Data Recovery Case File · Desktop Drives · The Tripped Supply

The drive that blows the whistle on itself: power supplies cutting out, decoded

His account had the tidy tragedy of hindsight: a veteran 500GB Barracuda, attached to the desktop and left running whenever the PC was on, felt unusually hot one day — so he switched it off to cool. It never started again. The original supply appeared damaged; a replacement behaved identically — "every time I try to switch the drive on, the power supply cuts out" — with no sound and no sign of life from the disk. His suspicion: the motor. His suspicion was right, and the tripping supplies were how it announced itself.

What a tripping supply measures

A power supply that cuts out the instant a device joins it isn't malfunctioning — it's protecting, and in doing so it's taking a measurement on your behalf: this load is drawing far more current than any healthy version of itself should. On a hard drive there are two classic authors of that draw. One is a shorted component on the circuit board — protection devices failing closed after electrical trouble, a story other pages in this archive tell. The other, fitting his heat-then-cooldown prelude like a glove, is a stalled motor: a spindle commanded to spin that cannot turn converts the command into raw current demand — locked-rotor draw, in the trade's language — and a competent supply slams the door on it within moments. The bench separated the two suspects in minutes: his board, tested in isolation, was electrically clean; the drive's motor, asked to turn by hand-adjacent means it should have laughed at, would not. The hot casing had been the bearing dying loudly in its own way — running hot as it failed — and the cooldown he sensibly performed was simply the moment the worn assembly contracted, settled, and never agreed to begin again. The supplies weren't the mystery. They were the gauge.

The recovery

A seized spindle with healthy platters gets the treatment this situation earns: the data surfaces leave the dead mechanism rather than the dead mechanism being fought. Under laminar airflow the platter stack was migrated to a donor chassis of the same family — moved as a clamped, registered unit, its members' mutual alignment surrendered for not one instant, because on a multi-platter drive that alignment is the data — and seated onto the donor's willing motor. The rebuilt drive spun true on its first properly-supplied start, came ready on the PC-3000, and Data Extractor imaged the ~300GB estate without a single distressed region: the platters, as the whole diagnosis had promised, untouched by everything that had failed around them. Contents verified; delivered on modern media; the veteran and its exhausted bearing honourably retired.

Outcome

Full recovery — and two behaviours worth taking from his story, one to copy and one to stop. Copy: believing the heat. A drive running unusually hot is a drive telling you about friction it shouldn't have; powering it down was right, and the only upgrade is copying the data off before the cooldown rather than after. Stop: auditioning supplies. Once one PSU has cut out on a device, further attempts with other supplies aren't tests — they're repeated stress events against a fault the first supply already measured, and against a genuinely shorted board they can push current through paths that turn a recoverable fault into a burnt one. One trip is a reading. Two is a confirmation. Three is a habit the drive can't afford.

Our case files are drawn from genuine enquiries received by our laboratory over the past ten years, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery procedure our engineers apply to that fault, using the equipment listed.