Bench notes · Portable hard drives · HUL-2025-0925
A Rattle Inside, and No Machine Sees It.
The drive lived in a toolbox wedged behind the seats of a surveyor’s van, and had done all summer — it comes with me on every job
. Then a barn conversion out past Kilham, with a fortnight of rutted farm track each way, and after that there is a rattle the second it spins up
. She kept working off it for another few days. Then nothing would list it — it is not in Disk Management either
— not her laptop, not the office machine, and I opened a new cable, same result
.
Sounds like yours? Ring us.
0800 6890668
What causes it.
A working hard drive contains almost nothing that moves. Platters on a bearing, and an arm carrying the heads over them. That is it. A rattle at spin-up therefore means something on that list has worked loose: a screw out of its thread, a broken-off part, or play where the heads should sit rigid. Her own testing had done half the diagnosis for us. Two computers, a cable from a fresh packet, the same silence each time: the fault was sealed inside the casing before she went to the post office. And each rotation gives the loose piece another pass at a surface.
The tools this one needed.
The steps we follow →| The gear | What this one did | Why we have it |
|---|---|---|
| PC-3000 Data Extractor | Read the rebuilt drive inside strict limits, one surface at a time | Copying is run at firmware level, with the PC-3000 in charge |
| DeepSpar Disk Imager 4 | Its head map gave us five clean surfaces and one that was scored | Grades each head first, reads one surface at a time, and keeps resets, timeouts and power under control |
| UFS Explorer Professional Recovery | Rebuilt the NTFS tree from the image, folder structure and all | Opens the volume formats most tools cannot: APFS, ReFS, XFS, ZFS, Btrfs |
The stages.
Find out what has come loose
Under laminar flow the lid was lifted, and the fault was obvious. A crash stop had sheared at the actuator and the broken piece was loose in the casing. There was play in the head stack that no drive is built to survive, and one suspension was bent out of line. Each surface was then examined on the stereo microscope: heads with that much movement can strike a platter on any seek.
Fix the mechanics, then map what is left
A single matched donor supplied both the wrecked head stack and the sheared stop. The rebuilt drive was then asked for a map before it was asked for a copy. The answer was plain enough: five surfaces reading clean, and one scored right round its outer edge. That scored surface is the price of the days she kept working after the noise started.
One controlled pass, and then retire it
The read was done in a single sitting. Clean surfaces first. The scored band went last, at a crawl, and the imager was given whatever time it needed. Donor heads buy a window; they mend nothing. The drive did the one job it was rebuilt for and was then taken out of use.
How it finished.
Five of the six surfaces returned intact. The scored one carried part of a church survey, and the report lists each file that did not come back. The archive is held on two drives now, and neither one travels in the van.
What people read after this.
More from the Pocket-size drives file.
Is yours doing the same?
Turn it off. Send it to us. The diagnosis comes first. It says which files are still readable and which are not.