Data Recovery Case File · Cameras, Drones & Cards · Half-Answering Is Its Own Fault

Memory Card Reporting as a Faulty Disk

His enquiry described a card that is neither working nor entirely dead, which is its own category. Around 50GB on a microSD card that "my phone stopped reading — the computer doesn't recognise it either. Comes back as faulty disk. There's no physical damage to the actual card and it's currently in a USB stick." What he wants back is the family videos and photographs. Two things worth separating. The card is sitting inside a USB adapter, which is an untested variable and a genuinely common culprit — that costs nothing to eliminate and should be done first. And the message itself is informative: a card reported as faulty rather than absent is a card that answered the introduction and then failed to deliver, which is a different fault from silence and, in one specific way, a more hazardous one.

The decode: the adapter first, then what half-answering means

Eliminate the adapter before anything: a microSD inside a USB adapter is two devices in series, and the adapter is the cheaper and more commonly faulty of the two. It has its own controller, its own contacts, and a card socket whose springs wear. A failing adapter produces exactly this picture — a device that appears and misbehaves — while the card inside is in perfect health. Testing the card directly in a native microSD reader, or in a different adapter, costs nothing and occasionally ends the whole matter. It is worth doing before any conclusion is drawn about the card, and worth mentioning to anyone quoting for the work.

Why the phone stopping first is a clue: the sequence he describes — the phone stopped reading it, then the computer called it faulty — suggests the failure travelled with the card rather than with any one host. That leans away from the adapter, though it does not rule it out, since the card may have been moved into the adapter after the phone failed.

What "faulty" means as distinct from "absent": a card that is entirely dead does not appear at all — the computer cannot distinguish it from an empty slot. A card reported as faulty has done something: it answered the electrical introduction, was accepted as a device, and then failed to provide valid data about itself or its contents. That is partial controller response, and it means the controller is alive enough to talk and not alive enough to work.

Why that is more hazardous than silence: the counter-intuitive part, and the reason for urgency. A silent card invites nothing — no software can scan it, so nobody tries. A half-answering card invites everything: it appears in disk tools, so scanners will attempt it, operating systems will offer to repair or format it, and phones will offer to set it up again. Each of those runs a struggling controller through further work, and a format offer accepted at this point would be genuinely destructive. The hazard is not the fault itself but everything the fault tempts people into.

What the recovery involves: access through hardware that manages instability rather than inheriting it — timeouts enforced so a stall cannot hang the operation, dropouts absorbed and resumed — and a single prioritised image taken while the controller still responds at all. Everything else happens on that copy. Where the controller fails entirely during the attempt, the memory is read at chip level past it, which for a card of this type is well-established work. Family video and photographs are then validated by playing and rendering them rather than by counting files.

On the bench

The adapter was eliminated first, on principle — the card tested directly in a native reader so that the fault could be attributed to one device rather than two. With the card confirmed as the source, access ran behind the DeepSpar USB Stabilizer 10Gb, write-blocked, with timeouts enforced and resets absorbed so that a half-responsive controller could not stall or be pushed further by repeated attempts. A single prioritised image was taken while the device still answered, and the filesystem was reconstructed on the copy; the chip-level route on the PC-3000 Flash stood ready had the controller given out. The family photographs were validated by rendering and the video by playback before delivery on fresh media.

The outcome

The adapter eliminated at no cost, the card imaged in one managed pass while it still responded, and the family media validated and delivered. 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 with a card called faulty: test it in a native reader before blaming the card, because an adapter is a second device with its own controller and contacts and is the commoner failure of the two; a card reported as faulty rather than missing has answered and then failed to deliver, which means the controller is partially alive; and that state is more hazardous than silence, because it tempts scanners, repair offers and format prompts — all of which spend a struggling controller and one of which is destructive.

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.