Data Recovery Case File · Portable Drives · Case 2,000 · The Site Eight Finale
The Most Dangerous Thing a Failing Drive Can Do Is Look as Though It Has Recovered
The two thousandth file in this archive closes a site that has been about decisions made before the evidence arrives — and it ends with a drive that offered its owner the wrong evidence. A 20TB drive dropped at the weekend, which "makes a ticking noise whenever plugged in, but after around a minute the ticking stops and seemingly goes back to normal." He plugged it in four or five times to be sure. Then he stopped. Nothing told him to.
| Media | 20TB external hard drive subjected to impact — producing repeated positioning attempts on power-up, ceasing after approximately one minute; not presenting content at any point |
| Reported situation | External drive of very large capacity dropped · drive producing a repeated ticking on each connection · ticking ceasing after approximately one minute · drive appearing to return to a normal state thereafter · drive connected four or five times to confirm · drive not used since · nature of the fault unknown to the owner · assistance sought |
| Fault class | Head positioning failure following impact — repeated seek attempts terminating in retirement to a rest position; silence indicating cessation of attempts rather than restored function |
| Equipment used | Cessation of noise interpreted as the drive abandoning its attempts rather than resuming function · no further power-up permitted · laminar flow bench inspection with head and platter assessment before any rotation · matched donor assembly as indicated · imaging planned across multiple supervised sessions against capacity |
The decode: five things this case establishes, at the close of a site
The first — the silence is not the fault resolving. A drive that cannot position its heads attempts the operation, fails, resets and attempts again, on a timer. Each cycle is audible. What ends the noise is not success but the drive exhausting its retry sequence and retiring the heads to their rest position, where they make no sound at all. The quiet drive is a drive that has stopped trying. It is the failure completing, not the failure passing.
Why that reading is so hard to arrive at unaided: everything about the experience argues the other way. A noise that stops is the universal signal of a problem resolving — an engine settling, a machine finishing its cycle. A drive is the one thing that goes quiet when it gives up, and nothing about it announces the inversion.
The second — confirmation is not free, and it is the most expensive habit in this archive. He connected it four or five times "to make sure it wasn't working." That instinct is universal and it is understandable and every repetition cost him something. On a drive dropped days earlier, each power-up spins platters beneath heads that may be damaged, and the answer after the fifth attempt was the same answer available after the first. He bought no information and spent five opportunities.
Why the drive's behaviour actively encouraged it: a device that seems to recover invites another look. Had it simply ticked and stayed noisy, he would very likely have stopped after one attempt — the apparent recovery is precisely what generated the repetition. The symptom recruited him.
The third, and the technical heart — capacity is the compounding factor here. Twenty terabytes is not merely more data; it is more platters stacked in the same housing, more heads on the same assembly, and tracks packed more tightly than any previous generation. Every one of those raises the consequence of an impact and narrows the tolerance for head misalignment. It also changes what a successful recovery requires: a drive of that size takes a very long time to image completely, so it must sustain hours of continuous reading. A drive that gives up after sixty seconds cannot yet do the one thing the recovery needs it to do — which is why the work is head replacement first and capture second, rather than an attempt to read it as it stands.
Why the outlook is nonetheless reasonable: a drive that fails to position and retires quietly has not been dragging damaged heads across the recording. Positioning failure is the fault class in which surfaces are most often untouched, and his stopping preserved that.
The fourth — he made the decision that mattered without any evidence supporting it. Nothing changed between the fifth attempt and his decision to stop. The drive did not get louder or quieter. He simply decided that testing further was not worth it, on no information at all, and that unprompted judgement is the whole reason this case is recoverable. The evidence never arrived. He acted anyway.
The fifth, and the close of this site — that is what every case here has been about. The owner who established what he actually needed before choosing a method. The repairer who restored power without anyone asking whether it was safe to. The man who kept a dead drive for twenty-five years rather than foreclose the option. The machine that ran a repair by default, before anyone could decline. In every one, the decision that determined the outcome had to be made before the information that would have justified it. That is not a failure of the people involved. It is the shape of the problem: storage does not warn, symptoms mislead, and the moment of choice arrives first. Which is why these decisions have to be made on principle rather than on evidence — because the evidence, reliably, is late.
The principle, stated plainly for the two thousandth file: when a device changes its own behaviour without being repaired, assume it has stopped trying rather than started working. Silence is not recovery. And when the urge arrives to check once more whether the problem is real — that urge is the thing to distrust, because on failing storage the checking is the damage.
On the bench
Cessation of noise was interpreted as the drive abandoning its attempts rather than resuming function — a drive unable to position its heads attempting the operation, failing, resetting and repeating on a timer, each cycle audible, with silence following exhaustion of the retry sequence and retirement of the heads to a rest position. Very high capacity compounds impact consequence through platter count, head count and track density, and requires sustained operation for complete imaging. No further power-up was permitted; inspection ran under the laminar flow bench.
The outcome
Cessation of noise read as abandonment rather than recovery, no further powering permitted, and inspection performed under filtered air before any rotation. Free assessment, one fixed written figure including VAT, 50% of parts and labour upfront with the balance only on successful recovery. The decode, for the two thousandth file: your drive going quiet is the failure completing rather than passing. It stopped trying. And you stopped testing it without anything telling you to — which is the decision this whole site has been about, and the reason there is still something here to recover.
When a failing device seems to fix itself
Unplug it and leave it alone — that appearance is the thing to distrust. A drive that goes quiet has usually exhausted its retry sequence and parked its heads, so the silence is the failure completing rather than passing; a machine giving up sounds exactly like a machine that has finished. Resist the urge to connect it once more to be sure, because on failing storage the checking is the damage, and the answer after the fifth attempt is the same one you had after the first. If it was dropped, every power-up spins the platters under heads that may be damaged.
Unplug it and stop testing — call Manchester Data Recovery on 0161 871 0788; cessation of noise read as abandonment rather than recovery, no further powering permitted, inspected under filtered air before any rotation.
Request a quote online →
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.