The parts cost less than the labour, and on most modern designs there is nothing separable to work on. What matters is getting the data off — and how that is done depends on something you cannot see from the outside.
A repair utility is trying to produce a working stick, and the quickest route to that is frequently to erase it. A recovery is trying to produce your files. The two aims collide at exactly the wrong moment.
It is what people search for, and being clear about it saves a wasted afternoon.
The components are surface-mounted on a board the size of a stamp, frequently sealed in resin, and a replacement costs a few pounds. There is no economic version of repairing one, and no technical version either on most modern designs.
What can be done is getting the data off, once. That is a different job with different odds, and how it is done depends on something you cannot see from the outside: whether the stick is built from separate components or as one fused package.
Superficially identical sticks can present completely different prospects.
A controller chip and one or more NAND packages on a small board. If the controller fails, the flash can be desoldered and read directly, then reassembled in software. Common on older and higher-capacity sticks.
Controller and flash fused into a single package with nothing separable inside. There is no chip to remove — access is through internal test points, which needs the correct pinout for that exact part.
Independent of both, and the best odds of any USB fault. The break is at the solder joint and the memory on the other side of it is almost always untouched.
A stick reporting far more than it holds, with firmware that accepts writes into nothing. Works perfectly until real usage passes the true size, then loses everything beyond it.
Worth noting exactly what yours does before contacting anyone.
The file system header is damaged or the controller is struggling. Your data is almost certainly still there — decline the prompt and stop using it. Formatting writes through the component that is failing.
The controller is crashing and being reset by its own watchdog after hitting flash it cannot read. Every reinsertion runs the same loop. Retrying achieves nothing and risks worse.
Electrically silent, no drive letter, nothing in Disk Management. Either the controller has failed outright or the connector has broken internally — and the latter is the better of the two.
A logical fault. The flash is fine and the file system is not, which on healthy hardware is often recoverable with read-only software installed elsewhere.
This is where the time goes on the harder jobs, and why two identical-looking sticks differ by days.
Pulling data directly off flash — by desoldering a chip, or reading through test points — does not produce files. It produces a raw dump that has to be de-interleaved across the flash planes, corrected for ECC, and unscrambled according to the arrangement that particular controller used.
Only after all of that does anything resembling a file system appear, and only then can files be carved out. That work is manufacturer-specific, which is why a stick that still enumerates and simply will not mount is a far quicker job than one that is electrically silent.
And a point worth making plainly: flash is not archival media. Sticks left in drawers for years develop read errors from charge leakage — the insulation holding electrons in each cell is very good and not perfect. If something matters, it should be on at least two devices that get powered up occasionally. USB sticks and cards in Manchester are from £250 +VAT after a free 48-hour diagnostic.
Not economically, and not in a way that returns it to service — the components are surface-mounted on a tiny board and a replacement costs a few pounds. What can be done is recovering the data off it once, which is a different job.
No. That message usually means the file system header is damaged or the controller is struggling, not that your data has gone. Formatting writes through the component that is failing and makes recovery considerably harder.
Almost certainly not. The break happens at the solder joint and the memory on the other side is generally untouched — snapped connectors have the best odds of any USB fault. Send the broken piece too.
The controller is crashing and being reset by its own watchdog after hitting flash it cannot read. To the computer that looks identical to unplugging it. Repeated reinsertion runs the same loop and achieves nothing.
One where the controller and flash are fused into a single sealed package, standard on most modern sticks. There is no separate chip to remove, so recovery goes through internal test points using the correct pinout for that exact part.
No. Flash holds data as electrical charge behind an insulating layer, and that charge drifts away slowly — faster in older or warm-stored cells. Anything you want in five years should be on at least two devices that get powered up from time to time.