The Siemens 840D — in all its variants: 840D, 840Di, and 840D sl — is the most common Sinumerik control in UK workshops. Unlike Fanuc's sparse alarm codes or Heidenhain's verbose dialogue, Siemens uses a block-numbered alarm architecture that tells you which subsystem is failing just from the number range. Learn the ranges and you can triage before you touch a multimeter.
This guide covers the alarm groups you'll actually encounter on service calls — not the full 1,000+ page Siemens manual, but the patterns that let you diagnose quickly, protect machine data, and know when to escalate.
The 840D Alarm Architecture
Siemens alarms are grouped by thousands. The first digit tells you the subsystem:
| Range | Subsystem | Meaning |
|---|---|---|
| 000000–009999 | General NC | System software, internal NCK errors |
| 100000–199999 | Reserved | — |
| 200000–299999 | PLC | PLC status, sign-of-life, user alarms |
| 300000–399999 | Safety / E-stop | Safety chain, Integrated Safety alarms |
| 400000–499999 | Machine data | Axis config, parameter errors |
| 500000–599999 | — | — |
| 600000–699999 | Cycle alarms | Machining cycle parameter errors |
| 700000–799999 | OEM user alarms | Machine-builder-specific PLC alarms |
| 800000–899999 | — | — |
| 810001–810009 | Internal NC | Power-on system faults |
| Fxxxxx | SINAMICS drives | Power, thermal, comms, motor faults |
Rule of thumb: 2xxxxx = PLC, 3xxxxx = safety, 4xxxxx = configuration, 7xxxxx = OEM. F-numbers = drives. Everything else = NC system.
NC System Alarms (0xxxxx, 1xxxxx, 810xxx)
These are the ones you can't field-repair — they indicate internal NCK (Numerical Control Kernel) faults. But the diagnostic before you call Siemens matters.
810001–810009: Power-On System Faults
These appear at boot and are usually non-recoverable:
| Code | Likely Cause |
|---|---|
| 810001 | NCK software fault during power-up |
| 810004 | NCK hardware fault — MMC/NPU comms failed |
| 810007 | NCK battery-backed SRAM lost — total reload required |
810007 is the one you must prevent. The 840D uses battery-backed SRAM for machine data, PLC program, and drive configuration. If the NCK battery dies and you didn't back up, the machine is a paperweight until the OEM reloads it. Always back up before battery changes. Always.
1000–1019: Internal Processing Errors
These indicate computation or memory errors in the NCK. Common triggers:
- Corrupted NCK software after a failed firmware update
- DRAM failure on the NCU module
- Power supply brownout during operation
Action: Power cycle first. If the alarm returns, check the NCU module's diagnostic LEDs. A red SF (System Fault) LED with these codes means hardware replacement. No field fix.
PLC Alarms (200000–299999)
These are the most common service-call alarms because they span everything from PLC failure to machine-builder logic.
2000 — PLC Sign-of-Life Monitoring
The NCK polls the PLC at a fixed interval. If the PLC doesn't respond:
Causes:
- PLC in STOP mode (check the PLC status LED on the NCU — green RUN vs yellow STOP)
- PLC cycle time exceeded — an OB is stuck in a loop
- OB1 (main cycle) or OB40 (process alarm) not called — PLC program fault
- Service switch left in the wrong position after maintenance
Field check: Open the PLC diagnostics screen (Diagnosis → PLC → PLC Status). If it says STOP, cycle to RUN. If it immediately stops again, read the diagnostic buffer — it'll tell you which OB caused the stop.
2001 — PLC Has Not Started Up
Unlike 2000 (which means the PLC was running and stopped), 2001 means it never initialised. Common after:
- NCU replacement without reloading the PLC program
- Corrupted PLC archive on the CompactFlash card
- Power supply to the NCU/PLC side failed (check 24V rail)
700000–799999: OEM User Alarms
These are machine-builder-specific. The text is whatever the OEM programmed — "Coolant level low", "Way lube pressure fault", "Spindle chiller alarm". These are technically PLC alarms but use the 7xxxxx range because they're triggered by the OEM's PLC logic, not Siemens base code.
Diagnostic: Use the PLC status screen to trace the alarm back to the input that triggered it. The alarm text often includes the PLC flag or input address — match it to the electrical drawing.
Safety & E-Stop Alarms (300000–399999)
3000 — Emergency Stop
The safety chain is open. Unlike Heidenhain's single error 182, the 840D breaks E-stop causes into sub-codes:
| Sub-code | Meaning |
|---|---|
| 3000 | General E-stop — check all physical buttons and safety relays |
| 3001 | Internal E-stop — NCK-triggered, usually from axis limits or safe operating stop |
| 3003 | Safety door open — interlock switch signal lost on a guarded door |
Diagnostic: Siemens machines route E-stop through a safety relay (often Pilz or Siemens SIRIUS). Check the relay's channel LEDs — they show exactly which input dropped. Don't bridge safety circuits to diagnose. Use the PLC I/O screen instead.
Machine Data Alarms (400000–499999)
These appear after parameter changes, battery loss, or NCU swaps.
4000 — Gap in Axis Assignment
An axis is referenced in a channel but not assigned to a physical drive. This happens when machine data is partially reloaded — the axis table references a slot that has no drive object.
Fix: Check MD10000 $MN_AXCONF_MACHAX_NAME_TAB and MD20070 $MC_AXCONF_MACHAX_USED. The axis names in both must match real drive objects.
4010 — Invalid Identifier in Machine Data
A machine data field contains a value that doesn't correspond to a valid option. Common after manual editing of the machine data file.
Fix: Reload a known-good MD backup. Don't hand-edit machine data unless you're certain of the option list.
SINAMICS Drive Alarms (Fxxxxx)
The 840D sl integrates with SINAMICS S120 drives over DRIVE-CLiQ. These are the most common production-stopping faults.
F30003 — DC Link Undervoltage
The drive's internal DC bus dropped below threshold. This is a power supply problem, not a drive problem.
Check order:
- Incoming mains voltage — is the machine getting clean three-phase?
- Line reactor/commutation choke — loose connection or blown
- Line Module (Active Line Module or Smart Line Module) — check for internal DC link fault
- Regenerative feedback — if the Line Module can't dump braking energy, the DC link sags
F30005 — I²T Overload
The drive has been running above rated current for a sustained period. The I²T model is thermal — it calculates motor heating and trips before damage.
Common causes:
- Mechanical binding on that axis — check slideways, ballscrew, gibs
- Tool overload — feed rate too high, tool blunt
- Counterbalance failure on vertical axis — servo carrying the load
- Parameter mismatch — motor data in the drive doesn't match the physical motor
F30021 — Ground Fault
Insulation failure in the motor or motor cable. This is a hard fault — the drive will not restart until the ground is cleared.
Field isolation:
- Disconnect the motor power cable at the drive end
- Megger test the motor and cable separately (500V DC)
- If the motor reads low (< 1MΩ), it's winding insulation breakdown — motor replacement
- If the cable reads low, inspect for coolant ingress at connectors or damage in the cable chain
F07011 — Motor Overtemperature
The motor's KTY84 or PTC thermistor is reporting over-temperature. Check:
- Motor cooling fan — running? Filters clean?
- Ambient temperature in the cabinet and machine enclosure
- Actual motor temperature via drive parameter r0035 — if it reads unrealistic values (e.g. 25°C when the motor is hot), the sensor circuit is open
- Thermistor wiring through connectors — these are the first to corrode
Cycle Alarms (600000–699999)
These are the "your program is asking for something impossible" alarms. They appear during cycle execution, not at power-up.
| Code | Scenario |
|---|---|
| 61000 | No tool offset active |
| 61002 | Geometry axis not defined for the cycle |
| 61101 | Feed rate = 0 or negative |
| 61601 | Thread pitch incorrectly defined |
These are operator/programmer errors, not machine faults. Check the program block, verify tool data, confirm the workpiece coordinate system.
The Field Protocol — 840D Edition
When you walk up to an 840D that's down:
1. Back Up Everything First
The 840D stores data across multiple subsystems. A full backup captures all of them:
HMI: Menu Select → Service → Series Start-Up → PLC → Copy PLC → Archive
HMI: Menu Select → Service → Series Start-Up → NC → Copy NC → Archive
HMI: Menu Select → Service → Series Start-Up → Drive → Copy Drive → Archive
Without a full backup, one wrong move and the machine needs an OEM reload. Three separate archives. Do all three.
2. Read the Diagnostic Buffer
The alarm screen shows current alarms. The diagnostic buffer shows history. Menu Select → Diagnosis → PLC → Diagnostic Buffer. Scroll back to see the sequence. A drive fault followed by an E-stop followed by a PLC alarm means the drive fault is the root cause — the others are cascade.
3. Check PLC Status First
Green RUN = PLC is healthy, trace the specific alarm. Yellow STOP = PLC is down, that's your problem, everything else is secondary.
4. SINAMICS Drives: Read r-Parameters
Don't guess drive faults. Connect to the drive (HMI → Diagnosis → Drive → Service Overview) and read the r-parameters:
- r0035: Motor temperature
- r0068: Actual current
- r0070: DC link voltage
- r0947: Fault history (last 8 faults)
The fault history is under-used. An F30003 followed by an F30021 five minutes later tells you the ground fault happened during the undervoltage — probably a cable that arced when the voltage sagged.
5. One Subsystem at a Time
If you're seeing NC, PLC, drive, and safety alarms simultaneously, resolve in this order:
- Power — undervoltage, phase loss, 24V rail
- PLC — if the PLC is stopped, fix that first
- Safety — E-stop chain, door interlocks
- Drives — individual axis/drive faults
- NC — system-level configuration
Fix power and the cascade often clears itself.
Battery Management — The #1 Preventable Disaster
The 840D uses a BR2032 lithium battery on the NCU module to hold SRAM during power-off. When the battery voltage drops below 2.8V:
- 2100: NCK battery warning — you have time, but not much
- 2101: NCK battery alarm — replace immediately
- 2102: NCK battery alarm on power-up — replace before next power cycle
Replace the battery with the machine powered on. The NCU holds SRAM from the 24V supply during the swap. If you power down with a dead battery, the machine data, PLC program, and drive configuration are gone.
Check the battery every six months. Replace annually regardless of voltage. A £3 battery prevents a £3,000 service call.
When to Escalate
Some 840D errors are non-negotiable:
- 810xxx boot faults — internal NCK hardware or firmware
- DRIVE-CLiQ topology faults — the control can't see a drive component; could be hardware, cable, or configuration
- NCK/PLC sync loss — the two processors are not communicating; hardware
- SINAMICS internal faults — component-level drive failure (gate driver, IGBT, control board)
- Safety Integrated configuration loss — safety parameters are corrupted or missing; requires Siemens commissioning tool
If the diagnostic buffer says "contact Siemens" or "replace hardware," it's not ambiguity. Siemens documentation only uses those phrases when field repair isn't possible.
Summary
The 840D's numbered alarm ranges are its biggest diagnostic advantage over Fanuc and Heidenhain. Learn the thousands blocks, check PLC status before anything else, and never touch a machine without three separate archives — NC, PLC, and drive.
If you're staring at an 840D fault and production is stopped, we cover Siemens controls across the UK — from legacy 840D powerline on 1990s machines to 840D sl with DRIVE-CLiQ and Safety Integrated on current-generation 5-axis mills. Call 07561 040702 or read about our control system repair services.