A pocket SSD that stopped being recognised anywhere. Inside are a USB bridge and the drive’s own controller, they fail independently, and from the outside they look identical — so establishing which had gone was the whole diagnosis.
A documentary producer carried a pocket SSD between edit suites, holding rushes and working sequences. It stopped being recognised on any machine — no drive letter, nothing in Disk Management, no response across three different computers and two cables.
An external SSD is not a single component. Inside are a storage device and a translator, and each fails independently.
The USB bridge converts between USB and whatever the drive speaks natively — SATA on older units, NVMe on current ones. It has its own controller, its own firmware, and its own power handling, and it is the part most exposed to physical stress because the connector is soldered to it.
The drive controller is the SSD proper: it manages the flash, maintains the translation tables, handles error correction and wear levelling.
A dead bridge with a healthy drive is the commonest external fault and the easiest outcome — the disk comes out and reads normally on a direct connection. A dead drive controller behind a working bridge is a substantially harder job, and one that no amount of trying different cables will distinguish from the first.
There is a quick test before anything is opened, and it is worth knowing because it costs nothing.
A USB device announces itself before any storage is involved. The bridge presents its own descriptors — vendor and product identifiers, a device class, sometimes a serial number — and the host enumerates all of that before it asks a single question about disks. So if the enclosure appears in the operating system’s USB device list while presenting no volume, the bridge is alive and the problem is behind it. If nothing appears at all, the bridge is the suspect.
On this unit the bridge enumerated cleanly and reported itself normally. Everything downstream of it was silent.
The enclosure was then opened and the internal drive connected directly, bypassing the bridge entirely. It did not respond on a direct connection either, which settled the question: the SSD controller had failed and the bridge had simply been faithfully relaying its silence.
Access was established through vendor-specific diagnostic mode, which sits outside the standard command set and allows the reserved service area to be addressed directly. That is where the flash translation tables live, and without them the NAND is undifferentiated data — every page readable and nothing reconstructable, because there is no record of which logical address any given page holds.
The controller responded, but intermittently: it dropped off the bus under sustained reads and returned after a rest. So imaging ran in short bursts with the drive powered down between them, and any region that came back inconsistent across passes was re-read until successive attempts agreed. An unstable controller will return plausible-looking rubbish as readily as it returns an error, which is why agreement across passes is the test rather than absence of errors.
One band of flash proved unreadable rather than merely unstable, and the distinction is worth drawing because it decides whether more effort helps.
Flash stores data as charge trapped behind an insulating layer, and that charge drifts. As it does, the voltage thresholds separating one stored value from another blur, and the controller’s error correction works harder to resolve each read. Modern drives respond by retrying at shifted reference voltages — effectively asking the same cells the question a different way — and this recovers a great deal. But once the drift exceeds what the correction can resolve even with retries, the data is not obscured; it is gone.
Those regions were logged as gaps rather than guessed at. A recovery that silently substitutes zeroes for unreadable flash produces files that open and are wrong, which is worse than a file that is honestly missing.
Video files were played end to end rather than counted. A clip with a gap in the middle will very often open, display a duration, show a thumbnail and stop partway through — so a file listing proves nothing at all about whether the footage is usable.
Affected clips were named individually in the manifest. On an edit that is the only useful form of the information: a producer needs to know which shots are incomplete so they can be reshot, re-sourced or cut around, not how many.
About 96%, including every working sequence and the overwhelming majority of rushes. The shortfall traced to degraded flash rather than to anything done afterwards.
Two things worth taking from it. Send an undetected external complete rather than opening it — on a growing number of models the bridge is part of how the drive is addressed, and on some there is no standard connector inside at all. And stop replugging it. Every reconnection gives an unstable controller another opportunity to commit something unhelpful to its own metadata. SSD recovery is from £300 +VAT after a free 48-hour diagnostic.
Send the device over with a short account of what happened to it. The diagnostic costs nothing and commits you to nothing, and the figure that follows is fixed in writing before anything is opened.
Nothing can begin until the device is on the bench, which makes this the only step that needs anything from you. Pack it properly, put your details in with it, and send it over. What follows is a free diagnostic and a written figure, in that order.
Posting it? Use something tracked and insured — whatever is on the drive is worth considerably more than the postage. Bringing it in? Weekdays, 9am to 5:30pm, and it still wants packing as above for the journey.
Tell us what the device is, what it is doing, and anything already tried — an engineer will come back with a realistic view of the odds and a price band, before you commit to posting anything.
We’ll be in touch shortly. For anything urgent, call 0161 871 0788.
There are two candidates inside the case and they present identically from outside — the USB bridge, and the drive’s own controller. There is a free test: if the enclosure still appears in your machine’s USB device list while showing no volume, the bridge is alive and the fault is behind it. Nothing at all, and the bridge is the suspect.
Better not to. On a growing number of models the bridge is part of how the drive is addressed rather than a passive adapter, so a bare drive can read as unformatted while being perfectly healthy — and accepting the prompt to initialise it at that point is the thing that causes real damage. Send the unit whole.
Not from being carried; there is nothing inside to jolt. What does accumulate is write cycles, and a drive worked hard across several sites gets through them faster than a desktop one. Heat matters more than most people expect too — a compact case with no thermal path holds the controller at its throttling point for as long as a transfer runs.
From £300 plus VAT after a free 48-hour diagnostic and a written quote. Controller-level work carries a 50% deposit with the balance payable only on success.
Start with an instant online quote, or call and talk it through with us first. You'll have a clear, fixed price before any work begins.