Data Recovery Case File · Cameras, Drones & Cards · The Tool Waits Forever
The Card That Hangs Disk Management
His enquiry was well-reasoned and reached a sensible conclusion. A 128GB SD card that "seems to be totally dead. Disk Management hangs when it's connected, as does Device Manager. I've tried a couple of different readers and PCs, same results. My assumption is the firmware is probably corrupted. There's no physical damage." His elimination is thorough and his conclusion is close to right. But the symptom he has described is more specific than "dead", and worth separating — because a card that hangs the tools is not the same as a card that is absent. A device that is not there causes nothing to happen. A device that responds and then stops answering makes software wait, and that distinction points somewhere particular.
| Media | 128GB SD card — causing system disk and device management tools to hang on connection across multiple readers and computers; no physical damage present |
| Reported situation | Card connected causing management utilities to freeze rather than report an error · behaviour reproduced on several readers and machines · no physical damage · firmware corruption suspected by the owner |
| Fault class | Controller responding partially then failing to service requests — host tools blocking indefinitely; memory typically intact behind the fault |
| Equipment used | No further host testing · enumeration attempted write-blocked under strict timeouts (DeepSpar USB Stabilizer 10Gb) · chip-level read past the failed controller (PC-3000 Flash, Rusolut VNR) · contents reconstructed and validated |
The decode: why hanging is informative, and why his conclusion is nearly right
Why an absent card causes nothing: when nothing is connected, or when a device fails to power at all, the system simply carries on. No entries appear, no tools freeze, nothing waits. So a card that produces no reaction whatsoever is a different symptom from his.
Why his card hangs things: because it is answering, at least at first. Enough of the card is alive to be noticed — power is accepted, the controller begins the introduction — and then it stops responding partway. The host has already committed to talking to a device, so it waits. Disk Management and Device Manager both enumerate storage as part of building their view, and neither has a facility to abandon a device that has started answering and gone quiet. They block, and the window freezes. So the hanging is not a Windows problem: it is the precise signature of a controller that is partially functional rather than dead.
Why that is a hopeful finding: it places the fault firmly at the controller — the chip conducting the introduction and managing the memory — rather than at the memory itself. His assumption about firmware is in the right area: controllers run their own code and maintain their own translation tables, and corruption there produces exactly this behaviour, a device that starts and cannot continue. The NAND holding his files sits behind that controller and is normally entirely unaffected.
Why his testing is complete: several readers and several machines reproducing identical behaviour eliminates readers, ports, drivers and operating systems. There is nothing further to establish, and each additional attempt hangs another machine while powering a failing controller.
What must not be done: two things specifically. Any tool that offers to initialise or format a partially-responding card, because a format that succeeds writes over the structures a chip-level reconstruction rebuilds from. And repeated attempts to force detection — some people work through readers for days, and on a controller in this state that is spending margin for no information.
How it is actually read: not by coaxing the controller back. The memory is accessed directly through the card's internal pads, bypassing the failed chip entirely, then descrambled, error-corrected and rebuilt through its translation layer in software until the filesystem stands up.
On the bench
No further host testing was attempted, since several readers and machines had already exhausted what that establishes. Enumeration was attempted write-blocked behind the DeepSpar USB Stabilizer 10Gb under strict timeouts — equipment that does not hang when a device stops answering, which is the difference between gathering information and freezing a computer. With the controller confirmed unable to service requests, the memory was read at chip level on the PC-3000 Flash with the pinout confirmed against the Rusolut VNR database, then descrambled, error-corrected and reassembled through its translation layer. The contents were validated before delivery.
The outcome
The controller confirmed as partially responsive, the memory read at chip level past it, and the contents reconstructed and validated. Free assessment, one fixed written figure including VAT, 50% of parts and labour upfront with the balance only on successful recovery. The decode, for anyone whose card freezes the tools: hanging is a specific symptom rather than a general one — a device that is truly absent causes nothing at all to happen, while one that answers and then stops makes software wait indefinitely, because system tools have no way to abandon a device that has started responding; that signature points at a partially functional controller, which is where firmware corruption shows itself, and the memory behind it is normally intact; several readers and machines is a complete elimination; and never let anything initialise or format a partially-responding card.
Card that freezes Disk Management or your file manager
Stop connecting it — you've already learned everything a computer can tell you, and each attempt hangs a machine while powering a failing card. Read the symptom as good news rather than as "dead": a device that isn't there causes nothing to happen at all, whereas one that hangs your tools is answering at first and then stopping, and the freeze is your system waiting for a reply that never comes. That's the signature of a controller that's partially alive — often a firmware or translation-table problem — rather than failed memory, and the NAND holding your files normally sits untouched behind it. Don't work through more readers, don't run recovery software, and above all don't let anything initialise or format a card that half-responds, because a format that succeeds writes over exactly what a chip-level recovery rebuilds from.
That's a controller, not your data — call Glasgow Data Recovery on 0141 404 0294; enumeration attempted under strict timeouts, memory read at chip level past the fault, contents validated.
Request a quote online →
Our case files are drawn from genuine enquiries received by our laboratory over the past ten years, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery procedure our engineers apply to that fault, using the equipment listed.