Flash & SSD · case notes · HUL-2025-0639
The Flash Was Fine. Nothing Could See It.
The Dell Latitude had been locking up for a month. It goes solid for a minute, then carries on as if nothing happened
. A researcher in Newland, and a month is a long stretch to work round. By Tuesday morning all it would produce was Boot Device Not Found
. She moved the SSD to her partner's laptop, then to a USB caddy, and it never showed up in either
. IT at work offered a rebuild and nothing else
, which is no help to an unfinished MA and the interview recordings sitting underneath it.
Sounds like yours? Ring us.
0800 6890668
What causes it.
Every SSD has a small processor doing two jobs at once. It introduces the drive to the host, and it keeps the table recording which flash page currently holds which logical block. That processor is the controller, and it usually gives out long before the memory it looks after is anywhere near worn. Those minute-long stalls were the table coming apart underneath her, though nobody could have known it then. She had narrowed it down herself before the parcel ever arrived: if no host will list a disk at all, the fault lies past the connector, and an SSD has very few parts to blame. Flash work with the controller out of action is nothing like a disc job. Nothing spins. There is no mechanism. The SATA connector stops mattering.
The tools this one needed.
The steps we follow →| The gear | What this one did | Why we have it |
|---|---|---|
| PC-3000 SSD | Kept the chip in its factory service state from start to finish | Made only for SSD controllers; there is nothing mechanical in it |
| PC-3000 Flash | Identified the V-NAND from its chip database and read the raw image | Goes round the controller and reads the NAND itself, matched to a maker-ID library kept current |
| UFS Explorer Professional Recovery | Pulled folders, chapters and recordings back out of the ordered image | Opens the volume formats most tools cannot: APFS, ReFS, XFS, ZFS, Btrfs |
The stages.
Is the memory worn out, or is the controller gone
No host port saw it again after that. Everything was done at controller level, on the bench, and every start attempt failed in the same place: where the translation tables load, which a disk does before it identifies itself to anything. Then the wear counters, and that was the day's one piece of good news — the memory had years left in it.
Getting into the factory's own service mode
The controller had to be pushed into technological mode: a factory condition that no ordinary computer would ever ask for, entered with vendor-specific commands. Put there, the chip surrenders its raw contents; nothing else will persuade it to. Without that one step the drive is scrap. With it, the drive goes home.
Order the pages before looking for anything
A raw dump comes off scrambled, and that is deliberate. Wear-levelling drops successive pieces of a file into whichever package has space at that moment, so the mapping must be reconstructed before anything reads as a document at all. With the mapping solved, NTFS mounted without complaint — folders in the right places, chapters under the names she had given them.
How it finished.
The thesis went back complete on fresh media, with every interview recording and the whole reference library beside it. The Latitude itself was never at fault — a replacement SSD went in and the machine is fine. She says the finished work will live in three places from now on. We intend to check.
What people read after this.
More on SSD & flash.
Is yours doing the same?
Turn it off. Send it to us. The diagnosis comes first. It says which files are still readable and which are not.