Troubleshooting Sinumerik 810D Powerline NCK 5-2-1 Boot Loop

1. Problem Overview

The SINUMERIK 810D powerline controller, when used in a ShopMill-equipped three-axis mill such as the XYZ560 (manufactured circa 2004), is susceptible to a specific class of boot failure that manifests as either of the following observed states:

1. State A (Cold-boot failure): The HMI (MMC 102/HMI Advanced) reaches the splash screen and freezes on the message "Waiting for NCK connection: #.# seconds". The seven-segment display on the front of the CCU (also called NCU in 810D powerline nomenclature) stabilises on "4" rather than advancing to "6" (ready). The wait timer increments indefinitely; the HMI never establishes a logical connection to the NCK.
2. State B (Post-clear failure): After an operator performs the NC and PLC Clear & Reload procedure described in the Siemens field service document "810D NC PLC Clear & Reload Procedure", the seven-segment display does not stabilise on 6. Instead, it cycles 5 - 2 - 1 - 5 - 2 - 1 indefinitely (also reported as 1-2-5-1-2-5 by some observers; both describe the same three-stage loop). A red LED on the CCU is typically off, which means a hardware "hard stop" alarm is not raised, but the NCK never reaches a running state.

2. Affected Hardware and Software Identification

Before any intervention, capture the exact part numbers and software versions displayed during the HMI splash. These determine the correct backup image, the correct PCU firmware, and the correct PLC toolchain. The reference machine reported in the source case shows the following stack:

For functional scope, the official Siemens manual Milling with ShopMill - SINUMERIK 810D/840D powerline (PDF, 6FC5 297-0AE70-0BP0, edition 07.07) documents the ShopMill 6.4 cycle set and confirms that the 810D powerline can host ShopMill versions 6.x. Knowing the exact installed ShopMill build is essential when reloading from a backup because ShopMill 6.2 is not bit-compatible with ShopMill 6.4 cycle libraries.

3. Seven-Segment Display Code Reference

The seven-segment display on the CCU/NCU front panel is the lowest-level diagnostic indicator. The single digit reflects the current state of the firmware boot loader and the NCK. A normal cold boot displays each digit in sequence for a fraction of a second before settling on 6. A 1-2-5-1-2-5 loop or a 5-2-1-5-2-1 loop indicates a state that the firmware cannot exit.

Interpretation of the observed loops:

• 1-2-5-1-2-5 loop — the CCU POSTs successfully (1), reads the PCMCIA card directory (2), and starts loading the HMI/OS image from the card (5). It then restarts. The NCK image is not the failure point; the HMI image or the card itself is.
• 5-2-1 loop — same root cause from a different observation angle. Step 5 cannot complete, so the loader reverts to step 1 and re-attempts the full boot.
• Stuck on "4" with HMI waiting — the CCU has loaded the NCK and is broadcasting, but the HMI's MMC 102/PCU20 cannot establish the named-pipe or TCP link to the NCK. Causes include: wrong HMI version on the PCMCIA, wrong NCK version on the CCU, mismatched PLC, MMC boot block damaged, or the PCMCIA card's HMI directory being corrupt.

4. Root Cause Analysis

Combining the two observed states yields a small set of root causes. Rank them by probability based on the symptom set.

5. Pre-Procedure Safety and Prerequisites

Prerequisites before the Clear & Reload procedure:

1. A working image of the system PCMCIA card for the specific controller. For an 810D powerline with ShopMill 6.2 and NCK boot V01.03.04, the card must contain the matching NCK directory tree (/nc), the HMI directory tree (/mmc2), and the PLC archive (/plc). The card must be formatted FAT16 with the Siemens-specific SYSTEM attribute set on the boot files.
2. A blank or freshly formatted PCMCIA card of at least 16 MB (the 810D powerline ShopMill 6.2 image is approximately 8 MB compressed).
3. A PG (programming device) with STEP 7 V5.x and the matching ShopMill commissioning tool installed, plus a serial / MPI cable for PLC debug access.
4. The machine's commissioning archive (NC machine data, compensation tables, tool magazine table, PLC program backup) — the original startup file set supplied with the machine at delivery.
5. The exact MLFB of the CCU and the PCU20 firmware on a label. For PCU20 the MLFB is in the form 6FC5210-0xxxxx-xxxx; the source case has 6FC5210-0DF00-0AA0.

6. NC and PLC Clear Procedure (Detailed)

The Clear & Reload procedure wipes the NCK static RAM (machine data, programs, tool data) and the PLC work memory. It does not touch the PCMCIA card. Run it only when you have a confirmed-good image to reload.

1. Confirm S3 (rotary mode switch on the CCU front) is at position 0 and S4 is at position 0 before applying power. Power the cabinet on.
2. Wait for the CCU POST to complete. The seven-segment will sequence 1, 2, 3, 4, 5. If it settles on 6, the system is up; the PLC clear is performed by setting the switch and resetting, not by a full power cycle.
3. To clear the PLC only: set S3 to position 2, press and hold the Reset pushbutton on the CCU front for 5 s, then release. Watch the seven-segment; it should briefly display 6 with a fast-blinking PLC LED, indicating the PLC has been cleared. Return S3 to 0.
4. To clear the NCK only: set S3 to position 1, power on, wait for the seven-segment to reach 6, then press Reset. Return S3 to 0.
5. To perform a full clear (NCK + PLC), set S3 to 1, power on, wait for 6, press Reset, then cycle S3 to 2 and press Reset again. The seven-segment must reach 6 at each stage; if it does not, the CCU has not finished the previous step and the clear has not happened.

7. PCMCIA Card and Backup Restoration

The PCMCIA card (PC card, Type II, 5 V) holds the system image. A common failure mode is oxide on the contacts after long storage, or a soft sector failure after the card was last written months or years ago.

7.1 Card inspection and re-seat

1. With the cabinet de-energised and locked out, open the CCU cover and press the ejector to remove the system card.
2. Inspect the contact edge for oxidation, dust, or bent pins. Wipe with isopropyl alcohol (IPA, 99 %) and a lint-free swab. Allow to dry fully before reinserting.
3. Reseat the card firmly. Power on and observe the seven-segment sequence.

7.2 Card image verification on a PC

1. Read the card with a PCMCIA-to-USB or PCMCIA-to-IDE adapter on a Windows 98 / 2000 / XP machine (Linux with dd also works for raw image capture).
2. Mount the FAT16 partition and run chkdsk /f to repair soft file-system errors.
3. Verify the presence and size of these critical files:
   • mmc2\hmi.exe or mmc2\hmi_adv.exe — should be several MB.
   • mmc2\shopmill.lng or matching language file.
   • nc\nck.bin or nc\*.arc — the NCK runtime image.
   • plc\*.awl or plc\*.s7p — the PLC source archive.
4. Compare file modification dates with the original commissioning record. A date in the 1990s is a strong indicator that the card has not been re-flashed since delivery and is the most likely source of corruption.

7.3 Card re-imaging from a known-good backup

1. Use a card image captured at commissioning time, or download the matching image from the Siemens support portal (entry by MLFB and the controller's CF/PN order number). The image is delivered as a ZIP containing a raw card image (.img) or as individual files in the directory layout above.
2. Write the image to a new PCMCIA card using the PCMCIA-to-IDE/USB adapter. Use a card of the same capacity and type as the original.
3. Mark the new card with the controller's MLFB and the date.
4. Insert into the CCU, power on, and observe the seven-segment. A healthy card brings the system to 6 in under 90 seconds.

8. CCU/NCU Hardware Diagnostics

If a freshly imaged, freshly seated card does not bring the CCU to 6, the CCU itself is suspect. The Siemens 810D powerline CCU contains the NCK, the PLC CPU, the PROFIBUS master, and the buffered SRAM. The following steps isolate the failure.

1. Battery check — measure the 3 V cell on a powered-down CCU after 5 minutes. A voltage below 2.5 V indicates the cell has failed. The source case has already had the cell replaced; verify the new cell's voltage and the polarity. Reverse polarity destroys the buffered RAM interface IC.
2. Power-rail check — measure the 5 V and 3.3 V rails on the CCU test points. The 5 V rail supplies the digital logic; ripple above 50 mV pk-pk indicates a failing bulk capacitor on the backplane.
3. POST observation — with the card removed, power the CCU. The seven-segment should display 1 and remain there (no card means no stage 2). If the CCU does not reach 1, the CCU POST is failing → hardware fault, escalate to Siemens service.
4. Single-card test — try a different, known-good PCMCIA card from a sister machine of the same age and software stack. If the second card also produces 1-2-5, the CCU is the problem.
5. CCU module swap — if a spare CCU is available, swap it. The replacement must be loaded with the same NCK version (V01.03.04 in this case) and the same machine data file. Note that swapping the CCU breaks the calibration if the position encoder offsets are not transferred; a re-reference of all axes is required.

9. Battery (RAM Buffer) Replacement Procedure

The CCU battery maintains the buffered SRAM that holds the NC machine data, the tool table, the PLC retentive flags, and the part-program work memory. When the battery dies, the CCU boots with default machine data on the next power-up, which frequently causes alarms that prevent the NCK from advancing to stage 6.

1. Power the cabinet on and allow the CCU to reach state 6 (or as far as it will go).
2. Locate the lithium cell holder on the CCU PCB. The cell is a 3 V BR2032 or CR2032, with the positive terminal up.
3. Using a non-conductive tool, lever the cell out of the holder. Wait 30 seconds; this allows the buffered SRAM supervisor to recognise the battery loss.
4. Insert the new cell, positive terminal up, in a single smooth motion. The supervisor will report Battery alarm on the next HMI splash, which is cleared by acknowledging the alarm with the HMI's Cancel key followed by the NCK reset.
5. Reload the NC machine data from the commissioning archive. If the archive is not available, regenerate it by hand from the machine's mechanical specification (ball-screw pitch, encoder resolution, axis travel limits, spindle ratios).

10. NCK Connection Wait State Recovery

When the HMI displays "Waiting for NCK connection: #.# seconds" and the seven-segment shows 4, the NCK is running but the HMI cannot bind to the NCK's communications port. Recovery:

1. Confirm the seven-segment really is 4 (not 1, 2, or 5). Use a torch; the digit is small.
2. Wait at least 60 seconds; the NCK may be performing the initial cyclic IPO synchronisation, which can take up to 45 s on a cold start.
3. If the wait does not clear, the HMI's NCK-binding file is suspect. On the PCMCIA card, edit mmc2\netnames.ini and mmc2\dhcpd.cfg (or the equivalent for ShopMill 6.2) to confirm the NCK address matches the CCU. The default is 192.168.214.1 for the NCK and 192.168.214.241 for the PCU.
4. Reboot the PCU only (not the CCU) by pressing the PCU reset button behind the operator panel door. The HMI re-launches and re-binds.
5. If the bind still fails, the NCK's static machine data block may have been corrupted by the dead battery. Re-load the machine data per the commissioning archive (see section 5).

11. Verification and Commissioning Checks

After the controller reaches state 6, perform the following sequence before returning the machine to production.

12. Field Notes and Escalation Path

• Don't trust the LED colour alone. A green LED on the PCU only means the PCU BIOS has finished POST; the HMI is a Windows application that can still be hung. The seven-segment on the CCU is the ground truth.
• Don't repeat the Clear & Reload procedure on a 1-2-5 loop. The clear cannot complete, and repeated toggling of S3 can leave the CCU in a state that requires a service bench to recover.
• Capture the splash text before doing anything else. The text is overwritten as soon as the HMI starts. Photograph the screen on the first 10 seconds of every power-up.
• Keep a clone of the original PCMCIA card on file. A 16 MB card is a small price for a guaranteed recovery path.
• Document the CCU's NCK and PLC versions on the cabinet door. When a fault recurs years later, the on-cabinet note saves the next engineer's first 30 minutes.
• Escalation: When the 1-2-5 loop persists with a confirmed-good, freshly imaged card, escalate to your local Siemens service centre with the CCU MLFB, the PCU20 MLFB, the NCK version, the ShopMill version, and the PLC version. A CCU repair / exchange is the most likely outcome. The Siemens support portal is the authoritative entry point for the latest service notes and spare-part availability: support.industry.siemens.com.