Missing or corrupt files · lab log · CDF-2025-1113
It Fell. Nothing Seemed Wrong.
In Cwmbran a shelf bracket above a desk gave way mid-backup, and the My Book landed with both leads still in it. On it was a family's whole video collection, camcorder transfers included. Nothing afterwards looked wrong: it powers up exactly like it used to
, there's no strange sound coming off it
, and Windows will happily offer Eject My Book
. No drive letter ever appears.
Does that match yours? Call us.
0800 6890668
Why that happens.
A drive letter sits at the end of a queue, and every step in front of it has to work: motor and heads, then the bridge board in the case, then the partition table, and the filesystem last of all. The first two were sound here — the unit announced itself and it spun up to speed — so the queue stalled at step three. That is the mark left by a write that was cut off halfway. It was the records that broke, not the mechanism. Enclosures of this kind carry a second rule. The decryption is done by the small board inside the case, which is why disk and board arrive at the bench as a pair and are never parted. Split them and the sectors you read back are noise.
What we used on this one.
Every step, in order →| The tools | What it did here | Why it helps |
|---|---|---|
| ACE Lab PC-3000 Express | Signed off heads, motor and platters before anything logical was considered | Opens the maker's service commands, which a plain ATA link will not pass |
| DeepSpar USB Imager | Copied it through its own bridge board, keeping decryption in the chain | Ordinary imaging speed, working through the USB bridge |
| R-Studio Technician | Mended the partition table and the NTFS records on the image | Reads nearly every file system, and rebuilds arrays you can rely on |
How it went.
Prove the mechanics survived the fall
Sounding right proves nothing by itself. Spin-up current, per-head read rates, both defect lists and the reallocation counters were all measured, and every figure landed where it should. Twenty minutes with the meters turned an alarming job into a routine one.
Image it with the bridge in place
The key for a My Book lives on the little USB board in the case, so the copy was made with the board still in the chain, and it read cleanly from sector zero. Take the naked disk out, put it on a SATA port, and you get a perfect image of nonsense: every byte there, and not one of them usable.
Do the repairs on the copy, not the disk
Seen from the copy, the damage was small and recent: a partition entry that stopped part-way through, at the second both plugs came out of their sockets. It was rebuilt on the image, with nothing ever written back to the drive, and the volume opened at the first ask — folders and dates just as they were.
Signing it off.
Nothing had gone astray. The number of files on the copy matched what the folders claimed. Disk and bridge were kept together as one item, first measurement to last, and with hardware like this that habit is most of the job.
Pages people often read next.
More on corruption and lost files.
Is that what yours is doing?
Power it down, get it to us, and hold off on any decision until the diagnosis says what can still be read.