Pocket drives · case notes · CDF-2025-0308
Dropped Once, and Then Only a Tick.
There was no second copy anywhere. The edited JPEGs had gone out to the couples and been cleared off the laptop months earlier, so two seasons of wedding work sat on one pocket drive — the one that slipped from a bag left half-zipped and hit a Chepstow driveway. It does six or seven ticks, then it goes quiet
, and the blue light is on but nothing else happens
. What was on it: the raw files from two whole wedding seasons
.
Does that match yours? Call us.
0800 6890668
Why that happens.
Every platter carries servo patterns written at the factory. Without them a head has no idea which track it is above. Knock the stack askew and those patterns come back as noise, so the arm crosses the disk, finds nothing it recognises, retreats to the park position and tries once more — and that repeating cycle is what you hear as a tick. As for the LED: it is not evidence. Power reaching the USB board lights it whether the mechanism behind is healthy or wrecked. He plugged it in a second time, watched the same routine, and then stopped. That decision to stop is what preserved the archive.
What we used on this one.
Every step, in order →| The tools | What it did here | Why it helps |
|---|---|---|
| DeepSpar Disk Imager 4 | Rated every head before any reading started, so the weak surface went last | Maps the heads first, works surface by surface, and keeps resets, timeouts and power in hand |
| PC-3000 Data Extractor | Ran the post-transplant imaging zone by zone, from within the drive's firmware | Imaging driven from inside the drive's own firmware, by the PC-3000 |
| UFS Explorer Professional Recovery | Rebuilt the exFAT volume out of the image, then opened every CR3 file | Reads the volume formats that defeat most software: APFS, ReFS, XFS, ZFS, Btrfs |
How it went.
Rubber off, bridge unplugged, a plain SATA disk inside
Strip off the rubber jacket, lift out the little USB bridge, and underneath sits a 2.5in SATA disk no different from thousands of others — apart from having met concrete. On our own power the spinning becomes our decision: it turns when there is a reason to turn it, not because some laptop has gone looking for a volume.
Fit the donor stack, then survey before reading
The stack came out of a donor from the same family, running the same firmware revision. Once it was in, the patient reported its proper identity again — and even then nothing was read from it. The survey came first: every head timed and marked for how cleanly it pulled data, so this particular disk set the order rather than a standard recipe. One head was steady. The second was slow and kept faltering, which sent its surface to the end of the queue.
The steady surface first, the shaky one after
Work started on the sound side, timeouts held short and the power cycled at every stall, which meant most of the archive was already banked before the faltering head had been asked to read a thing. Its own platter came last, at whatever rate it cared to give. Bench to sign-off took six days. When a disk has met concrete, working slowly is not a concession; it is how the job gets finished.
Signing it off.
Not one file was signed off until the CR3s had been opened and drawn on screen — a name sitting in a directory listing is worthless to a photographer if the image behind it will not render. The recovered set left on new media, and a duplicate was written before the day was out. Two wedding seasons now sit where they should have sat from the beginning: across two separate drives.
Pages people often read next.
Filed under Pocket-size drives as well.
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.