Portable hard drives · bench notes · PLY-2025-0925
A Noise on Spin-Up, Then No Drive at All.
A surveyor kept the drive in a toolbox behind the front seats of her van, where it had ridden all summer — it goes wherever I go
. Then came a barn conversion out beyond Peter Tavy, two weeks of rutted farm track twice a day, and after that it rattles the moment it starts spinning
. She carried on working off it for a few more days. Then no machine would show it — Disk Management cannot see it either
— not the van laptop, not the office desktop, and a cable out of the packet changed nothing
.
Sounds like yours? Talk it through.
0800 6890668
Why it happens.
Almost nothing in a hard drive is meant to travel. Discs spin on a spindle; an arm carries the heads to and fro above them; that is the whole of it. A rattle at spin-up therefore narrows to a very short list — a screw backed out of its thread, a component that has sheared away, or slack where the head assembly should be rigid. Her own testing had answered half the question before the parcel went off: two separate computers, an unopened cable, the same silence every time, which put the fault inside the sealed casing. Meanwhile each revolution gives that loose piece one more run across a surface.
The kit this job called for.
How the job runs →| The kit | What it did here | Why we keep it |
|---|---|---|
| PC-3000 Data Extractor | Copied the repaired drive under tight limits, surface by surface | The PC-3000 runs the copy through the drive's own firmware |
| DeepSpar Disk Imager 4 | Head map: five surfaces answering cleanly, the sixth scored | Scores every head before it starts, works one surface at a time, and holds resets, timeouts and power in check |
| UFS Explorer Professional Recovery | Put the NTFS directory tree back together from the image | Mounts what other software will not: APFS, ReFS, XFS, ZFS and Btrfs volumes |
How it ran.
Establish what had worked loose
Under laminar flow the lid was taken off, and there was no mystery about it. The crash stop had sheared away at the actuator and the broken end was loose inside. The head stack carried far more play than any drive is built to tolerate, and one suspension had been pushed out of line. Each surface was then put under the stereo microscope, since a head with that much movement in it can touch a platter on any seek.
Put the mechanics right, then map the damage
A single matched donor supplied the ruined head stack and a sound crash stop together. Rebuilt, the drive was made to give up a map before anyone asked it for files. The answer was unambiguous: five clean surfaces, and one scored right around the outside. That sixth surface is what carrying on for a few days after the noise started has cost.
One pass under control, then retired
The copy was taken at one sitting. The good surfaces came first. The scored band went last and was read at a crawl, the imager given as long as it asked for. Donor heads only open a window; they repair nothing. The drive completed the single job it had been rebuilt to do, and was then withdrawn.
How it ended up.
Six surfaces went in and five came back whole. Part of a church survey sat on the scored one, and each file lost from it is set out by name in the report. Two drives hold that archive now, and neither of them goes out in the van.
Pages people open after this.
Other cases from the pocket-drive file.
Seeing the same thing?
Leave it switched off and post it in. Nothing happens until the diagnosis, which lists what can still be read and what cannot.