SSD & flash · bench notes · MHD-2025-6639
The SSD That No Machine Lists.
For a month, a Cox Green researcher's Dell Latitude had been freezing for a minute or so, then carrying on as normal
. One Tuesday morning the screen offered Boot Device Not Found
, and that was that. She tried her partner's laptop, and then a USB caddy; the 850 EVO wasn't listed on either of them
. The IT desk at work said a rebuild was the only option
, which was no use whatever to a half-finished MA and the interviews beneath it.
Much the same fault? Call us.
0800 6890668
In everyday terms.
Between them, those borrowed machines had already done most of the diagnostic work. When nothing at all will list a drive, the fault sits behind the connector rather than in front of it, and on an SSD the candidates are few: chiefly the controller, the small processor that presents the drive to a host and keeps the map that ties every logical block to the flash page holding it. Controllers commonly give out years ahead of the memory they administer, and looked at afterwards, those minute-long pauses were the map coming apart in front of her. Reaching flash without a working controller has nothing in common with platter work. Nothing turns, nothing is mechanical, and the ordinary interface has no comment to offer.
What the bench used here.
How a job is handled →| Tool | Why we used it | What it brings |
|---|---|---|
| PC-3000 SSD | Held the chip in the manufacturer's own service state throughout the dump | Built for SSD controllers alone, with no mechanical element to consider |
| PC-3000 Flash | Matched the V-NAND against its chip library and carried out the raw read | Reads the NAND chips directly, matched against a maker-ID library that is kept up to date |
| UFS Explorer Professional Recovery | Turned the reordered image back into folders, chapters and interview audio | Reads the awkward filesystems others stumble over: APFS, ReFS, XFS, ZFS, Btrfs |
The lab work.
Settle it first: exhausted flash, or a dead front end
It went nowhere near a host port again. All the work happened at controller level on the bench, where start-up kept collapsing at one revealing stage: the moment the translation tables are pulled into memory, before the drive will announce itself at all. Wear counters carried the better half of the news, since this memory still had a great deal of life left in it.
Put the controller into the state its maker holds back
Vendor commands moved the controller into technological mode, a factory service condition that no ordinary host ever asks for. In that state the chip will hand over the raw contents it otherwise declines to serve. Whether a drive of this sort is recovered or written off turns entirely on that one step.
Restore the order before looking for any file at all
A raw NAND dump is disordered by design, since wear-levelling scatters each file across the packages, so the translation had to be reconstructed before any of it resembled data. Once reordered, the NTFS volume opened intact, her folders sitting where she had left them and the chapters still under their own names.
How the job closed.
Her thesis, the interview recordings and the whole reference library returned complete on fresh media. There had never been anything wrong with the Latitude, which now runs on a new SSD. She has promised that the finished thesis will live in three separate places, and we intend to ask.
More of the jobs we've closed.
Filed beside SSD & flash.
Does any of that match your case?
Nothing needs deciding until the diagnosis is back: switch the device off, send it in, and let the findings settle it.