Data Recovery Case File · Solid State & Flash · One Success Is Enough
It Opened Once on the Other Machine
Her enquiry contained a verdict, a warning and a fact that outweighs both. A USB pen drive that "will not open on computer or laptop. A retailer says the stick is corrupted. A message comes up each time saying to format the stick, but that it will erase any data on the stick if you do. Opened once on a MacBook, but not again." The retailer's verdict is a reasonable summary of what a shop counter can establish. The format prompt is doing her the courtesy of stating its own consequences. But the sentence that decides everything is the last one: it opened once. A stick that has presented its contents even a single time is a stick whose data is present and reachable, and no verdict of corruption survives that.
| Media | USB flash drive — inaccessible on multiple Windows machines with a format prompt presented; contents accessible once on a Mac and not subsequently |
| Reported situation | Drive failing to open on two Windows machines · format prompt presented on each connection · declared corrupted by a retail assessment · contents opened successfully once on a Mac · no further access since |
| Fault class | Intermittent access with structures readable on some attempts — content present; marginal controller or filesystem region rather than data loss |
| Equipment used | No further connection attempts · imaged write-blocked under per-sector timeouts (DeepSpar USB Stabilizer 10Gb) · weak regions revisited on repeated passes · structures rebuilt on the image · files verified by opening |
The decode: why one success outweighs every failure
What a single successful open proves: everything that matters. For the contents to appear even once, the stick must have accepted power, completed its introduction, reported its capacity, presented a filesystem the computer could parse, and returned directory data. That entire chain worked. Data that has been erased, overwritten or destroyed does not appear once and then stop — so the single success establishes that her files are physically present and that the device is capable of delivering them. Every subsequent failure describes a device that is intermittent, not one that is empty.
Why the two platforms behaved differently: not a coincidence worth chasing, but worth understanding. Windows, faced with a filesystem it cannot parse, offers to format — that is its standard response to any unreadable volume. macOS declines to mount and often shows nothing at all. Neither is diagnosing the stick; they are two different default reactions to the same condition. The Mac happening to succeed once suggests the stick's readability varies between attempts rather than that one operating system is better.
What "corrupted" means from a shop counter: an honest description of what they could establish with the equipment available — a device that will not open. It is not a finding about the data. Retail staff are not equipped to distinguish an unreadable filesystem over intact content from genuine loss, and they would not claim to be.
The prompt to refuse, and it is unusually honest: the format message actually tells her it will erase the data. That is rare and helpful, and it should be taken literally. Formatting writes a fresh empty structure over exactly the region a reconstruction rebuilds from. Her files would usually still be physically present afterwards, but they would come back with no folders, no filenames and no dates — and on a stick that has already shown it can present its contents properly, that would be a self-inflicted downgrade.
Why she should stop connecting it: the harder discipline. Knowing it worked once creates an overwhelming urge to keep trying until it works again, and each attempt powers a device that is evidently marginal. Some of these give a small number of good reads before stopping entirely, and spending them on connection attempts rather than on a controlled image is the one real risk remaining.
What the bench does instead: images under per-sector timeouts, capturing everything readable in a controlled pass rather than hoping for a lucky mount, and revisits the difficult regions patiently — where repeated reads frequently succeed where a single hurried one fails.
On the bench
No further connection was attempted, since a device that has already presented its contents once should have that capability spent on a controlled capture rather than on another hopeful mount. The stick was imaged write-blocked behind the DeepSpar USB Stabilizer 10Gb with per-sector timeouts enforced, and the regions that had read inconsistently were revisited on repeated passes — the technique that matters most on intermittent devices, because a region that fails one attempt commonly succeeds on a later one. The filing structures were then rebuilt on the image rather than the stick, and the files verified by opening.
The outcome
The stick imaged under timeout control with the inconsistent regions re-read, the structures rebuilt and the files verified and delivered. Free assessment, one fixed written figure including VAT; where a chip has to be removed, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode, for anyone whose device worked once: that single success proves the whole chain — power, identification, capacity, filesystem and directory all worked, and erased or destroyed data never appears even once, so your files are present and the device is intermittent rather than empty; a shop's verdict of corruption describes what a counter can establish rather than the state of your data; the format prompt is telling you the truth about its consequences, so refuse it; and stop reconnecting, because the good reads are finite.
Device that worked once and not since
Stop trying to catch it on a good day — that single success has already told you what you needed to know, and every further attempt spends a limited resource. If your files appeared even once, the device accepted power, identified itself, presented a readable filesystem and returned your directory, all of which means your data is physically present. Erased or overwritten data never shows up once and then stops. So the fault is intermittent readability rather than loss, and the job is to capture everything in one controlled pass instead of hoping for another lucky mount. Refuse the format prompt, which on Windows is simply the standard response to any volume it can't parse — and note that when it warns you it will erase your data, it means it. A shop's verdict of "corrupted" describes their equipment's limits, not your files.
Stop reconnecting it — call Glasgow Data Recovery on 0141 404 0294; imaged under per-sector timeouts, inconsistent regions revisited patiently, structures rebuilt on the copy.
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.