Data Recovery Case File · Trust, Practice & Practicalities · The Refund Question
Data first, refund second, honesty throughout: the dead SSD's two claims sequenced — what assessment changes (nothing), what deeper work changes (told before it happens), and four terabytes delivered to the drive that arrived mid-job
His enquiry was cheerful, organised, and asked the practical question this archive is glad to answer in public. The casualty: a Samsung 870 Evo 4TB SATA SSD, "died on Sunday morning." The plan: "could do with having 4TB of data recovered to a new M.2 drive, which hasn't arrived yet but should be coming soon." And the collision he'd spotted before anyone else raised it: "I'll be needing to have the retailer refund the drive as well — so would the drive end in a state where they wouldn't accept it?" A genuinely failed drive earns a genuine refund; a genuinely failed drive full of your data earns a recovery attempt; and the two claims share one physical object. The honest map of that collision — which stages of recovery leave a drive exactly as it arrived, which ones change it, and who decides when — is this case file's real deliverable, right beside the four terabytes.
| Media | Samsung 870 Evo 4TB SATA SSD — completely unresponsive since a Sunday-morning failure · destination: the owner's incoming 4TB M.2 replacement |
| Reported situation | Sudden total failure of the PC's data drive · replacement ordered, in transit · retailer refund intended for the failed unit · the state-of-the-drive question asked up front |
| Fault class | Solid-state electronic failure — non-responsive at the interface; recovery and returns considerations running in parallel |
| Equipment used | Non-invasive diagnostic and controller-level access first · staged escalation gated on the owner's written go-ahead · verified extraction · delivery held for, then written to, the arriving replacement drive |
The decode: two legitimate claims, one honest sequence
The refund, legitimised first: nothing in his plan needed defending. A drive that dies within its cover is exactly what consumer protections and warranties exist for; claiming the refund is not sharp practice, and attempting recovery of your own data from your own failed drive is the most ordinary thing in this building. The only wrinkle is informational, and he'd found it himself: returns processes assess the unit they receive, so its state matters — which makes the recovery plan a matter of sequencing and disclosure rather than dilemma.
What changes the drive, and what doesn't: the honest inventory. Assessment and controller-level work — diagnosis, interface-level access, imaging attempts through the drive's own connector — leave a sealed unit physically as it arrived; a drive that has only been through those stages returns to any counter indistinguishable from one that sat in a drawer. Deeper work is a different category: opening a sealed case, board-level intervention, or lifting memory chips changes the unit's state in ways an inspection can see. The rule this bench runs, stated in his quote before anything began: no stage that alters the drive happens without the owner's explicit, written go-ahead — with the trade-off in front of him at that gate: proceed and prioritise the data, or stop with the drive pristine and the refund's simplicity intact. His data, his drive, his call — made with full information instead of after the fact.
The sequencing rule, generalised: data before returns, always — a refund can wait a fortnight; a returned drive is gone forever, and retailers do not run recovery departments. And disclosure over concealment, always: a failed drive honestly described stays honestly described; nothing about a professional recovery attempt needs hiding from anyone, and no reputable process would coach otherwise.
The recovery: the non-invasive road proved sufficient
Diagnosis found the drive's controller wedged rather than destroyed — unresponsive to ordinary hosts, but reachable through controller-level access techniques at the bench, all of it via the drive's own connector with the unit sealed throughout. The contents were imaged in full and verified; the escalation gate was never needed, and the written record said so plainly for whoever might ask later. Delivery waited, as arranged, for the courier: when his new M.2 drive arrived, the four terabytes were written to it, checked, and handed over — the old drive going back to him in exactly the state Sunday morning left it.
The outcome
His PC's entire estate delivered onto the replacement he'd chosen, the failed unit returned unaltered with a written account of everything done to it — nothing invasive, nothing to declare beyond the failure itself — leaving his refund conversation exactly as simple as it deserved to be. Free assessment, one fixed written figure including VAT, no recovery, no fee. Both claims honoured; neither at the other's expense.
Failed drive with data on it — and a refund or warranty claim pending
Sequence it: recovery first, return second — the counter can wait, and once the drive's handed over, the data conversation is closed forever. Ask any recovery firm the state question he asked, and expect the two-part answer: non-invasive stages leave the unit as received; anything that alters it should be gated on your explicit say-so, with the trade-off explained before it happens. Keep everything disclosed — a genuine failure honestly described supports both claims. And back up the replacement from day one; drives that die on Sunday mornings rarely book ahead.
Both are doable in the right order — call Edinburgh Data Recovery on 0131 202 0491; non-invasive first, your call at every gate, one written figure, no recovery, no fee.
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.