Data Recovery Case File · Portable Drives · Sounds and Codes, Read Together

The drive that almost starts: RPM hunting, a moment of recognition, then Code 43 — one fault wearing two disguises

This enquiry arrived as a well-observed pair of symptom streams. Stream one, the computer's: the Seagate Expansion would connect and show momentarily in Device Manager — then, on any real interaction or simply with time, disconnect into "USB device cannot be recognised," with Code 43 in the device properties ("USB set address request failed"); mounting the bare drive over SATA fared no better. Stream two, the drive's own: on every start it definitely spun up, but through "a few scratching/beeping noises whilst rapidly changing RPM" before finally "settling to a steady speed." A large amount of data aboard; save what you can, he asked — and, with a sore throat, could everything please run by email. It could, and did. Here's how the two streams describe one fault.

Two streams, one translation

Start with the drive's own testimony, because it's the deeper witness. Spin-up on a healthy drive is quick and quiet: motor to speed, heads out, calibration reads, ready. His drive's version — speed hunting up and down, strained noises, an eventual uneasy settle — is a mechanism fighting through that sequence: motor and heads labouring to complete the opening reads that everything else depends on, succeeding barely and late. (The "scratching" word rightly worries people; at the bench it proved to be the sound of that strain rather than of surfaces being scored — a distinction only inspection can make, and the reason the noise is a stop signal either way.) Now the computer's stream translates itself: a drive that eventually stabilises gets far enough to introduce itself — the momentary Device Manager appearance — but a mechanism that exhausted itself starting has nothing left for real requests, and stalls; over USB, the bridge waiting on those stalls eventually fails its own housekeeping with the host, and Windows logs the surrender at the only layer it can see: Code 43, a USB-protocol error that is, here, merely the last echo of a mechanical struggle three layers down. His SATA-direct test — good instinct — removed the USB layer and met the same silence, completing the proof: the messenger was never the problem.

The recovery

Every symptom in both streams argued against further start-ups, and the drive got its next one at the bench: inspection under laminar airflow confirmed the strain's author — a head assembly at the end of its serviceability, labouring over platters that had, per the "scratching" question's happy answer, survived unscored — and a donor stack retired the struggle. On the PC-3000 the drive started the way it used to, and Data Extractor imaged the full terabyte head by head, the once-impossible opening reads now routine, coverage closing in the high ninety-nines. The volume mounted whole from the image; his data — the "large amount" his enquiry weighed — was verified and delivered on new media, the whole exchange conducted by email throughout, sore throat fully accommodated.

Outcome

Full practical recovery — and a listening lesson to file beside the archive's other sound portraits. A drive that hunts — speed wandering, strained noises, a late uneasy settle — is passing its start-up on effort alone, and every additional power-cycle asks for that effort again with less in reserve; the correct count of further attempts, once the pattern repeats, is zero. And on the software side: Code 43 and its "cannot be recognised" cousins are USB-layer messages that faithfully relay failures from far below — worth noting, never worth troubleshooting with driver updates and hub-swapping when the device under them is audibly labouring. He read both streams, wrote both down, and stopped. The recovery was largely a matter of honouring that.

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.