Troubleshooting the SINUMERIK 840D NCU 720.1 0 / C 7-Segment Boot Loop

A SINUMERIK 840D sl controller built around the NCU 720.1 (Siemens article number 6FC5371-0AA00-0AA0 family) is normally a transparent, well-telemetered machine. When the unit refuses to complete its boot sequence and instead toggles the 7-segment indicator between 0 and c, while every status LED on the front bezel glows solid orange, the controller has stalled in the HMI / NC handshake phase and will not publish an HMI connection to the operator panel. The fault is recoverable in most cases, but only if you work the recovery tree in the correct order. This guide consolidates the on-machine evidence, the Siemens rotary-switch clearing procedure, the CF-card service boot, the PLC restart logic, and the hardware checks you must run before condemning the NCU.

1. Symptom Definition

The reported condition is reproducible and is the canonical signature of a stalled NCU 720.1 boot:

• Every front-panel LED on the NCU illuminates orange simultaneously and stays lit (no red fault LED is asserted; no green run LED is reached).
• The 7-segment display oscillates between 0 and c at roughly 1 Hz. The indicator never advances to 1, 2, 5, 6, or any run-state code.
• The HMI client on the operator panel (PCU / TCU) loses the connection to the NCU and displays "no connection to controller", "HMI not responding", or "NCK not reachable".
• Power-cycling the machine produces the same indicator pattern within 30-90 seconds of cold start.
• The buffer battery measures a healthy ~3.0 V DC across the battery compartment on the NCU, and the NCU fan ramps immediately at power-on.

2. NCU 720.1 Hardware Reference

Pin the hardware variant down before you start clearing things. The NCU 720.1 is a single-board CNC in the SINUMERIK 840D sl line that integrates the NCK, the PLC (S7-300 compatible CPU 317-2DP/PN), the HMI proxy and the CF-card mass storage. The front bezel carries the rotary mode switches, the 7-segment indicator, and the LED stack.

3. Decoding the 0 / c Indicator

SINUMERIK 840D sl uses the 7-segment indicator as a single-letter/single-digit state machine. The exact meaning of each code depends on the installed firmware (NCK, PLC, HMI versions are all independent). The most common mappings for the codes visible during a stuck boot are:

The alternating 0 / c display is a near-certain sign that the NCU is cycling between the HMI startup phase (0) and the PLC configuration phase (c) without successfully completing either. Each time the HMI proxy times out waiting for the PLC, the controller restarts the PLC configuration phase and the display flips back to 0 while the HMI subsystem retries. Two to four seconds between flips is the typical cadence on a machine with this fault.

4. Root-Cause Matrix

Match the symptom to the most likely cause and the cheapest fix first.

5. Prerequisites Before You Touch the NCU

1. Obtain a backup of the CF card using the service menu or by removing the card and cloning it on a Linux PC with dd. Even if the card looks dead, an fsck.ext3 -n image can save the project.
2. Locate the machine manufacturer's commissioning archive: the original SINUMERIK archive (.arc), the original CF card image, or both.
3. Have a known-good replacement CF card of the same type (typically a 4 GB or 8 GB industrial-grade card, Siemens 6FC5447-0AA00-0AA0 family) loaded with the matching HMI/NCK versions.
4. Confirm the buffer battery is fresh. If the battery measures below 2.7 V under load, replace it with a Siemens 6FC5247-0AA18-0AA0 (3 V, 850 mAh, lithium) before proceeding; a marginal battery will hide a successful recovery.
5. Note the current rotary-switch positions for S3 (NC) and S4 (PLC) before turning any switch. Both should be on 0 for normal operation.
6. Verify the 24 V DC cabinet supply holds within ±5% under load. Brown-outs look identical to a software fault.

6. Recovery Procedure A — Rotary-Switch NC/PLC Clear

The first action is the cheapest: clear the NC and PLC using the rotary switches on the front bezel. This procedure does not erase the project on the CF card; it only resets the volatile NC and PLC state.

1. Power down the machine at the main disconnect and lock it out.
2. Open the cabinet door and locate the NCU 720.1 in the operator panel.
3. Turn the NC rotary switch S3 from position 0 to position 1 (NC general reset / "clear NC").
4. Turn the PLC rotary switch S4 from position 0 to position 3 (PLC memory reset / MRES).
5. Restore power to the machine. The 7-segment display will show the boot codes.
6. Wait at least 90 seconds without interrupting. Do not turn either rotary switch until the display settles or until you have observed three complete 0 ↔ c cycles.
7. If the controller still oscillates 0 ↔ c, perform a PLC memory reset properly: hold S4 in 3 for 3 s, release to 2, hold 3 s again, then return to 0. The PLC will perform a full MRES, which erases the work memory and forces a re-download of the PLC project from the CF card.
8. After the PLC MRES, return S3 to position 0. The NC will request a re-load of NCK data from the CF card.
9. If the display now settles at 6, recovery is complete; proceed to Section 11 (Verification).
10. If the display continues to oscillate 0 ↔ c, leave the machine powered and proceed to Recovery Procedure B.

7. Recovery Procedure B — CF-Card Reseat and Service Boot

If the rotary-switch clear did not help, the next most likely cause is the CF card. Power the controller down before touching the card.

1. Power down at the main disconnect; wait for the NCU capacitors to discharge (a minimum of 30 s).
2. Press the CF-card ejector on the NCU front panel and remove the card.
3. Visually inspect the card edge and the socket for contamination, oxidation, or bent pins. Clean with isopropyl alcohol and a lint-free swab if needed.
4. Re-insert the card, ensuring it is fully seated and the ejector latches.
5. Restore power and observe the boot. The display should briefly cycle through b, d, 1, 2, 5, then 6 on a healthy boot. Anything else means the card is still suspect.
6. If you have a known-good spare card, swap it in. The known-good card must contain HMI, NCK, and PLC versions that match the cabinet hardware. Refer to Section 8 for the firmware matrix.
7. If you do not have a spare card but the original card is readable on a Linux PC, mount it read-only and run fsck -n on every partition. Repair with fsck -y on a copy, never on the original.

7.1 Force Service Mode via the HMI Service Switch

The NCU 720.1 supports a service-mode boot that loads a minimal HMI from a small rescue partition on the CF card. To enter service mode:

1. Power down.
2. Hold the service button on the NCU front panel (the small recessed button next to the rotary switches).
3. Restore power while continuing to hold the service button.
4. Release the service button when the 7-segment display shows 1 (HMI service mode).
5. The HMI will boot into the service menu and offer Restore, Backup, and Update options. Use a USB stick containing a known-good SINUMERIK archive (.arc) to perform a restore.

8. Firmware / Version Compatibility

Mismatched firmware on the CF card is one of the silent killers of an NCU boot. Validate the versions before doing anything destructive.

To read the versions from the CF card, mount it on a Linux workstation and inspect:

/card/system/configuration/version.ini
/card/hmi/version.txt
/card/nck/version.txt
/card/plc/<project>/version.txt

The presence of downgrade versions (e.g. NCK 4.8 with HMI 4.5) is a known boot-stall pattern. The HMI subsystem will refuse to register with the NCK and the controller will loop in 0 ↔ c.

9. Hardware Diagnostics

If Procedures A and B both fail, the NCU itself is suspect. Run through the hardware tests in order.

9.1 Power Supply

• Measure the 24 V DC supply at the NCU terminals. Acceptable range: 20.4 V to 28.8 V DC (Siemens SITOP / Logo!Power tolerance band).
• Scope the 24 V rail during HMI boot; inrush on the CF card + PLC can pull the rail down 1-2 V on a weak PSU. A drop below 19 V will reset the NCU during boot.
• Check the cabinet fan and filter; an overheated cabinet will trigger the NCU's thermal shutdown long before the operator sees a temperature alarm.

9.2 Battery and NV-RAM

• Battery voltage under load: must be ≥ 2.9 V DC.
• If the battery was disconnected or replaced, perform one full NCK reset (Procedure A) so the NC can rebuild its NV-RAM CRC. The first reboot after a battery replacement can take 2-3 minutes.

9.3 Fan

• The NCU 720.1 fan must be running immediately at power-on. If the fan is silent, the NCU is in thermal protection and may refuse to boot.
• Fan part number: Siemens 6FC5347-0AA02-0AA0 (12 V DC, 80 mA typ).

9.4 Backplane and Connector

• Remove and re-seat the NCU on the backplane to clean the connector pins.
• Inspect the backplane for discoloration, broken traces, or cold-solder joints, especially around the PLC and PROFINET connectors.

9.5 CF Card Socket

• If the CF socket shows bent pins or a cracked housing, replace the socket or the entire NCU 720.1 carrier.

10. Inline Boot-Flow Diagram

The diagram shows the normal boot path on the right (0 → 1 → c → 5 → 6) and the fault loop arrow returning from the PLC-config phase to HMI startup. Every time the PLC cannot complete configuration, the NCU restarts the HMI layer and the 7-segment display flips between 0 and c.

11. Verification Procedure

After any recovery step, verify the controller is genuinely healthy before handing the machine back to production.

1. 7-segment display settles on 6 and stays there.
2. HMI on the operator panel reaches the main operating area without any "NCK not reachable" or "PLC not reachable" red banner.
3. From the HMI, open Diagnostics → NC/PLC → Version and confirm NCK, HMI, and PLC versions are present and match the machine's commissioning record.
4. From the HMI, run Commissioning → Axis Test (or equivalent machine-builder test routine) on every axis. Confirm the drive enable, position feedback, and referencing all complete.
5. Trigger a controlled E-Stop and confirm the PLC drops the OK-to-move chain and the drives respond within the configured stop time.
6. Run the machine for a full shift under no-load and then under load. Monitor the NCU's internal log (Diagnostics → Logbook) for any re-occurrence of the boot-stall error codes.

12. When to RMA the NCU

Replace the NCU 720.1 if any of the following are true after Procedure B has been exhausted:

• The 7-segment display still oscillates with a known-good CF card loaded with matching firmware.
• The 24 V supply, battery, and fan are confirmed healthy and the unit still stalls.
• The PROFINET / PROFIBUS link LEDs never light, indicating an FPGA load failure.
• The NCU runs hot to the touch after five minutes of operation (more than ~55 °C on the bezel).

Siemens RMA channels for the 6FC5371-0AA00-0AA0 NCU 720.1 are handled through the regional Siemens DI / IA field service. Always return the original CF card with the NCU; Siemens uses it to recover the machine license.

13. Preventive Maintenance Schedule

14. Field Notes

The 30-minute battery-removal trick that was attempted in the original incident is a textbook last-resort move. It works because it forces the NCU to reload all NV-RAM state from the CF card, bypassing whatever was preventing the NCK from completing its boot-time consistency check. It also erases any PLC retentive data and may invalidate the machine's axis referencing, so do not use it as a routine procedure. The correct primary action is the rotary-switch NC/PLC clear; reserve the battery-pull for the case where the rotary clear is not enough but you have a CF card backup in hand.