinner VC610 in the source is:

+5V = Green
PR   = Green
NF   = OFF
PS   = Red
SF   = Red
PF   = Red
CB   = OFF
PFO  = Amber

This is a diagnostic fingerprint. +5 V and PR green confirm that the NCK CPU has been released from reset and is executing firmware. NF OFF means NCK has not reported a hard kernel crash. However, the triplet of PS red, SF red, PF red, CB OFF tells a very specific story: the NCK has started, but the PLC is in STOP (PS red) and the peripheral / system faults indicate that one or more PROFIBUS slaves (drives or ET200) cannot be reached. The HMI therefore waits for a PLC that is in STOP, and the cyclic PLC process image never comes up.

6. S3 and S4 Rotary Switch Functions

On NCU 710.1 / 720.1 / 730.1 and on legacy CCU 1 / CCU 3 the two hexadecimal rotary switches S3 and S4 behind the hinged cover set the operating mode of the NCK and the PLC at power-up. Their normal-operation positions on a standard 840D sl are:

On a more modern NCU 730.2 / NCU 760 the same two switches exist physically but their primary use is to enter the commissioning / series-startup mode. Position 1 on S3 is the explicit "NCK in commissioning mode" setting and Position 1 on S4 is the explicit "PLC in commissioning mode" setting. Position 0 on both switches is the correct setting for normal operation — the original post from the service engineer correctly observed that S3 = S4 = 0 is normal. The concern that this might have "erased data" is a misunderstanding; the destructive settings are 6 and 7 on the legacy CCU family. Verify, do not touch, the switches until you have a confirmed good NCK / PLC archive on external media.

7. MPI / PROFIBUS Cable and Connector Checks

When NF / SF / PF are red and CB is OFF, a physical-layer bus problem is the most probable cause. Inspect the PROFIBUS cables between NCU X141 (drive bus, to 611D bus connector X161) and X142 (I/O bus, to ET200 / IM361). Use the following checklist:

1. Confirm that the bus connector is the Siemens 6GK1500-0FC10 (with PG socket) or 6GK1500-0EA02 (without PG socket). Third-party connectors often leave the terminating resistor in the wrong position.
2. Check that the terminating resistor is ON only at the two physical ends of the bus. On a 611D system the terminating switch is on the last drive module and on the NCU X141. All other modules must be set to OFF.
3. Measure the bus shield at the cabinet entry. The shield must be 360° clamped to a copper ground bar at the point of cabinet entry; a "pigtail" shield termination is a known cause of intermittent NF / SF errors that look like a soft boot hang.
4. With a Siemens BT200 bus tester (or an oscilloscope on channel A vs. channel B) verify the differential voltage between pins 3 (RxD/TxD-P) and 8 (RxD/TxD-N) is > 4 V at every node when no traffic is on the bus, and > 1 V during token passing.
5. Disconnect the drive bus at the NCU and observe whether the HMI can establish the connection. If the HMI now progresses past "waiting", the fault is on the drive bus and you should suspect a faulty 611D module, broken fiber-optic, or a defective closed-loop controller card.

8. Series Startup with PCIN and a Floppy / PC Card Image

When the CFast / PC card in the NCU is suspected of holding a corrupted boot image but the user has a known-good series startup file on floppy (typical for older 840D) or on a PC card, a series startup is the recovery path. The procedure requires access to the ASCII-mode loader on the NCU, which is normally entered by setting S3 = 1 and S4 = 1 (or by holding a service key combination depending on the firmware version), then connecting a PC running PCIN or SIMATIC S7-Transfer via the X122 RS-232 service port or via the X120 Ethernet service port.

1. Power down the machine and isolate at the main disconnect. Lock-out / tag-out the cabinet per local safety rules.
2. Insert the original series-startup floppy / PC card into the PCU or an external USB floppy drive connected to the PCU.
3. Set S3 = 1 and S4 = 1 on the NCU. Apply 24 V. The NCU now boots in commissioning mode. The +5 V and PR LEDs will be green; NF will pulse red briefly; the seven-segment display will show "0" or "1".
4. On the PCU open PCIN (Start → Programs → SINUMERIK → Tools → PCIN). Configure the protocol: Protocol = Binary / Text, Baud = 19200, Parity = Even, Stop bits = 1, Data bits = 7 for the legacy RS-232 path, or configure Ethernet to 192.168.214.10 / 24 for the X120 system network.
5. From the PCU, transfer the series-startup archive to the NCU using PCIN's Send Data function. The NCK writes the file to its internal Flash, then reboots automatically.
6. Return S3 and S4 to 0. Power-cycle the NCU. The HMI should now reach the channel menu within 60–120 seconds.

9. NC and PLC Archive Recovery Procedure

If a full archive backup exists (the typical *.arc file produced by Commissioning → Archive in HMI Advanced / HMI sl, or the older *.teaching format) the cleanest recovery path is:

1. Connect the PCU to the NCU over the X120 system network (default IP 192.168.214.1 on the NCU, 192.168.214.10 on the PCU).
2. Boot the NCU. If the HMI cannot reach the NCK, use the PCU Service Center shortcut (Ctrl + Alt + S on HMI Advanced, or via the CFast image boot menu on HMI sl) to access the commissioning menus directly.
3. From the Commissioning menu choose NC area → Clear NC to wipe the corrupted NV data, then choose Archive → Read-in and select the saved *.arc file.
4. Power-cycle. The NCK will re-initialize, the PLC will re-load from internal Flash. The HMI will be able to complete the handshake.
5. If the PLC is in STOP, restart the PLC from Commissioning → PLC → PLC STOP / RUN or, in legacy CCU 1, toggle S3 to 0–1–0.

10. Power Supply and Grounding Verification

Under-voltage is a common root cause of intermittent "waiting for connection" faults. Verify:

• 24 V DC at the NCU input is between 21.0 V and 28.8 V (Siemens tolerance band) under full load (all drives enabled). If the rail sags below 20.5 V at the NCU terminals during brake release, the NCK will reset cyclically and the HMI will see a flapping handshake.
• The NCU PE terminal is bonded to the cabinet PE bar with a short, fat conductor (≤ 100 mm, ≥ 6 mm²). A long ground lead allows the drive switching noise to inject into the NCK reset logic.
• The 24 V return (−) is bonded to PE at exactly one point, normally at the SITOP supply. A second bond creates ground loops that produce the same NF / SF red pattern.
• The MCP 24 V is sourced from the same 24 V rail that feeds the NCU, not from a separate branch. The MCP and NCU share a common reference; if the MCP sits on a different ground, the PLC cannot handshake with it and the MCP LEDs flash in unison — exactly as observed on the Spinner VC610.

11. Verification and Commissioning Tests

After the boot hang is cleared, run the following checks before returning the machine to production:

12. Frequently Asked Questions