Data Recovery Case File · Solid State & Flash · Seen, Then Refused
It Introduced Itself and Failed Immediately After
This enquiry quotes an error precisely, which makes it unusually easy to answer. A USB memory stick that has failed, where "the system says 'device cannot start' in the device list." That message is not a generic complaint about a missing drive. It is reported for a device that was successfully detected and then failed to become operational — which means the stick got partway through its introduction and stopped, and that places the fault quite precisely.
| Media | USB flash drive — enumerated by the host and listed in the device inventory with a start failure reported; no volume presented |
| Reported situation | Stick failing without warning · listed in the host device inventory · host reporting that the device cannot start · no volume presented to the file manager · contents required |
| Fault class | Enumeration succeeding with initialisation failing — controller unable to complete start-up; memory readable past the controller |
| Equipment used | No driver reinstallation or repair attempted · enumeration and initialisation state read write-blocked under strict timeouts (DeepSpar USB Stabilizer 10Gb) · chip-level read past the controller where required · translation layer reconstructed in software · files validated by opening |
The decode: the two stages, and which one failed
What has to happen when a device is connected: two distinct stages. First enumeration — the device draws power, announces itself on the bus, and states what it is, so the host knows something is there and what to load for it. Then initialisation — the device is brought into service, its parameters read, its capacity established, and its storage made available to the system.
Why the message is specific: that error is reported when the first stage succeeded and the second did not. The host knows the device exists, has identified it well enough to list it, and then could not bring it into a working state. A device that failed at enumeration would not be listed at all — it would simply be absent. So being listed with that error is a more advanced position than nothing appearing.
What it places the fault on: the stick's controller. These devices run their own firmware on power-up, reading internal structures before they can report capacity and serve requests — the block mapping, the wear tables, the configuration. Where those are damaged or the controller has partly failed, it can complete the introduction and then be unable to do anything further. That is precisely this message.
Why the memory is very likely intact: a controller failing to initialise has not touched the stored data. The memory devices hold what they held; the component that knows how to interpret them has stopped working. Nothing has been erased and nothing overwritten.
Why the fix people try first cannot work: the obvious response to that message is to reinstall drivers, uninstall the device and rescan, or try a different machine. All reasonable, all futile — the failure is inside the device, and the host's software has nothing to do with it. The same message will appear on every computer.
Why a stick is a better prospect than a solid-state drive with the same fault: because of what is inside. A stick carries a single memory package and a relatively simple controller, and the pattern linking addresses to physical locations can be reconstructed from the memory contents themselves. So the memory can be read directly through the manufacturer's test points, bypassing the failed controller entirely, then descrambled, error-corrected and reassembled in software. That route is routine on this class of device and is why controller failures on sticks are usually recoverable.
What to stop: further connection attempts and driver work, neither of which can change anything.
On the bench
No driver reinstallation or repair was attempted, the failure being internal to the device and unaffected by anything on the host. Enumeration and initialisation state were read write-blocked under strict timeouts behind the DeepSpar USB Stabilizer 10Gb, confirming that the device completed its introduction and failed to be brought into service — which places the fault on the controller rather than the memory. The memory was then read at chip level past the controller, descrambled and error-corrected, and the translation layer reconstructed in software. Files were validated by opening.
The outcome
The failure located at the initialisation stage, the memory read past the controller and the contents reconstructed and validated. Free assessment, one fixed written figure including VAT; where a chip has to be removed, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode, for anyone seeing that message: it is specific and it is not the worst news. Connecting a device has two stages — announcing itself, then being brought into service — and that error means the first succeeded and the second failed. So the controller cannot complete its start-up, while the memory holds everything it held. Driver reinstallation cannot help, because the fault is inside the device.
System reporting that a device cannot start
Stop reinstalling drivers and trying other machines — the fault is inside the stick and nothing on the computer can change it, so the same message will appear everywhere. What that error actually reports is quite specific and reasonably encouraging. Connecting a device happens in two stages: it announces itself on the bus and states what it is, then it's brought into service with its capacity established and storage made available. That message means the first stage worked and the second didn't, so your stick is being seen and identified — a device that failed earlier wouldn't be listed at all. It means the controller can't finish its own start-up, which has nothing to do with your data. On a stick the memory can usually be read past a failed controller.
Drivers won't fix it — call Newcastle Data Recovery on 0191 406 1051; enumeration and initialisation state read under strict timeouts, memory read at chip level past the controller, files checked by opening.
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.