Logical & corruption · job record · BHD-2025-8932
Every Name Present, Every Size Zero.
A portfolio and two assignments were on the stick, and nothing else was. It came out of a laptop in Solihull the afternoon before — I yanked it out halfway through the save
— and by the next morning the folder listed every name, every size at 0 bytes
, with the type column empty and each entry described as a bare File
. A few more hopeful double-clicks, then her ward mentor said leave it alone and let somebody look
.
Sounds like yours? Give us a ring.
0800 6890668
What that means.
FAT keeps the catalogue and the goods in separate places. A short table holds each file’s name, its attributes, its length and the cluster it starts at; the bytes sit elsewhere on the chip and have no idea the table exists. Pull the power partway through an update and that one record is left unfinished — length zeroed, type blank — while every cluster of her coursework stays exactly where it was written. Windows repeats the table without questioning it. That is how a full folder gets reported as an empty one.
What did the work here.
The order of the work →| Equipment | The job it did | What it adds |
|---|---|---|
| PC-3000 Flash | Read the whole chip at sector level, taking nothing the filesystem said on trust | Goes straight to the NAND chips, checked against a maker-ID library that is kept current |
| R-Studio Technician | Rebuilt the broken entries by checking them against FAT’s spare table | Broad file-system support, with RAID rebuilds that can be trusted |
| UFS Explorer Professional Recovery | Carved by file signature wherever an entry was past repair | Handles the file systems other tools give up on: APFS, ReFS, XFS, ZFS, Btrfs |
What we did.
Get a full copy off before touching anything else
Power a stick up and its controller starts tidying after itself, wear levelling and garbage collection, with no idea which blocks a recovery is relying on. So the first job was a flat sector-by-sector read onto bench disks. Once that copy existed, no later attempt could cost her a single file.
Work from whichever allocation table survived better
There are two copies of that table on a FAT volume, and the interrupted write had made a worse job of one than the other. Where the better copy lined up with a directory entry that had survived, the record could be put back whole: the name, the length, the entire cluster chain. Files restored that way are the originals, not something reconstructed to look like them.
The rest had to declare what they were
A handful of entries could not be put back at all, so the files had to speak for themselves. A Word or PDF header spotted among the raw clusters, then the bytes following it joined up in order, and a name given only where the header or the opening lines earned it.
How it closed.
Most of the portfolio opened under the name it always had. A smaller pile came back carved, with machine-made names, and she sorted those by hand over an evening. It all went home on a new stick, along with one rule: wait for the light to stop before pulling anything out.
Other pages worth a look here.
Others under Logical & corruption.
Is yours behaving the same way?
Switch the device off, post it in, and decide nothing until the diagnosis comes back to tell you where you stand.