Case log · Pocket hard drives · CDF-2025-0925
Now It Rattles, and Nothing Will See It.
The Expansion had spent an entire summer in a toolbox jammed behind the front seats of a surveyor’s van — it travels with me to every job
. Then came a barn conversion out towards Rudry, and a fortnight of rutted farm track to reach it, after which the drive starts rattling the moment it powers up
. For a few more days she kept using it anyway. Then nothing would show it at all — Disk Management does not list it either
— not the laptop, not the office PC, and I got a new cable out of the packet, no difference
.
Does that match yours? Call us.
0800 6890668
Why that happens.
Count the moving parts in a working drive and you run out fast: platters spinning on a bearing, and an arm that swings the heads across them. Nothing else in there moves. So a rattle at start-up can only mean one item on that short list has come adrift — a screw backed out, a part snapped off, or slack where the heads ought to be held rigid. Her own testing had already done half our work. Two computers, a cable straight out of its wrapper, the same silence each time: whatever was wrong was sealed inside the casing before the parcel went in the post. And every rotation hands the loose piece another go at the surface.
What we used on this one.
Every step, in order →| The tools | What it did here | Why it helps |
|---|---|---|
| PC-3000 Data Extractor | Kept the read after the rebuild inside tight limits, one surface at a time | Imaging driven from inside the drive's own firmware, by the PC-3000 |
| DeepSpar Disk Imager 4 | Its head map showed five surfaces reading clean and a sixth that was scored | Maps the heads first, works surface by surface, and keeps resets, timeouts and power in hand |
| UFS Explorer Professional Recovery | Put the NTFS structure back together off the image, folders included | Reads the volume formats that defeat most software: APFS, ReFS, XFS, ZFS, Btrfs |
How it went.
See what is loose in there
Under laminar flow the lid came off, and the answer was there immediately. A crash-stop had sheared away at the actuator and its fragment was rolling free in the casing. The head stack had play in it that no drive is designed to survive, and one suspension sat bent out of true. Each surface was then checked on the stereo microscope, since heads with that much play can touch down on a platter during any seek.
Rebuild the mechanics, then map what survives
One matched donor covered both the wrecked stack and the snapped stop, and the rebuilt drive was first put to work producing a map rather than a copy. It came back clear enough: five surfaces reading clean, and a sixth scored the whole way round its outer edge. That sixth surface is what the extra days of use after the noise began ended up costing her.
A single measured pass, then out of service
It was read out in one unbroken sitting. The clean surfaces went first; the scored band was held back to the end and taken at a crawl, the imager given all the time it wanted. Donor heads buy time, they do not mend anything. The drive did the one job asked of it and was then retired for good.
Signing it off.
Five surfaces came back intact. What the scored one took with it was part of a chapel survey, and the report named every file that did not make it. That archive lives on two drives now, and neither of them travels in the van.
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.