SINUMERIK 810D CCU1 Alarm 2001 "PLC Not Started": Complete Recovery Guide

The SINUMERIK 810D CCU1 (Compact Control Unit 1) is the integrated CNC and PLC controller used on a wide range of milling and turning machines, including retrofits on Emco, Deckel, Maho, and similar platforms. Because the CCU1 can boot its firmware and PLC image directly from either the on-board JEIDA memory or an inserted PCMCIA card, any PCMCIA card carrying a different SINUMERIK software generation (for example, an 840D sl image inserted into an 810D CCU1) can be unintentionally picked up by the boot loader on the next power-up. The result is a corrupted PLC image and alarm 2001 "PLC not started" on the operator panel.

This reference documents the boot architecture, switch configuration, recovery sequence, and verification steps needed to bring the controller back to a runnable state without losing machine-specific compensation data, tool tables, or PLC programs. Official startup and commissioning manuals are available through the Siemens Industry Online Support portal under the SINUMERIK 810D product tree.

1. Affected Hardware and Software

The CCU1 board revision 570 665 9002.03 corresponds to an early CCU1 production batch. On this revision, the internal JEIDA memory is the primary boot device and the PCMCIA slot is only read when a card is physically inserted at power-up. Newer CCU boards add a switch position (S3 = 4) that forces a permanent boot from internal memory even with a card inserted, but revision .03 typically does not include that position.

2. Root Cause: Accidental PCMCIA Firmware Update

When the CCU1 powers up, the boot loader executes the following sequence:

1. Check rotary switch S3 position.
2. If S3 = 2, treat any inserted PCMCIA card as a firmware source and run an automatic update without operator confirmation.
3. If S3 = 0 (normal) but a card is present and contains a boot image with an executable auto-update marker in its root directory, the loader can still hand off to the card image depending on the CCU revision and the contents of the card root.
4. Compare the version stamp on the card with the version stamp in JEIDA memory. On some CCU1 revisions, a mismatch alone is enough to trigger an automatic copy operation.

Inserting an 840D sl / 840DE PCMCIA card into an 810D CCU1 slot under these conditions overwrites the JEIDA-resident NCK and PLC firmware. After the next reset, the integrated PLC runtime no longer matches the NCK version, the PLC stays in STOP, and the operator panel raises alarm 2001.

3. CCU1 Memory Architecture

The CCU1 contains three distinct non-volatile regions plus the battery-backed SRAM. Understanding what each region holds is essential before attempting a recovery.

When an 840D card overwrites the JEIDA Flash, the CCU1 loses its PLC runtime image. The SRAM still holds the last valid machine data set, but the PLC can no longer start because there is no PLC operating system to interpret the ladder/S7 code stored in the Flash file system.

4. S3 and S4 Mode Switch Reference

Two rotary switches are mounted on the front of the CCU1 behind the operator panel or behind the front cover. They are the only manual controls available during a service recovery and must be operated with the controller powered off.

4.1 S3 — Operating Mode Switch

4.2 S4 — PLC Service Switch

The classic CCU1 PLC reset handshake is the sequence S4 = 3 → 2 → 3 → 2 → 0 in three-second intervals while S3 is in position 1. This pattern is documented in the SINUMERIK 810D startup and commissioning manual and is the first action to attempt when the PLC refuses to start.

5. Alarm 2001 — Definition and Triggers

Alarm 2001 is raised by the NCK whenever the integrated PLC interface reports that the PLC is not in RUN state at the end of the startup handshake. The full alarm text on the operator panel is 2001 PLC not started or 2001 PLC no start depending on the HMI language pack.

The seven-segment display on the CCU1 showing 6 during boot is the standard indicator that the integrated PLC has not completed its handshake with the NCK. The SF (System Fault) LED on the CCU confirms that the PLC operating system cannot initialise.

6. Prerequisites Before Recovery

Before touching S3/S4 or any cabling, gather the following items. Attempting a recovery without them usually ends in a longer downtime than the original incident.

1. Original NCK backup — archive from the machine builder containing machine data, setting data, tool data, and compensation data (.ARC or .NKS format).
2. Original PLC program — the S7 project (.S7 / .AWL) provided by the OEM. For Emco machines, request the factory backup directly from Emco service if the local copy is suspected to be stale.
3. Correct firmware image — a PCMCIA card containing the matching 810D CCU1 software (not 840D, not 840D sl). Do not improvise with cards from other machines.
4. Serial transfer tool — WinPCIN (Windows) or PCIN (DOS) plus a serial cable connected to the COM port on the CCU1 front panel for archive transfer.
5. MPI / PROFIBUS cable — for STEP 7 online access to the PLC if WinPCIN alone cannot restore the PLC program.
6. Battery check — verify that the on-board lithium battery still holds voltage above 3.0 V. A weak battery will wipe SRAM as soon as power is removed and erase all retained machine data.
7. Schematics and parameter list — for cross-checking machine data against the actual machine configuration (axes, spindles, encoder type, drive assignments).
8. STEP 7 V5.x project — for online diagnostic buffer reads (PLC > Module Information > Diagnostic Buffer).

7. Step-by-Step Recovery Procedure

7.1 Verify Physical State

1. Power off the machine at the main disconnect.
2. Open the CCU1 housing.
3. Remove any PCMCIA card from the slot. Label it and set it aside — do not reinsert it.
4. Inspect the on-board battery. Replace it if it is older than five years or below 3.0 V.
5. Check the S3 and S4 positions. Both should be at 0 for a clean power-on baseline.

7.2 Cold Start the CCU1

1. Confirm S3 = 0 and S4 = 0.
2. Power on the machine at the main disconnect.
3. Watch the seven-segment display on the CCU1 and the SF / PLC LEDs.
4. If the segment display reaches a steady state (typically a number like 1, 2, or 6) without alarm 2001, skip directly to Section 7.5 — verify the existing image first.

7.3 PLC Reset via S4 Sequence

If the PLC stays in STOP and alarm 2001 returns, perform the PLC overall reset handshake:

1. Power off.
2. Set S3 = 1 (boot/service mode).
3. Set S4 = 3 (PLC stop).
4. Power on. Wait until the segment display finishes cycling and the PLC LED stops flashing (approximately 30-60 s).
5. Set S4 = 2 for 1 s, back to 3 for 1 s, back to 2 for 1 s, then to 0 — all transitions within 3 s.
6. The PLC LED should extinguish, then relight. This indicates a successful overall reset.

7.4 NCK and PLC Reload

Reload the NCK data first, then the PLC image. Loading PLC data first on a CCU1 in this state has been observed to leave alarm 2001 active even when the PLC is healthy.

1. Leave S3 = 1 and S4 = 0.
2. On the operator panel, navigate to Services > Setup > Password and enter the manufacturer password. The standard SINUMERIK default is SUNRISE, although many OEMs (including Emco) override this with a machine-specific code.
3. Select Services > Data Backup > NCK Data and load the NCK archive from the serial port or PCMCIA.
4. Wait for the NCK restart. The display should return to the standard operator panel view.
5. Select Services > Data Backup > PLC Data and load the PLC archive.
6. If prompted for the archive format, select Punch strip / AWL for plain .AWL files, or Binary for S7 binary archives. Selecting the wrong format leaves the PLC image unbootable.
7. When prompted, allow the PLC to restart automatically.

7.5 Switch Back to Normal Operation

1. Power off.
2. Set S3 = 0 and S4 = 0.
3. Power on.
4. Watch the seven-segment display. It should advance to 6 briefly and then extinguish as the HMI takes over the display.
5. Confirm that alarm 2001 is no longer present.

8. Verification

A full recovery is only confirmed when all of the following return green:

1. Operator panel reaches the standard channel view without alarm 2001.
2. SF LED on the CCU1 is off.
3. PLC LED is solid green (RUN).
4. All axes reference to their end stops / reference cams without alarm 1007 or 1008.
5. Spindle reaches command speed in each gear range without alarm 2200.
6. Tool table numbers and tool offsets match the OEM commissioning sheet.
7. M functions fire correctly on a dry-run part program.
8. Handwheel and jog keys produce axis motion at the configured rapid rate.
9. NCK MD5xxx drive assignments (drive bus address, motor type, encoder type) match the actual hardware configuration.
10. PLC cycle time (visible in STEP 7 or on the operator panel diagnostic page) is within the expected 10-25 ms range.

If any axis refuses to reference, the most likely root cause is a mismatch between the loaded NCK machine data and the actual encoder / drive configuration. Cross-check the following MD values against the OEM commissioning sheet:

9. PCMCIA Card Best Practices

• Label every PCMCIA card with the machine serial number, software generation, and date of creation.
• Store PCMCIA cards in anti-static sleeves in a dry location between 5 °C and 35 °C.
• Never insert a card of unknown origin into a running CCU1. If a card is in your hand and you cannot identify its contents, do not insert it.
• For routine program transfer, use WinPCIN over RS-232 rather than the PCMCIA slot.
• If a firmware update is genuinely required, prepare the card, set S3 = 2, and confirm the card generation matches the controller before powering on.
• After any successful update, archive the contents of the card with a unique serial number so it can never be confused with another machine's card.
• Maintain at least two known-good PCMCIA cards per machine — one as the live update card and one as a verified backup stored off-machine.

10. Firmware Restoration via PCMCIA (If Required)

If the JEIDA Flash is so corrupted that the CCU1 will not even reach the segment display state, a clean firmware reload from PCMCIA is required.

1. Prepare a PCMCIA card containing the correct 810D CCU1 software generation. For SW 1.6 on board 570 665 9002.03, use the matching 810D CCU1 image — not 840D, not 840D sl.
2. Power off, insert the card.
3. Set S3 = 2.
4. Power on. The CCU1 will detect the card, prompt on the operator panel (or on the seven-segment display if the HMI is also missing), and run the update.
5. Do not interrupt the update. Progress is shown on the seven-segment display.
6. When the display reaches its final state and the HMI boots, set S3 back to 0 and reload NCK and PLC data from backup as in Section 7.4.

11. Troubleshooting Matrix

12. Field-Proven Caveats

• OEM database drift: The most common cause of a phantom "failed" recovery is that the OEM has updated the machine internally but not updated the backup files in their service database. Always cross-check the OEM-provided archive against an in-house backup if one exists from commissioning. A backup made on the day of installation is almost always more reliable than one requested years later.
• MMC language pack: Swapping the MMC100.1 (for example Swedish) for MMC101 (English) does not change the underlying machine data, but it does change the displayed alarm text and the HMI menus. Be sure the OEM service team reads the alarm text from the same language pack you are looking at, otherwise you will both be debugging different alarms.
• Battery replacement timing: Always replace the lithium battery with power applied to the CCU1. Removing the battery with power removed wipes SRAM. If the SRAM has been wiped, the controller falls back to default machine data on the next boot and axes will need to be re-referenced.
• Spare NCU: An MMC101 from a different machine family is not a drop-in replacement for the 810D CCU1 and can cause cascading PLC errors even if it boots. Keep NCU / MMC swaps within the same software generation.
• STEP 7 access: For deeper diagnostics, connect STEP 7 V5.x via MPI to the CCU1 and read the S7 diagnostic buffer (PLC > Module Information > Diagnostic Buffer). The buffer entries reveal whether the PLC failed in OB100 (startup), OB1 (cyclic), or due to an I/O fault — each requires a different fix.
• SF vs PLC LED pattern: SF red alone with PLC green means the PLC operating system is loaded but a peripheral or program fault is active. SF red with PLC dark or yellow means the PLC operating system itself is missing — go back to Section 10 for a firmware reload.
• Serial cable pinning: The CCU1 COM port uses the Siemens V.24 pinout, not a standard PC V.24. A straight-through cable will not work. Use a Siemens-specific serial cable or wire per the commissioning manual pinout table.
• Password lockout: After three failed password attempts, the HMI imposes a waiting period. Plan password entry ahead of time and do not test random codes during a recovery.

13. Frequently Asked Questions