Devices · disks that spin
The first hour decides more than the fault does. Heads, spindle, board, firmware — those are the four things that go, and not one of them heals while a computer keeps trying the disk again. Cut the power, leave it cut, and post it in from Cardiff so a bench can work out which of the four you have.
Describe the symptoms to an engineer
0800 6890668
| What shows up | What has gone wrong | Do this now |
|---|---|---|
| The disk in drive X: is not formatted. Do you want to format it now? | Windows cannot read the file table, so it labels the volume RAW. The data is still down there | Cut the power. Never agree to format it |
| Data error (cyclic redundancy check) | A sector came back with the wrong checksum — usually media on the way out | Abandon the copy; take a full image |
| 0x8007045D — the request could not be performed: an I/O device error | Commands are sent, and nothing answers before the timeout | Power it off and write the model down |
| Windows detected a hard disk problem | SMART attributes have gone past their thresholds | Image it whole; take no files off it |
| 0x800701E3 — a fatal device hardware error | Part-way through a transfer the disk falls off the bus | Stop the retries — this one needs a lab |
| SMART Status BAD, Backup and Replace | Its own self-test has already condemned it | Never boot from it a second time |
| The size shows as 0GB, or the model name is gibberish | The service area will not read, so the drive misreports itself | Firmware work, on the bench |
The diagnosis costs you nothing. Every item that arrives is logged under a case number of its own, and an engineer works out what has actually failed — then tells you plainly which files can come back and which cannot. The price follows that: one fixed figure, in writing, and no work begins until you have read it and said go ahead.
Before any repair is attempted the disk goes onto a hardware imager. It reads at whatever pace the drive allows, stepping around the awkward regions rather than hammering at them, and nothing else happens until that copy is safe.
Heads come from a donor drive matched to yours, a jammed spindle is freed off, and a scorched board hands its ROM chip to the replacement.
All of it happens against the copy. The file system is put back together on that image, and the list of recovered files reaches you before a penny is charged.
Nothing is billed while you are still deciding. A complete listing of what came off the drive reaches you first, and the invoice only follows your yes. Files travel back on media bought new for the job, return postage ours, and we do not close it here until every one of them opens on the machine you will use.
So what are the chances of a disk failing this year? One large public dataset exists — Backblaze's Drive Stats — and it puts 2025 at 1.36% annualised across a fleet of more than 344,000 disks it runs itself, down from 1.57% a year earlier. Drives in Cardiff houses and small offices lead easier lives and fail less often; they also give far less notice.
Nothing had spun inside it for weeks, and two heads were dead. With a donor stack fitted, the first pass off the imager lifted 96%, and the ledgers, invoices and family pictures went home checked.
Rarely. The clicking comes from a head stack that has lost the servo pattern it steers by, and whether the platters below are damaged is a separate question with a kinder answer. Put in a donor head stack, crawl across the surface, and most clicking disks hand back most of what they held. Every restart takes something off that.
Usually it is — just tell us plainly what was done. Once a lid has come off in an ordinary room, no figure is quoted until an engineer has been over the platters, and dust already resting on those surfaces can take a job outside no-recovery-no-fee.
The diagnosis is free, and nothing chargeable starts until a single figure has been put in front of you in writing. Logical faults are the cheapest end of it; donor spares and work in filtered air push a job further up. That number is agreed before any work begins.
Yes. IDE, early SATA and SAS disks turn up on this bench most weeks. Age changes which adapter and which imager come out of the drawer, and nothing else.
A failing drive has only so many starts left in it. Spend none of them. Keep the power off, and the free diagnosis will tell you what is still readable.