Overview of CNC Maintenance Scope
Computer Numerical Control (CNC) machinery combines a controller, servo amplifier stack, spindle drive, feedback devices, mechanical transmission, hydraulic and pneumatic subsystems, and a wide range of auxiliary I/O. A CNC maintenance technician's role spans mechanical, electrical, electronic, software/firmware, lubrication, and process-troubleshooting disciplines. When the controller platform is either Fanuc (Series 0i, 30i, 31i, 32i family and successors) or Siemens (SINUMERIK 840D sl, ONE, 828D, and predecessors), diagnostic primitives differ but the maintenance scope is broadly parallel.
This reference focuses on field maintenance procedures common to both platforms. Where platform-specific behavior diverges, the article calls the divergence out rather than glossing over it. Categories referenced here are derived from publicly available documentation maintained by the relevant control vendors at their respective support portals, and any specific alarm code, firmware behavior, or parameter default should always be verified against the as-installed firmware revision before being acted on. Do not rely on verbal references from one machine to another: same platform name, different letter revision, different parameter map.
Common CNC Failure Categories
Field-reliable triage relies on grouping symptoms into a small number of physical domains. The table below summarizes recurring failure modes that surface on both Fanuc and Siemens installations across heavy-duty aerospace machining, automotive powertrain lines, prototyping shops, and high-mix job shops.
| Domain | Symptom signature | First 60 seconds | Cross-platform category |
|---|---|---|---|
| Mechanical | Positioning drift, surface-finish degradation, audible cyclic noise, tool-breakage rate uptick | Listen for bearing rumble, check way-cover seals, inspect lubrication, feel for backlash at handwheel | Common |
| Servo / drive | Excess following error, axis vibration at low feed, SV/Spindle alarm on power-up, occasional overcurrent trip | Inspect encoder feedback cable, check amplifier 7-segment status, record alarm ID and time-stamp, capture trace before reset | Common |
| Spindle | Toolholder tap-test fails, vibration during cut-off, spindle load deviation under constant surface speed | Measure drawbar force, check spindle bearings (cold-start warm-up), verify encoder air gap, verify toolholder cleanliness | Common |
| Hydraulic / pneumatic | Cylinder drift, chuck-pressure low, tool-changer slow / fault, way-lube warning | Inspect regulators, FRL, leakage, verify pressure gauge against spec, check pump filters | Common |
| NC controller | Won't enter READY, persistent NCK or PROFIsafe fault, persistent reset storm, HMI lockup | Capture NCK diagnostic log, check backup-battery voltage, review mode selection and key-switch position | Common |
| I/O and fieldbus | Discrete output stuck, servo-enable missing, hand-off faults, intermittent station failure | Check PROFIBUS/PROFINET cycle, EtherCAT state, DeviceNet status LEDs, double-check addresses | Common |
| Safety | E-stop latched, guard interlock won't reset, SLS / SLP trip, dual-channel disagreement | Verify guard circuit, dual-channel integrity, SI module diagnostics, cross-check Estop pushbutton wiring | Common |
| Process | Tolerance loss, tool-life inconsistency, chip-load alarms, surface finish issue | Inspect tool wear, coolant state, fixture repeatability, look for chip pack at flute | Common |
Fanuc Control Platform: Service Points and Architecture
Fanuc's controller family (Series 0i-MF, 30i-B, 31i-B5, 32i-MB and successor platforms) shares a common architecture: a CNC main board, a PMC (Programmable Machine Controller) processor, separate servo amplifier modules (SVM, βi series, αi series), a spindle-drive module (SPM), and a high-resolution pulse coder (typically αi or βi encoder protocol). Maintenance touches the following recurring service points:
- Backup battery: SRAM / CMOS battery (typically 3 V lithium) backing the part-program, parameters, and PMC ladder. Replace on a calendar interval, typically 3 to 5 years depending on environmental exposure; verify the controller does not flag battery-low at every cold start. A dropped battery in a powered-down machine is the most common cause of catastrophic parameter loss on Fanuc controls.
-
PMC ladder and FSSB: Fanuc Serial Servo Bus (FSSB) parameters and the servo-axis assignment table. Disturbances here manifest as
SV0401 / SV0404-class alarms referencing excess error in stop or excess error in move. Always verify FSSB parameter numbers (the1902 / 1910-1980series, revision-specific) are not corrupted by an interrupted firmware update. -
Servo parameter integrity: Detune, fast-feed, acceleration and jerk limits (typical parameters
1420, 1424, 1870-series, revision specific). Loss of these after a battery event is a top cause of axis-overcurrent trips during rapids. -
Spindle parameter set: Gear-ratio parameters and motor code plates. Field-replaceable spindle motors require re-entry of the motor code to avoid
SP9012 / AL-41-class speed-detect-circuit alarms at orientation. - Display and touch panel: HMI displays on a Fanuc CNC are an inventory consumable. The projected-capacitive touch screens lose calibration under heavy glove operation and industrial coolant exposure; manufacturer replacement procedure is documented per series.
Fanuc documentation entry points are hosted on the FANUC America Customer Resource Center - Manuals and the corporate FANUC product portal. Always pull the manual that matches the controller's letter-revision (for example 0i-MF or 31i-B5): parameter numbering changes between revisions and the alarm index reset by letter-revision is not consistent.
Siemens SINUMERIK: Service Points and Architecture
SINUMERIK controllers (840D sl, 840D, 828D, ONE) pair an NCK with a SIMOTION- or SINAMICS-based drive bus. Service points a maintenance technician works against on a Siemens platform include:
-
CF card and OEM area: SINUMERIK's PLC, NCK, and drive firmware sit on a removable CFast card. Treating it as a generic storage device (writes, remove, hot unplug during write) is a leading cause of NCK reset active and CF-card boot stalls. Always power down before removing media, and verify the
MD18800 $NC_TRAFO_TYPE-class machine-data boot when the boot record is fresh. -
Drive-object commissioning: SINAMICS S120 / S210 drives require
DO-level parameter sets. Topology-loss alarms in theF08501-class family are dominated by loose DRIVE-CLiQ connectors between the NCU and the line module, motor module, or sensor-module stack. - PROFINET / PROFIBUS diagnostics: SINUMERIK exposes per-port statistics on the HMI diagnostics page. NCU/PCU/IM modules have per-slave diagnostic buffers; field remedy for station-failure alarms is largely resolving cable fault, addressing mismatch, or controller-name assignment.
- Safety Integrated (SI): SINUMERIK Safety Integrated brings SPL (Safe Programmable Logic), SLP / SLS zones, and Safe Stop categories. SI faults demand safe validation of channel agreement before commissioning clears them.
- Trace buffers: The HMI's trace page (startup/commissioning to trace) records servo and spindle signals at the drive cycle rate; capture before reset, not after.
- HMI Pro / Operate: Operator-panel software versions and sinumerik-integrated PLC (PL) firmware must match the bundled NCU archive - rolling back part of the stack while leaving the rest is a common commissioning-archive mismatch source.
Siemens documentation entry points: the SINUMERIK product page for catalog and selection data, and the Siemens Industry Online Support portal as the canonical location for vendor KB articles on individual alarm codes and machine-data defaults. Per-machine-data definitions should always be cross-checked against the machine-data manual that shipped with the installed NCU software version.
Mechanical Subsystem Inspection
Mechanical service work is dominated by three subsystems: ballscrews, linear-motion guides (LM rails / box ways), and the Automatic Tool Changer (ATC). Skipping any of these during a periodic inspection is a leading contributor to repeat downtime events.
Ballscrew and LM-rail Touch-up
- De-energize per Lockout / Tagout on the controller, main disconnect, and any retained-energy storage (spindle air-bearing, hydraulic accumulator, counterbalance cylinder).
- With the axis at mid-stroke, hand-feel backlash using the handwheel or by commanding a small jog in each direction. Any single-direction free travel greater than the OEM backlash spec indicates a worn nut or pre-load decay.
- Using a dial indicator on the spindle housing, command a slow jog over 50 mm of travel in each direction. Repeatability outside the expected reproducibility budget indicates a worn bearing pack or coupling flexure.
- Inspect wipers and bellows. Worn wipers let coolant and chip ingress attack the nut; replace in matched sets.
- Re-lubricate per OEM spec, not by feel. Mixing grease families (for example Li-soap with PAO-thickened synthetic) is one of the most common lubrication failures.
Automatic Tool Changer (ATC)
- Inspect the ATC arm cam followers and gripper pads. Worn grippers fail the tap-test check and cause tool wedging in the spindle taper.
- Verify ATC air / hydraulic pressure against OEM spec at the regulator, not at the pump.
- Inspect the carousel rotation sensor and proximity switch targets for mechanical damage.
- Run a full tool-change test cycle in single-block / MDI before returning the machine to production.
Servo, Spindle, and Drive Verification
Electrical and electronic service work concentrates on the servo amplifier stack, the spindle drive, encoder feedback, and the NCK / HMI. The sequence below is platform-agnostic; on Fanuc controllers, observe via the SERVO and SPINDLE amplifier 7-segment, on SINUMERIK via the HMI diagnostics page and SINAMICS r0027 / r0039-class trace values.
Pre-Power Checks
| Check | Expected | Comment |
|---|---|---|
| Three-phase line voltage at the disconnect | Within OEM tolerance, balanced within roughly 2 % | Both phases unbalance and DC-bus ripple errors trace to incoming quality |
| 24 VDC control rail | 22.0 to 25.5 VDC | Undervoltage at the I/O produces sporadic fieldbus and discrete-output failures |
| Encoder cable shield | Bonded at controller end only, no chassis short | Double-grounded shields create encoder-error alarms at long-axis end |
| PE / chassis bonding | < 0.1 ohm to facility ground | Use a low-impedance bond tester, not a standard multimeter |
| DC-bus pre-charge indicator | Charges within spec time, no LED alarm | Discharge before touching bus capacitors |
Power-Up and Trace
- Power up the controller without axis enable. Capture the boot log; verify NCU / Series revision and license flags.
- Trace each axis velocity and following error at zero command, 30 second window. Any drift outside spec indicates encoder contamination or amplifier thermal issue.
- Enable single axis in jog; verify smooth motion and absence of oscillation or audible chatter at hand-change speed.
- Run a circularity / roundness test at part-cycle feedrate; quadrant transitions are the worst case for ballscrew backlash and bi-directional stiffness.
Spindle-specific
- Tap-test the spindle taper for cleanliness before each ATC event. Wipe with a clean, lint-free swab and approved solvent only.
- Verify drawbar / retention force with a calibrated pull-test fixture per OEM interval.
- Capture a constant-surface-speed spindle trace while applying a known load. Compare against the OEM spindle-power map.
- For belt-driven spindles: verify belt tension with a sonic or force gauge, not by deflection feel.
Hydraulic, Pneumatic, and Lubrication
Hydraulic and pneumatic subsystems fail in three modes: regulator drift (especially with thermal cycling), internal leakage (visible as cylinder drift under no command), and contamination (visible as slow valve actuation or pump noise).
| Subsystem | Verify | Indicators of failure |
|---|---|---|
| Hydraulic pressure | Verify at the work-port gauge, not at the pump gauge | Cylinder drift under hold, slow traverse, chatter at low feed |
| Pneumatic FRL | Drain bowls, verify regulator setpoint under load | Slow ATC, weak drawbar, intermittent auxiliary motion |
| Way lubrication | Cycle count vs. dispensed volume; verify distribution block at each way | Bearing and slide wear appear long before the lube-low alarm fires |
| Spindle lubrication (oil-air) | Stator flow at each metered outlet | Spindle bearing over-temperature trip without load |
| Coolant | Concentration, pH, suspended solids, tank sump screens | Surface finish degradation, corrosion on cast iron slides |
For oil-air lubrication, the oil column should be visible and intact at every outlet. A broken or short oil column is a near-certain indicator of a mis-pumped piston distributor.
Preventive Maintenance Schedule
The cadence table below is a starting point, not a substitute for the OEM recommended schedule shipped with the machine. Compress intervals under harsh duty cycles, especially high-spindle-on-time and high-axis-stroke environments.
| Interval | Action | Notes |
|---|---|---|
| Daily | Way-lube top-off, ATC taper wipe, coolant concentration, audible walk-around | Logged in the operator checklist |
| Weekly | Hydraulic / pneumatic gauges, FRL drain, way-cover inspection | Trend rather than react |
| Monthly | Ballscrew and LM-rail hand-feel, ATC gripper pad inspection, encoder cable shield inspection | Document backlash and reverse-pitch |
| Quarterly | Spindle drawbar pull-test, oil-air flow verification, axis trace capture | Save traces to a structured archive |
| Annual | Coolant tank clean, controller backup (NC + PLC), battery voltage check, full thermal image of cabinet | Maintain change-control records on every parameter edit |
| 3 to 5 Years | Battery replacement, contactor and breaker inspection, motion-control fan replacement | Use OEM-specified battery part |
Alarm Triage and Diagnostic Workflow
Alarm triage is half engineering and half discipline. The flowchart below captures a vendor-agnostic sequence; platform-specific alarm numbers are recovered from the named vendor KB resources.
Two practical rules minimize downtime during triage:
- Capture before clear. Pull the trace buffer, the HMI alarm history, and the NCU archive before any reset that clears non-volatile context.
- Match by exact letter-revision. Search the vendor KB by alarm code and the platform's letter-revision, not by platform name. A "31i" search returns both "31i-A" and "31i-B5" hits; parameter numbers differ.
Safety, LOTO, and Commissioning Procedures
CNC machinery violates multiple hazardous-energy categories at once. A formal Lockout / Tagout (LOTO) covering the controller, main disconnect, hydraulic accumulator, counterbalance cylinder, spindle pneumatic suspension, and any cap-stored high-voltage on the drive bus is required before reaching into the work envelope.
- Identify every energy source (electrical, hydraulic, pneumatic, gravity / counterbalance, thermal, capacitor-stored).
- Notify the operator and the area lead. For collaborative cells, lock the work envelope every cell.
- Isolate at the listed disconnect points. Verify zero-energy state with a known-good instrument at the workport.
- Apply lock and tag per site procedure. Confirm test point indents and any residual-discharge wait time on the drive bus.
- Hand-back of LOTO requires formal release, including operational check of any safety channel that was disturbed.
After any safety-channel disturbance - E-stop trip, guard replacement, encoder replacement, SI module replacement - a documented acceptance test must be performed. Acceptance tests record dual-channel agreement, response times, and SLS / SLP zone behavior at four corners of the operating envelope.
Field Cautions and Reference Notes
The following items are field-proven caveats reported across Fanuc and Siemens platforms over multi-year service windows.
| Domain | Caveat | Why it matters |
|---|---|---|
| Backup power | Battery failure during a Fanuc SRAM-only system produces a parameter loss that mirrors parameter corruption | Always restore from a known-good image-and-parameter backup |
| DRIVE-CLiQ / FSSB | Connector seating visually passes yet electrically fails under vibration | Replace rather than re-seat; keep a spare known-good loop |
| Firmware | Updating only one segment of the SINUMERIK stack (NCK without HMI or PLC) introduces subtle latent faults | Use OEM-released stack bundles; do not cherry-pick revisions |
| Coolant hygiene | Tramp-oil and high solids content accelerate way-pack wear and ballscrew lubrication failure | Skim and refresh on a calendar interval independent of concentration |
| Air / oil supply | Sub-micron water and oil in shop air foul FRLs and ATC valves within months | Install a desiccant dryer when failure recurs quarterly |
| Documentation | Parameter numbering changes with letter-revision on both platforms | Quote the manual's NCU-software or letter-revision, not just the platform name |
| OEM archives | Backup archives are version-specific; restoring a 5.6 archive onto a 6.5 NCK fails boot | Document backup version on the archive envelope |
For verified parameter numbers and alarm-code explanations, always defer to the platform-specific documentation:
- FANUC America manuals index
- FANUC product portal
- SINUMERIK product page
- Siemens Industry Online Support
None of the points above is a substitute for the machine builder's installation manual and the as-built electrical schematic. Field engineers should not deviate from site-specific procedures or from local regulatory requirements without documented authorization.
How do I recover from a Fanuc backup battery loss on a 31i-B5 controller?
Replace the battery first, then restore parameters and PMC ladder from a known-good backup image. Verify FSSB axis assignment (parameters in the 1902 / 1910-1980 series, revision-specific), the motor code plate data for any field-replaced spindle, and re-run the axis circularity test before returning the machine to production. Avoid trying to re-enter parameters manually unless no valid backup exists - manual transcription errors are the most common cause of a partial recovery that turns into an unplanned downtime event.
What causes a DRIVE-CLiQ topology fault on a SINUMERIK 840D sl installation?
DRIVE-CLiQ topology faults in the F08501-class family are dominated by loose or contaminated connectors between the NCU, line module, motor modules, and the Sensor Module stack. Power down, re-seat each connector with its correct torque, inspect for coolant ingress, and run the topology auto-detect from the HMI commissioning menu. Topology faults after a recent cabinet event almost always mean a disturbed connector rather than a failed module.
How often should way-cover wipers and bellows be replaced?
Quarterly visual inspection is the minimum. Replace wipers and bellows as a matched set when seals show hardening, tearing, or coolant ingress past the way-pack. Operating past a damaged seal typically shortens ballscrew and LM-rail life by years, so replacement on a calendar interval is cheaper than waiting for the part-quality alarm to fire.
Why does the spindle load meter deviate under constant surface speed?
Drift on the spindle load meter at constant surface speed usually traces to tool wear, chip pack at the flute, or a toolholder-taper cleanliness issue. Confirm by running the same program with a fresh tool in a known-clean taper; if the deviation persists, capture a spindle trace and compare against the OEM spindle-power map for the gear range in use. Bearing failure surfaces as a low-frequency modulation on the trace and a cold-start bearing temperature delta larger than the OEM spec.
Can I hot-swap a SINUMERIK CFast card while the NCU is powered?
No. The CFast card carries the active NCK and HMI partition on a live system. Always power the controller down before removing or inserting media, and always confirm the boot record version before returning the machine to production. Hot removal during a write will lock up the boot record and require a vendor-supported re-image procedure.