Data Recovery Case File · Portable Drives · The One Sound That Means Stop
Hard Drive Making a Scratching Noise: the One Sound That Means Stop
His report was carefully observed, well eliminated, and describes the most urgent symptom in this entire archive. A LaCie Rugged Mini 2TB that "is not being registered by computer. Light flashes then stays constantly on. I have tried multiple cables." Then the sentence that outranks all the others: "makes constant quiet scratching sound then a short sharp loud scratching sound (sounds like an intermittent beep)." This archive decodes clicking, beeping, ticking and silence at length, and treats most of them as diagnoses to be made calmly. Scratching is different. It is the one sound that means stop this second — not after another cable, not after one more attempt to see whether it mounts. Everything else on this page is secondary to that, and the reason is physical rather than cautionary.
| Media | LaCie Rugged Mini 2TB portable hard drive — not registering on the host; enclosure light flashes then remains lit; continuous quiet scratching with periodic sharp scratching audible |
| Reported situation | Drive unrecognised across multiple cables · audible scratching during operation · host machine confirmed working |
| Fault class | Probable head-to-platter contact — active surface damage during operation; immediate cessation of power the determining factor in outcome |
| Equipment used | Laminar flow bench inspection of platter surfaces before any power applied · Atola Insight Forensic assessment · head assembly replaced with matched donor parts · PC-3000 Express with Data Extractor prioritised head-mapped imaging around damaged zones |
The decode: why scratching is different, and what each second costs
What the sound is: a hard drive's heads are supposed to fly on a cushion of air a few nanometres above the platter, never touching it. Clicking means the heads are moving and failing to find what they seek — unpleasant, but the mechanism is still operating as designed. Scratching or grinding means contact. Something is touching a surface that spins thousands of times a minute, and that surface carries the magnetic layer holding the data. His description is precise enough to be worrying in a specific way: a constant quiet scratch suggests continuous light contact, and the short sharp loud events suggest heavier contact on each revolution or at particular positions.
Why every second of power matters: unlike almost every other fault in this archive, this one destroys data while it runs. Contact scores the magnetic layer, and a scored track is not recoverable by any technique at any price — the information is physically removed. Worse, contact generates debris, and debris circulating inside a sealed drive damages further tracks and can wreck the heads entirely. So the difference between a good outcome and a poor one is often measured in how many minutes the drive was left powered after the noise started. That is why this page's advice is unusually blunt: unplug it, and do not plug it in again for any reason, including to demonstrate the sound to somebody.
Why the cable testing was the wrong instinct here — kindly: trying multiple cables is normally excellent practice and this archive recommends it constantly. But it applies to a drive that is silent or simply unrecognised. Once a scratching noise is present, each additional test is another period of powered contact, and the diagnostic value is nil — no cable fault has ever produced a scratching sound. The light behaviour he describes is likewise irrelevant next to the noise: it belongs to the enclosure's bridge board and tells us nothing about the mechanism.
What can still be done: honestly, and without false comfort. The drive will not be powered again in its current state. It is opened under filtered air and the platter surfaces are inspected before anything else, because the extent and position of the scoring determines what is possible. Where the damage is confined to particular zones, the head assembly is replaced with matched donor parts and the drive is imaged with the damaged regions deliberately avoided until last, capturing everything readable elsewhere first. Where scoring is extensive, the honest answer is partial recovery or none, and it is given plainly. The outcome genuinely does depend on how long it ran — which is the one variable an owner controls.
On the bench
The drive was not powered on arrival. It was opened under the laminar flow bench and the platters inspected first: the scoring located and measured, the debris assessed, and the heads examined for the damage the contact had done to them — an inspection that decides the plan rather than following one. The Atola Insight Forensic then assessed the mechanism after the failed head assembly was replaced with matched donor parts, and imaging ran on the PC-3000 Express under Data Extractor with the damaged zones mapped and deliberately deferred — the healthy surface captured first and completely, so that nothing outside the scored regions was ever put at risk by a stall. What the scoring had taken was reported honestly alongside what came back.
The outcome
The readable surface imaged after head replacement, with the scored regions identified and the losses stated rather than glossed. Free assessment, one fixed written figure including VAT, 50% of parts and labour upfront with the balance only on successful recovery. The warning, put as plainly as this archive puts anything: clicking, beeping and silence are diagnoses to be made calmly, but scratching or grinding means the heads are touching the platters — and that fault destroys data continuously while the drive is powered, in a way no technique can undo. Cable testing, further attempts and demonstrating the noise all cost real files. Unplug it at the first scratch, leave it unplugged, and let the surfaces be looked at before anything is switched on again.
Drive making a scratching or grinding noise
Unplug it now — before reading the rest of this. Scratching is not clicking: it means the heads are in contact with the platter surface, and that scores the magnetic layer permanently while generating debris that damages more of it. Every additional second of power costs files that no laboratory can bring back at any price. So don't try another cable, don't try another computer, don't plug it in to show someone the noise, and don't move it about while it's running. Keep it flat and still. Ignore what the enclosure light is doing; it belongs to the bridge board and tells you nothing about the mechanism. Don't freeze it, tap it, or open it yourself. The next step is inspection of the platter surfaces under filtered air before any power is applied at all — because the extent of the scoring, which is determined largely by how long it ran, decides what can still be recovered.
Unplug it and don't switch it on again — call Leeds Data Recovery on 0113 322 3083; surfaces inspected before power, heads replaced, readable regions imaged first and losses reported honestly.
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.