Sinumerik 810D Boot Failure: 7-Segment Red Dot and NCK Reset Fix

Problem Summary: 810D Hangs at Red-Dot State With NF/SF Active

A Sinumerik 810D controller on a Hardinge 600 II mill failed to complete power-up. The PCU booted, but the NCK side stalled and the seven-segment display ended up blank with a red decimal point. Status LEDs on the CCU module read +5 (green), NF (red), SF (red), CB (yellow), PR (green), PS (blank), PF (blank), PF0 (blank). On the Windows NT/PCU side, MMC100 reported MMC100 version 53.29.08: waiting for connection indefinitely, never advancing to the operator menu. This article documents the diagnostic interpretation, the NCK reset that recovered the controller, and the field-proven rules that would have prevented the user's secondary data loss.

Indicator Mapping: 7-Segment Display and CCU LEDs

Every 810D CCU ships with a single seven-segment display and a bank of status LEDs. Together they form the lowest-level diagnostic channel. The user-reported LED pattern is the starting point for any subsequent procedure.

The combination NF=red, SF=red, PR=green is informative: the PLC has booted (PR is on) and is communicating on the bus (CB is on), but the NCK has either crashed during init or has been intentionally held in a reset state. The seven-segment readout is the second source of truth. Cross-reference the LED interpretation with the official SINUMERIK 810D Diagnostics Manual for the specific CCU part number installed on your machine.

Boot Sequence Reference: Normal vs Faulted

During a healthy power-up, the 810D walks a fixed state machine and the seven-segment display steps through a sequence of digits. The user's machine emitted 1 → 3 → 7 → 5 → blank-with-red-dot before latching. The exact code-to-phase mapping varies slightly by NCK software release, but the typical progression on SW 4.x / 5.x systems is:

A clean seven-segment count normally ends in the HMI handing off to the operator menu (no digit, no red dot). The fact that the user's unit stepped through 1, 3, 7, 5 and then collapsed to blank + red dot means the NCK reached the end of its initialization, attempted to bind the HMI link, and aborted. From the LED side this is consistent: NF and SF latch the moment the NCK stops without handing off to HMI-RUN. The PLC's PR LED stays on because the PLC310/317 ran to completion independently of the NCK fault.

Root Cause Analysis

On a 810D, a blank-with-red-dot boot fault with NF/SF active and PLC PR alive almost always points to one of three things:

1. Corrupted NCK machine data – a non-volatile memory cell failed and the NCK's integrity check failed, so it never hands off to MMC. This is the single most common cause on a 10+ year old 810D.
2. Battery-low or SRAM failure on the CCU – the buffer batteries that hold SRAM alive for machine data and part programs have aged out. The CCU boots, the NCK detects inconsistent SRAM, and refuses to start.
3. MMC / HMI link failure – the HMI/Windows side does not come up (the "waiting for connection" loop the user observed is an MMC timeout), and the NCK eventually latches the link error.

The user's resolution (an NCK reset) covers case 1 and 2 in one operation: the reset reloads default machine data and rebuilds the SRAM CRC. If the underlying CCU SRAM is hard-faulted, the reset will still come up once, then fail again on the next power-cycle. The user reports the machine is "working again" – which is the expected outcome for a transient NCK data corruption.

Pre-Reset Precautions: Back Up Before You Touch the Switches

Standard backup targets on a 810D are:

Three ways to capture them safely, ordered from easiest to most invasive:

1. RS-232 dump with PCIN or equivalent. The user already had a custom serial transfer tool; PCIN is the classic Siemens utility, while a generic Y-modem/X-modem client can also be used. Set the MMC's RS-232 port under Services → Data In/Out and dump each directory.
2. Network share to a Windows PC. If the HMI runs on Windows NT (PCU 50/70), the CF card or HDD contains the NC directories as plain files. Map a share, copy the directories wholesale.
3. CF card swap. On 810D with PCU 50/70, the CF card on the HMI side holds the active filesystem. Pull it (with the system powered down) and image it with a card reader. This is the highest-fidelity backup because the directory metadata is preserved exactly.

NCK Reset Procedure on the CCU

The NCK reset is initiated from the rotary selector switches on the front of the CCU module. The exact switch labelling varies by CCU version (6FX1111-x, 6FX1120-x), but the procedure is the same in concept. Always cross-check the position with the SINUMERIK 810D Operator Components Manual for your specific CCU part number before applying power.

1. Verify the controller is in a safe state: E-STOP active, axes stationary, drives disabled at the contactor.
2. Remove the cover to access the mode and init rotary switches on the CCU faceplate.
3. Set the mode switch to the NCK reset position. On most 810D CCU revisions this is the position adjacent to the Run position – confirm the labelling on your module.
4. Power-cycle the controller. The 7-segment display will walk the boot sequence and then sit at a steady digit while the NCK clears.
5. Wait at least 30 seconds for the NCK to commit the default machine data to SRAM.
6. Set the mode switch back to the Normal / Run position.
7. Power-cycle again. The MMC should now bind to the NCK and the operator menu should appear within 60–120 s. On the Windows side, the MMC100 version 53.29.08: waiting for connection loop will resolve once the NCK is up.

What an NCK Reset Actually Clears

Restoring from ARC Backups via the HMI

Once the operator menu is reachable, restore the NCK and PLC archives first – they carry the machine data, compensation tables, and PLC project without which the machine cannot move.

1. Insert the backup media (CF card, USB stick, or network share) that contains the _N_ARC_DIR directory.
2. Navigate to Services → Commissioning → Archive In/Out on the MMC100.
3. Select Read in NC archive for the appropriate NCKxxxxx.ARC file. Confirm when prompted.
4. Select Read in PLC archive for the appropriate PLCxxxx.ARC file. The PLC must be in stop mode for this step.
5. After both archives are loaded, perform an NCK restart (Services → NCK Restart).
6. Verify the 7-segment display clears to blank-without-red-dot and NF turns off.

Restoring via RS-232 (PCIN or Equivalent)

If the HMI is unavailable but you can talk to the controller over RS-232, the same archives can be fed in through the serial port. This is the path the user in the case study had on hand with their custom serial utility.

1. Connect a PC running PCIN (or a Y-modem client) to the MMC's COM port on the operator panel.
2. On the HMI: Services → Data In/Out → RS-232, set the baud rate (default 19200, 8N1 on most 810D systems) and the protocol to FE2 / binary for archives.
3. Send NCKxxxxx.ARC first, then PLCxxxx.ARC.
4. NCK restart as above to commit the data.

When the CCU Itself Is Dead: Hardware Replacement Path

An NCK reset only resolves a software/CRC problem. If the CCU's SRAM is hard-faulted (NF returns red on the second power-cycle after the reset), the module is dead. Indicators of a dead CCU:

• 7-segment display never advances past digit 1, 2, or 3 with red dot.
• PS, PF, or PF0 LEDs go red even after a known-good power supply.
• +5 LED dim or flickering.
• Battery voltage below 3.0 V after a known charge cycle (battery harness fault).

Replacement is straightforward: source an equivalent CCU (6FX1111 or 6FX1120 family, same firmware class as the original), transfer the backup NCK/PLC configuration, and reload the ARC archives. If a Siemens service contract is still active, request a CCU exchange with the latest firmware compatible with the original NCK SW. The user in the case study did not need this path – the reset alone brought the machine back – but it is the correct escalation if the reset does not hold.

Field-Proven Commissioning Checklist

1. Document the LED state and the seven-segment sequence with a slow-motion video before any reset attempt.
2. Back up _N_ARC_DIR, _N_CST_DIR, _N_CUS_DIR, _N_MPF_DIR, _N_WKS_DIR, and the tool data directories before the reset.
3. Verify the buffer batteries on the CCU are fresh (≥ 3.0 V open-circuit). Replace them on a 5-year cycle regardless of measured voltage.
4. Confirm the mode switch position against the operator's manual for the specific CCU part number before applying power.
5. Apply the NCK reset, wait 30 s, return switch to normal, power-cycle.
6. Verify NF = off, PR = on, seven-segment blank without red dot, MMC binds to the NCK within 120 s.
7. Reload NCK and PLC archives from the backup set.
8. Restore part programs, user cycles, and tool data from the backup.
9. Jog each axis and verify soft limits, feed override, and E-STOP behaviour before placing the machine in service.

FAQ