GreatLight CNC Machining Factory logo
CNC Machining
Rapid Prototyping
Materials
Industries
News
About GL

Get Instant Quote

Maintenance Guide

Siemens CNC System Maintenance Method

This guide is for maintenance engineers and shop supervisors running Sinumerik 840D, 828D, or 802D controls on production machines. It covers what to check weekly, monthly, and yearly, which faults point to which hardware, and when a controller problem is really a machine problem.

Battery backupEncoder feedbackDrive coolingData archive
Siemens CNC system maintenance method on an axis drive module
Quick answer

Key takeaways

Batteries before anything elseThe buffer battery keeps the absolute encoder and SRAM alive. Measure it monthly, replace at 3.0 V, not when the alarm appears.
Backups are the real spare partWithout a valid NC and PLC archive, a dead module costs days. Export one every month to an off-machine drive.
Most faults are thermal or mechanicalEncoder drift, lost reference, and 25050 alarms often trace back to heat, chips, or a loose coupling before the control board.
Clean cooling, longer lifeClogged fan filters on the drive rack push module temperatures up. Swap filters on a schedule, not on failure.
Check the machine firstA Siemens alarm number describes the symptom, not the cause. Confirm the mechanics before ordering a board.
Scope

What this Siemens CNC system maintenance method covers

Sinumerik controls fail in a small number of predictable ways. The board itself is rarely the first thing to go. Batteries run flat, cabinets overheat, fans seize, encoder cables get nicked by chip conveyors, and hard drives fill with old archives. A maintenance routine that watches those five things will catch most downtime before it starts.

This method applies to 840D sl, 840D powerline, 828D, and 802D sl. The exact menu names differ between generations, but the order of operations does not. Everything here runs from the machine panel or the HMI, no laptop required for the daily and weekly checks.

One warning before we start. Do not pull a module, reseat a board, or clear an alarm while the machine is referenced and mid-cycle. Power down at the main switch, wait for the DC link to discharge, then work. A 600 V DC link does not care about your schedule.

We keep 127 high-precision CNC machines running across three plants in Dongguan and Singapore, so this is a routine we run on our own floor. The intervals below are what we use on two-shift production.

Daily

Daily checks: cabinet heat, air, and alarm history

Start with the cabinet, not the screen. Open the electrical cabinet door and put a hand near the top of the drive rack. It should feel warm, not hot. If you cannot hold your palm there for three seconds, the airflow is already failing and you are running toward a thermal alarm.

Read the alarm history next. On 840D, go to Diagnosis, then Alarm list, and look at the last 24 hours rather than the current screen. Intermittent 25050 contour monitoring alarms, 3000 series axis alarms, and 21612 encoder alarms are all worth logging with time and axis.

Check the compressed air supply to the cabinet. If your machine uses positive pressure or air purging, verify the regulator is set to the machine builder spec, usually 5 to 6 bar. A dropped filter regulator lets fine dust into the drive rack, and that dust is conductive.

Finally, wipe the operator panel and check the emergency stop circuit by pressing it once at the start of the shift. A stuck contact in the E-stop chain looks like a control fault but is a 20-minute fix.

Monthly

Monthly: battery voltage and archive discipline

The buffer battery is the single most common cause of lost parameters on Sinumerik controls. Measure it under load with a digital meter, not with the diagnostics screen alone. A healthy lithium cell reads 3.5 to 3.6 V. At 3.0 V, replace it. Do not wait for the 2100 or 2110 battery alarm, because by then the SRAM may already be corrupted.

Replace batteries with the control powered on if the module design allows it, so the SRAM stays alive. If your module requires a power-down swap, complete the change inside 30 minutes and have the archive ready.

Archive discipline is the other half. Export the NC and PLC archives to a USB stick, then copy that stick to a network drive the same day. Label files with machine number, date, and control type. A backup that only exists on the machine is not a backup.

Note which axis modules carry absolute encoders. Those encoders need their own battery or capacitor backup, and the replacement interval is often shorter than the control battery.

Quarterly

Quarterly: cooling, connectors, and grounding

Pull the fan filter mats on the cabinet and drive rack. Hold them up to a light. If you cannot see through, replace them. On a shop with cast iron or graphite dust, that interval may need to be monthly. Wipe the fan blades too, since a loaded blade spins slower and moves less air.

Check the drive cooling fans by listening and by feel. With the machine powered, hold a strip of paper near the fan outlet. If it does not flutter, the fan is weak or the heatsink fins are packed. Blow out the fins with dry, oil-free air at under 2 bar.

Inspect the encoder and motor power connectors at the drive. Look for darkened pins, green corrosion, or a connector shell that has walked loose. Vibration loosens these over months. Torque them to the manufacturer value and mark them with a paint line so you can see movement later.

Verify cabinet grounding. Measure from the PE bar to the machine frame. You want a reading under 1 Ω. High resistance shows up as random encoder faults and communication dropouts that nobody can reproduce.

Procedure

Step by step: a monthly Siemens CNC system maintenance method

Run this sequence on a stopped machine. Total time is about 90 minutes per control.

  • 1
    1. Lock out and record stateSwitch off at the main breaker, tag it, and wait 5 minutes for the DC link to discharge. Note the current axis positions and whether the machine is referenced.
  • 2
    2. Export NC and PLC archivesPower up in setup mode. Go to Services, then Archive. Save NC and PLC separately to USB. Verify file size is not zero before you remove the stick.
  • 3
    3. Measure buffer batteriesWith the control on, measure each battery under load. Record voltage in the log. Replace anything at or below 3.0 V, and replace all batteries in the same module together.
  • 4
    4. Clean cooling pathsReplace filter mats, blow out heatsink fins at under 2 bar, and spin every fan by finger to check bearing play. A fan that wobbles gets replaced, not greased.
  • 5
    5. Inspect connectors and cablesReseat encoder and motor power connectors, check for chafing near cable carriers, and torque to spec. Mark each connector with a paint line for later visual checks.
  • 6
    6. Review alarm and message logsPull the last 30 days of alarms. Group them by axis and by time of day. Repeats on one axis point to that axis, not to the control.
  • 7
    7. Re-reference and verifyHome every axis, run a warm-up program for 15 minutes, then check backlash and a test cut. Log the result and the date, so the next technician has a baseline.
  • 8
    8. Store the archive off-machineCopy the USB archive to the network folder named for that machine. Keep the last three monthly sets. Delete older ones only after a verified restore test.
Fault map

Symptom, likely cause, and first action

Use this before you order a replacement module.

SymptomLikely causeFirst action
Blank screen on power-upNo 24 V at the PCU, or dead monitor backlightCheck 24 V supply and the 03840 board LED
2100 / 2110 battery alarmBuffer battery below 3.0 VReplace battery with control powered on
25050 contour monitoringMechanical binding, worn guide, or servo tuningCheck axis load and backlash before the drive
Axis drifts after referenceEncoder coupling loose or contaminated scaleClean scale and re-clamp the coupling
Random communication dropoutsPoor cabinet grounding or chafed encoder cableMeasure PE resistance, inspect cable carrier
Overheat alarm on drive moduleClogged filter mat or seized fanReplace mat, test fan with paper strip
Lost parameters after weekendBattery flat plus no valid archiveRestore archive, then replace battery and re-check

The short version

Measure the battery, keep a current archive, and clean the airflow. Those three habits prevent most Sinumerik downtime, and the rest is diagnostics.

FAQs

Common questions

How often should I replace the buffer battery on a Sinumerik control?

Measure monthly and replace at 3.0 V, which on a two-shift machine is typically every 12 to 18 months. If the machine sits powered off for long periods, the interval shortens, because the battery drains even when the control is off.

Keep a dated log. A battery that drops from 3.5 V to 3.2 V in one month is telling you something else is wrong, usually a charging circuit or a failing module.

Can I replace the battery without losing parameters?

On most 840D and 828D modules, yes, if you change it with the control powered on and follow the manufacturer procedure for that specific module. Have a current NC and PLC archive on USB before you start, in case the swap goes wrong.

If the module only allows a power-down swap, work fast and keep the machine on a UPS if one is fitted. Thirty minutes is the usual safe window.

Do I need to back up the PLC and NC separately?

Yes. They are separate archives and a full restore needs both. Export them to separate files on the same USB stick, and name the files with machine number and date so a restore months later is not guesswork.

Test a restore on a spare or offline control once a year. An archive that has never been restored is a hope, not a backup.

What causes a 25050 contour monitoring alarm?

It means the actual axis position drifted further from the command than the monitoring window allows. In most cases the cause is mechanical: a tight guide, a worn ball screw, chips packed under a way cover, or a loose encoder coupling.

Before you touch drive parameters, measure the axis load and check backlash. If the mechanics are clean and the alarm still repeats on one axis, then look at servo tuning and the encoder signal.

Is an overheat alarm on the drive module a control problem?

Rarely. It is almost always an airflow problem: a clogged filter mat, a seized fan, packed heatsink fins, or a cabinet that has no cool air entering it. Fix the airflow first.

Replace filter mats on a fixed schedule, and test every fan with a paper strip at each quarterly service. A fan that is slow but still spinning is the hardest one to catch.

When should we call in outside help instead of maintaining it ourselves?

Call for help when a fault follows a module rather than an axis, when you have two or more unexplained encoder alarms on different axes, or when a restore fails. Those point to hardware that needs a controlled swap and a full re-commissioning.

Keep the alarm log, the archive dates, and the battery voltages ready. That record cuts diagnosis time more than any single tool.

Need machined spares or a second opinion on a failing axis?

Send us the drawing or the alarm log. We quote and return a DFM analysis within 12 hours, and we machine from one prototype to 10,000+ part runs.

12-hour quote100% inspection±0.005 mm

Follow

More from the shop floor

We publish setup notes, tooling trials and inspection data from the factory floor.

FacebookTikTokYouTubeLinkedInInstagramThreadsPinterest

Trusted by engineers and manufacturers worldwide

Tesla Ford Motor Company BYD Auto Denso Magna International Boeing Airbus Medtronic KUKA FANUC