Two diagnoses arrived with this stick, and both survive contact with the bench. On any computer it displayed "unavailable"; a local shop had "used software to try and recover the data", got nowhere, and gave two pieces of advice — "send it to a specialist", because the trouble was probably in "the microchip or nanochip inside the USB". Loose vocabulary, sound instinct. Recovery software runs on the computer and must reach the data through the stick's controller; when that controller has failed, software is a man shouting questions through a locked door. 'Unavailable' is the controller failing to do its one job — and the answer isn't better software, it's reading the memory chip directly, which no shop counter can do and every proper flash bench does weekly.
| Media | A USB stick reading 'unavailable' on every computer — shop software attempts unsuccessful, a chip-level fault suspected, specialist referral given. |
| Reported situation | USB stick · shows 'unavailable' everywhere · shop tried recovery software, no result · shop suspects the chip inside · referred on · files wanted. |
| Fault class | Controller or chip-interface failure beyond software's reach — the stick half-answers, enough to register as 'unavailable', not enough to serve data. Contents ordinarily intact in the NAND; reached by direct chip reading. More software passes are wasted wear. |
| Equipment used | Stereo microscope examination of controller and memory; the NAND lifted under hot air and read in full on a specialist programmer, the failed controller bypassed; its data arrangement reconstructed and reversed; files carved by content in PhotoRec, opened to check, delivered on fresh media. |
'Unavailable' is one of the stranger states a stick can occupy: more alive than invisible, less alive than working. It means the controller still musters enough function to answer the computer's first knock — so a device registers — but collapses on every request after that. Recovery software meets this and can do precisely nothing, because software has no private route to the memory: every read it requests is passed to the same controller that's failing, which declines it the same way. The shop's software didn't fail through weakness. It failed because the road it must travel is closed.
The shop's parting theory — the microchip inside — lands close enough to the truth to deserve credit. The failure is at chip level in the sense that matters: it belongs to the silicon inside the stick, not to anything a computer can influence from outside. And the referral itself was the professional move. Knowing what your tools can't reach, and saying so before the customer spends more time and hope, is the difference between a good shop and a stubborn one. This was a good shop.
The specialist route the shop pointed at works because it abandons the closed road entirely. The memory chip is separated from the failed controller and read raw on a programmer — every page, no permissions asked. The controller's proprietary scattering of the data is then reconstructed and reversed, and the files are carved from the rebuilt whole by content, each opened to prove itself. 'Unavailable' was only ever the controller's opinion. The chip was never consulted.
Board surveyed under the stereo microscope, the controller's failure confirmed and the memory checked clean. NAND off under hot air, read completely on the programmer. The scrambling reconstructed, the logical image rebuilt, PhotoRec carving the files out by content — then each file opened, checked and written to fresh media for return, with the stick itself retired from service.
The road software couldn't travel, bypassed altogether — the files off the chip the shop correctly suspected. Nothing owed for the verdict: free diagnostic, reported inside two working days of arrival, then one fixed written figure. Chip-off is physical work, half payable up front, balance only on proven success. And if you're a shop reading this: referrals like that one save your customers' data — keep making them.
Believe them, and go one better: stop the software attempts entirely, including your own. Each pass powers a failing controller and achieves what the last pass achieved — nothing, at cost. Don't reformat 'to see if that clears it' (it can, along with your reason for caring), and don't open the casing; chip work begins with a microscope, not a thumbnail. Post it as it stands, with a line about what's on it. The shop got you halfway by knowing its limits. The bench covers the rest.
Every case file on this site is written up from a genuine enquiry handled for customers across York and North Yorkshire, anonymised to keep client confidence. Each one sets out the diagnostic reasoning and the recovery method our engineers apply to that fault, on the equipment named.
The pin is the doorway, not the data — rebuilt where the board allows it, read at chip level where it doesn't
The chime is the doorbell; the controller never answers — so the memory is read directly, past it
Half-seen is not half-recovered — the computer's nagging loop ends at chip level, past the broken part
Free diagnostic inside 2 working days of arrival, one fixed quote, no fix no fee on logical faults — start online or ring the freephone.