Check Belts, Rails, Fans, and Sensors
Ringing, layer shifts, temperature faults, probing errors, and feed failures can share a visible symptom while having different causes. A loose belt, dirty rail, blocked fan, disconnected thermistor, unstable probe, or runout sensor may be involved, but the symptom is not proof of which component failed. Start with an exact machine record and a non-invasive check before making an adjustment.
Capture the state
Write down the printer model and revision, firmware, recent relocation or collision, usage and material context, enclosure or room conditions, nozzle and plate, and the first layer or job where the problem appeared. Record whether the fault is repeatable, axis-specific, temperature-dependent, triggered by a sensor state, or associated with a recent belt, fan, hotend, or toolhead change. If the machine uses Klipper or another configurable control system, preserve the relevant configuration version and validate the manufacturer’s or firmware project’s configuration checks rather than treating a configuration edit as a mechanical repair.
Use the observation as a routing signal:
| Observation | First component groups to inspect | What it does not prove |
|---|---|---|
| Repeated ringing or layer shift | belt path, pulley/set screw, rail obstruction, motion alignment, acceleration or input-shaping state | a universal belt tension or a single mechanical cause |
| Grinding, skipped motion, or uneven travel | obstruction, bearing/rail condition, pulley position, belt condition, motor or driver state | that tightening a belt is safe |
| Temperature fault or poor cooling | fan debris, fan wiring, thermistor, heater connection, airflow path | that a different fan RPM target applies |
| Probing, homing, or load-cell failure | probe/load-cell surface, wiring, obstruction, gantry or axis state | that the sensor itself is defective |
| Filament or runout warning | sensor connector, debris, feed path, filament position | that the spool or nozzle is the root cause |
Inspect before adjusting
Power down, cool, and secure the machine as the exact manual requires before reaching into an axis, hotend, fan, or connector. Check for foreign objects, loose or displaced fasteners, damaged cables, debris on fan blades, rubbing belts, shifted pulleys, contamination on rails, and connectors that are visibly loose or damaged. Do not run a fan into compressed air or touch a moving or heated assembly unless the model procedure explicitly defines a safe state.
Do not copy a belt frequency, torque, lubrication product, fan threshold, or adjustment direction from another printer. Prusa’s MK4/S and MK3.9/S guidance uses a low-bass-note belt check and model-specific fan monitoring; its self-test gives approximate fan ranges for that family and routes failures to debris, wiring, pulley, belt, bearing, lubrication, heater, or sensor checks. Those values are not portable thresholds. The CORE One L and Original Prusa XL likewise expose separate self-test ladders for axes, fans, heaters, load cells, and filament sensors. A passing check is evidence about the named test on the named machine, not a guarantee of print quality or long-term reliability.
For an Ender-3 V3 SE, the manufacturer support surface lists separate routes for axis motion, leveling, fans, temperature, runout, homing, belts, and linear bearings. It does not turn the page into a complete replacement procedure. Follow the linked service instruction for the exact revision and stop when the required tutorial, tool, wiring view, or endpoint cannot be verified.
Verify the affected function
After a permitted adjustment or cleaning, run the machine’s own homing, probing, self-test, fan, heater, or sensor check when available. Record each component result and any displayed value or error. Then use a small, repeatable test that exercises the affected motion or thermal path. A motion self-test does not prove dimensional accuracy; a fan result does not prove cooling under a full print; and a successful homing sequence does not prove that a shifted belt will remain stable.
Keep unresolved faults in the record instead of compensating indefinitely in the slicer. Use verify a repair or replacement for the post-service ladder and first-layer and Z-offset calibration when the changed component affects first-layer behavior. Stop for smoke, uncontrolled heating, exposed wiring, unexpected motion, repeated failed self-tests, or an adjustment whose safe boundary is unclear for the exact printer.