Siemens 810M Crashes: Z-Axis Encoder Fault, Not Cooling

The control drops out without a dependable pattern, and a power cycle brings it back only temporarily. On the documented Siemens 810M installation, replacing the cooling fans did not stop the failures; replacing the Z-axis rotary encoder did. Treat that result as the lead diagnostic path, but prove the surrounding power, connector, and feedback circuit before fitting parts.

Capture the crash before cycling power

Do not start with another blind restart. The machine initially failed anywhere from several times per day to once a week or once a month. The rate later increased to as many as 15 to 20 crashes in one day. That progression points toward an intermittent component or connection deteriorating over time, not a normal operating sequence.

1. Record whether the display blanks, freezes, resets, or continues showing a static image.
2. Record the states of the control power supply, PLC, drives, contactors, cabinet interlocks, and diagnostic indicators before removing power.
3. Note whether the failure occurs while an axis moves, while the spindle runs, while the machine is idle, or immediately after a command.
4. Record the active axis, Z-axis position, direction, and motion command when the crash occurs.
5. Copy any alarm or diagnostic-buffer information before restarting, if the control still permits access.

A restart erases the most useful distinction: whether control power disappeared or the processor stopped responding while power remained present. Use the machine schematic to identify the control supply and test points; do not infer them from the display wiring.

Check before continuing: classify one event as a power loss, processor reset, frozen control, or communications/display failure. If no state survives long enough to classify it, arrange temporary measurement or event capture on the control supply and Z-axis feedback circuit.

Reject the usual quick fixes

The buffer batteries had already been replaced, with no change in the running failure. A buffer battery normally preserves retained data when normal power is absent; replacing it does not repair an intermittent crash during powered operation unless battery monitoring or its circuit is directly implicated by a diagnostic.

Cooling was also a reasonable early suspect. The 810M monitors a fan located on the control near the measuring-system card. Insufficient airflow through the control and power supply can lead to a delayed power-supply shutdown, and a defective power-supply monitoring circuit can produce a similar symptom. However, this machine could fail about 10 minutes after a cold morning start, and replacing the fans did not stop the crashes.

Check before continuing: confirm that the installed fans run, airflow is unobstructed, and no thermal or fan-monitor indication accompanies the event. If those checks pass and a replacement fan changes nothing, leave the cooling system in service and advance to the electrical paths.

Map the control, panel, and cabinet

The installation was a 1988 Bluthardt machine with the operating panel located well away from the electrical cabinet. That physical separation matters because control power, display signals, axis feedback, bonding, and interlock wiring travel between machine sections. Every connector and cable in that route becomes part of the fault boundary.

The monitor had previously been replaced. Its supply was tentatively described as 24 V, but the more specific installation information identifies 12 VDC from the graphics card. Read the schematic and measure the actual circuit before treating either value as a test specification. A display fault can resemble a control crash, while electrical disturbance from a monitor circuit can affect nearby signals; the state of the drives and control outputs decides whether the complete control failed or only the display path failed.

1. Mark the physical locations of the control rack, graphics card, monitor, power supply, measuring-system card, PLC hardware, drives, Z-axis encoder, and cabinet safety switches.
2. Trace the conductors and connectors between the remote panel and cabinet.
3. Identify cable shields, bonding points, and separations between feedback wiring and higher-energy conductors.
4. Compare connector labels and supply values with the machine schematic.

Check before continuing: verify that the drawing matches the installed wiring and that the observed failure affects the control rather than only the monitor. Stop here if undocumented modifications prevent safe identification of the supplies or feedback connections.

Prove the power and interlock path

Intermittent power connections can reset a control without leaving a persistent alarm. Inspect plug contacts at the control console, cabinet, control rack, and drives. Look for loose retention, oxidation, contamination, damaged pins, conductor movement at terminations, cable damage from chips, and cracked or cold solder joints where accessible through an approved repair process.

Check any cabinet-door safety switch that removes power when the door opens. Mechanical vibration can make a worn or poorly adjusted switch chatter even when the door appears closed. Because vibration did not correlate with this installation's failures, prove switch continuity or monitor its associated circuit rather than replacing the switch on suspicion.

Measure the control supply at its receiving terminals during operation, not only at the source with the machine idle. Capture the minimum value or transient interruption during a crash with instrumentation suited to the circuit. A normal handheld reading after the reset does not exclude a brief dropout. Also inspect the power-supply monitoring circuit because a false monitor trip can shut the supply down even when airflow is adequate.

A power-supply replacement solved a superficially similar problem on 840D controls, where replacement units had gold-plated contacts. That is a useful connector and supply clue, not proof that an 840D supply remedy transfers to an 810M. A PLC CPU replacement also corrected a crash on a 3M, but changing the CPU before separating supply loss from feedback disturbance is uncontrolled substitution.

Check before continuing: observe stable control power, closed interlock circuits, and secure connections through a representative machine cycle. If the control remains powered when it crashes, shift the investigation from incoming power toward processor and axis-feedback inputs.

Trace the Z-axis feedback circuit

The successful repair was replacement of the Z-axis rotary encoder. An encoder is not merely a passive position label: its supply, signal channels, reference signal, shield, cable, connectors, and receiving electronics form an active feedback circuit. An internal encoder defect or intermittent connection can disturb the feedback interface, load its supply, inject noise, or present invalid transitions. On an older control, the resulting behavior may appear as a general crash rather than a neatly isolated axis alarm.

1. Inspect the Z-axis encoder body, coupling, connector, and cable for looseness, contamination, crushing, sharp bends, chip damage, or motion-dependent strain.
2. Inspect the full cable route, including every intermediate connector between the machine and measuring-system card.
3. Verify shield termination and bonding against the machine drawing. Correct an open, loose, or improvised shield connection before condemning the encoder.
4. Measure the encoder supply at the encoder and receiving end under operating conditions. Read the required value and tolerance from the encoder documentation or schematic.
5. Observe the feedback channels with suitable diagnostic equipment while moving the Z axis through its travel. Look for missing transitions, unstable levels, intermittent reference behavior, or a disturbance that coincides with cable flex or axis position.
6. Compare crash records with Z-axis motion, position, and direction. A failure while stationary does not clear the encoder because an electrical defect can remain active whenever it is powered.

Do not disconnect a live feedback device unless the machine documentation permits it. An uncontrolled axis response or loss of position reference can create a hazardous condition.

Check before continuing: obtain a repeatable Z-axis feedback anomaly, an intermittent cable or connector result, or a clean power path that leaves the encoder assembly as the leading replaceable element.

Replace the encoder without creating a second fault

Use a replacement compatible with the installed Z-axis feedback interface and mechanical arrangement. Read the exact encoder identity, electrical interface, pinout, coupling arrangement, and alignment requirements from the fitted device and machine documentation. Do not select a replacement from the control model alone.

1. Back up the control and machine data available through the installed system.
2. Place the machine in a safe state and remove the energy required by the machine's service procedure.
3. Mark the coupling position, connector orientation, cable routing, shield termination, and mounting relationship before removal.
4. Fit the replacement without forcing the shaft or coupling. Route the cable away from damage and restore its original strain relief and shielding.
5. Restore power and follow the machine procedure for position reference, alignment, or axis calibration.
6. Jog the Z axis at controlled speed and verify that displayed position and physical direction agree before automatic operation.

The documented machine stopped crashing after the Z-axis encoder was changed, while prior battery and fan replacements had failed. That sequence makes the encoder the confirmed corrective component for this installation.

Check before continuing: verify correct Z-axis direction, stable displayed position, successful reference operation, and no new axis or feedback alarm.

Prove the complete machine cycle

A successful power-up is not a repair verification. The original interval ranged from minutes to weeks, so the observation period must include the operating states that previously exposed the problem and must be long enough to exceed the recent failure pattern.

1. Run repeated Z-axis moves across the working range, including both directions and normal acceleration states.
2. Run the production sequence that formerly crashed the control.
3. Test from a cold start because a failure had occurred about 10 minutes after morning power-up.
4. Monitor control power, encoder feedback status, cabinet temperature, and diagnostic indications during the test.
5. Record cycle count, operating time, and every alarm or reset. Do not count a power cycle as successful operation.
6. After the monitored run, recheck encoder mounting, coupling, connector retention, cable strain relief, and shield connections.

Final check: release the machine only after the complete control stays online, the Z axis references and tracks correctly, and the former production cycle runs repeatedly without a reset. Keep the captured baseline so a future intermittent failure can be compared against known-good power and feedback behavior.

FAQ