Spotting Fan Bearing Wear Before Hash Rate Drops
Early acoustic and thermal signals that a field-replaceable cartridge is failing on high-hash ASIC units.
Hash rate often falls only after a fan has struggled for days. On rigs rated well above 0.05 TH/s, thermal protection firmware pulls clock before a bearing fully seizes. Catching wear early keeps output steady and avoids emergency weekend visits.
Listen at consistent PWM
Run your monitoring stack at a fixed fan duty cycle for five minutes during a quiet shift. A healthy cartridge produces steady airflow noise. Intermittent ticking or a rising whine usually means bearing play—not dust, which tends to muffle rather than whine.
Compare inlet delta-T across neighbors
Place the same temperature probe location on adjacent units mining the same pool job. A drift of more than 6°C above row average, with unchanged ambient, points to airflow loss. Log readings daily; sudden slope change beats a single snapshot.
Inspect latch wear and shroud gaps
Field-replaceable modules depend on clean latch engagement. A cartridge seated at an angle passes air around the shroud rather than through the heat sink. Re-seat once; if delta-T persists, schedule swap before PWM hits 100% continuously.
When to call for swap
If noise progresses over three shifts or PWM pegs at maximum while delta-T still climbs, treat the module as failed-in-progress. Continuing stresses VRM thermals even when hash has not yet dropped.