Data Recovery Case File · Desktop Externals & Aging Drives · The Drive They Opened to Look

Spinning, but never seeking: the no-motion fault decoded — what removing a drive's cover really costs (and what it doesn't), why unpowered exposure is the survivable kind, and years of machining programs recovered from a 40GB legacy disk

His enquiry came from a working factory floor and contained one of this archive's more forgivable disasters. The patient: a 40GB drive from a CNC machine's control PC running Windows XP. The fault: "the drive spins but the heads don't move to access data." And then the sentence that changes the handling: "someone decided to remove the cover of the drive to see what was happening — not helpful." The context is better than it sounds: "a new drive has been installed and the CNC software is all up and running fine. It would be any saved CNC program files I would be after." So there is no downtime, no production crisis — just an archive of machining programs representing years of setup, proving and refinement, sitting on a drive that has been opened to the room. This page covers the three things that matter: why heads that never move is a different fault from heads that try and fail; what an opened cover genuinely costs (less than folklore says, provided one rule was kept); and why a small legacy drive holding tiny files is an unusually hopeful patient.

The decode: no-seek versus failed-seek, the opened cover, and why small legacy drives help

Heads that never move: this archive has documented drives that spin, attempt a seek, protest and retract — a mechanism trying and failing its calibration. His drive reports something different and more specific: no attempt at all. That narrows the suspects usefully. Most commonly it is adhesion — heads that failed to park properly and are now resting on, and stuck to, the platter surface, immobilising the assembly while the motor turns beneath (or tries to). Alternatively the head circuitry — the tiny preamplifier that drives the assembly — has failed, so no movement command reaches anything; or the drive's firmware has declined to unpark after failing to read the service areas it needs. All three are ordinary professional work; none of them consumes the recorded data. The instruction they share is the same: stop powering it, because a motor turning against stuck heads is the one configuration that turns a recoverable fault into surface damage.

What an opened cover actually costs: now the wince. A hard drive's heads fly over the platters at a height measured in nanometres — a gap so small that ordinary room dust, lint and skin particles are, at that scale, boulders on a runway. That is why drives are opened only under filtered air. But the honest arithmetic is kinder than the folklore, and it hinges on one thing: was it powered afterwards? An unpowered drive with its cover removed collects contamination that a controlled clean can often remove — the platters are still, nothing is being ground into anything. A drive powered with contamination aboard is a different story: heads sweep debris across the surfaces at speed, and scoring follows. His colleague looked; the drive was not run as an experiment afterwards; and that single distinction is why this case has a good outcome available. (The kindly note for the record: curiosity about a mysterious machine is human, and nobody at a factory bench is expected to know about flying heights. The lesson is simply the doctrine — a drive that matters is opened once, by someone with filtered air, and never again.)

Why the patient is a hopeful one: three quiet advantages. The drive is small — 40GB of low-density surfaces, mechanically simpler and more tolerant than modern high-density platters. The files wanted are tiny: machining programs are compact text, meaning that even a partial image typically contains all of them, and they carve and validate cleanly by content. And there is no time pressure, because the machine is already running on a new drive — which allows the patient path: sourcing era-matched donor parts for a legacy model takes as long as it takes, and nothing is gained by rushing a mechanism that only gets one good session.

On the bench

The drive was opened properly this time — under the laminar flow bench — and the exposed platters were inspected first: contamination assessed, the surfaces checked for scoring from any post-exposure run, and the deposits cleaned by controlled process rather than left to be swept by heads later. The Atola Insight Forensic characterised the no-seek fault directly through the drive's power-on behaviour, and the head assembly was serviced with era-matched donor parts until the mechanism would unpark and calibrate. Imaging followed immediately on the PC-3000 Express under Data Extractor, the 40GB banked in one patient pass — and the machining programs were located, carved and validated from the image, each opened and checked for structural integrity before delivery. Years of proved setups, returned to a business that had already moved on without them.

The outcome

The CNC program archive recovered from a contaminated, opened legacy drive and delivered verified. Free assessment, one fixed written figure including VAT; where a drive has to be opened, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The lessons, posted for every workshop and machine shop: heads that never move differ from heads that try — adhesion, head-circuit or service-area faults, all recoverable, all made worse by further power; an opened cover is survivable if the drive is not run afterwards, because unpowered contamination can be cleaned while swept contamination scores; and the most valuable data in a factory is often the least backed-up — machine program files live on the least-considered computer in the building, and a copy belongs somewhere off the machine tonight.

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.