Devices · memory sticks & flash
An odd amount of what matters most ends up on a memory stick — the sole copy of the accounts, a 'quick' backup still sitting there three years on. Snap one, or kill its controller, and the flash inside normally lives. The chip is where our work starts.
Speak to an engineer about it
0800 6890668
| The wording | What that tells you | First step |
|---|---|---|
| Unknown USB Device (Device Descriptor Request Failed) — Code 43 | No identity comes back from the stick — controller fault | Quit plugging it back in |
| USB device not recognized | The controller never enumerates | Try a different port once, no more |
| You need to format the disk in drive X: | A damaged file system above sound data | Do not format |
| Please insert a disk into removable disk (0 bytes) | The controller can no longer find the NAND | Bench work at chip level |
| Error 0x80070570: file or directory is corrupted and unreadable | Broken FAT/exFAT | Fixable — write nothing more |
| Windows was unable to complete the format | Outright controller failure | Just as well it refused |
As soon as it lands with us, your device gets its own case number. An engineer works out the fault, says what can genuinely be pulled off, then puts one fixed price in writing — no charge for diagnosis, no obligation, and no chargeable work until you approve it.
Chip-off suits older sticks — lift the NAND package, read it alone. A monolith offers no chip to lift: controller and memory are one part, read via test pads.
Controllers interleave the data, scramble it and shift blocks about for wear levelling. We undo that transformation in software so the real files come back.
Nothing is payable until you have signed off the list, and your data returns on new media.
You sign off a complete list of the recovered files first; only then does the recovery fee fall due. Everything returns on fresh media, postage paid, and the job stays open until you confirm the files open at your end.
An honest word on reliability: there are no dependable published failure rates for consumer flash — nobody runs them in fleets and reports the figures the way they do with hard disks. What this bench sees: broken-off connectors account for most physical damage, while dead-stop failures pile up among no-name and giveaway sticks.
Trodden on in a site office, connector ripped clean away. Reading the monolith meant going in via its test pads, with the controller's layout worked out from scratch. Seven years of invoices and CIS records went back before the tax deadline.
Normally, yes. The connector is only the plug — your data sits in the flash chip further along the board. Even a stick broken clean in two often reads fine, provided the chip or monolith wafer has no cracks.
No. That's a controller failing to report, not a blank chip. Reading the NAND directly skips the controller altogether.
Cheap sticks are built from poorer flash and weaker controllers, and a few lie about their size altogether — a tiny chip posing as a large one until your data passes the true limit. Fine for carrying a film; a bad place for anything you hold only once.
The one true dead end is a memory die broken in pieces. Short of that, failed controllers, snapped boards and corrupt tables are all everyday chip-level jobs.
Powering a damaged device up again costs you data. Start a case first; diagnosis is free whatever you decide.