A logistics firm in Slough keeps its job files, proof-of-delivery scans and driver paperwork on a 6TB WD drive inside the office machine that everyone maps to. On a Tuesday morning it was simply not there. Their own IT support did the sensible things: another SATA port, a second computer, then an external caddy. Nothing saw it anywhere. They took it out and sent it rather than carrying on.
| Media | A 6TB WD internal SATA drive out of a small office file server, holding job files, proof-of-delivery scans and driver paperwork. It is not detected on any machine, port or caddy. |
| What was reported | WD 6TB internal SATA drive, the machine suddenly stopped reading it, not detected on other computers, not detected through an external caddy, shared office data on it, recovery sought. |
| Fault class | Firmware. A module in the drive's service area had become corrupt, so the drive never completed the handshake that lets a host identify it. The mechanism was healthy and the user area was untouched. Firmware repair is logical work and stays inside the guarantee; a board fault, had it been one, would not have. |
| On the bench | Read directly behind a hardware write blocker on PC-3000 UDMA in technological mode, service area and module structure profiled before anything was brought ready, the corrupt module rebuilt from the drive's own redundant copies rather than from a donor, the drive brought ready and imaged on DeepSpar, the volume parsed in R-Studio from the image, files validated by opening a spread of them and written to fresh media. |
Testing across machines is the right instinct and it produces a narrow answer. Different port, different computer, different caddy, same nothing. That eliminates cables, controllers, drivers and enclosures in one go and leaves the drive. It is genuinely useful work, and it is also where a sensible person should stop.
Not detected is a failure of introduction, not of storage. Before a host can list a drive the drive has to identify itself, and to do that it reads code and tables held in a reserved area on its own platters. If that area is damaged the drive stalls before it ever gets to the part where it has data, and the machine sees an empty port.
Three causes look identical from outside and cost differently. Firmware, board and mechanism all produce a drive that is not detected. They are told apart on a bench, and the distinction matters commercially as well as technically: firmware repair is logical work and is covered by the guarantee, whereas board repair is electronics and is one of the things the exclusion names.
Rebuilding from the drive's own copies is better than borrowing. Service area modules are usually stored in more than one place. Repairing from the drive's own redundant copy keeps its calibration and its translation intact, whereas grafting modules from a donor risks a drive that reads confidently and returns the wrong sectors.
It was read directly on PC-3000 UDMA behind a hardware write blocker, in the mode that talks to a drive's own service area rather than asking it to come ready first. The module structure was profiled before anything was changed: one module was corrupt and the drive was stalling on it every time it was powered.
The module was rebuilt from the drive's redundant copies, after which it readied normally and identified itself correctly. It was imaged on DeepSpar in a single pass, R-Studio parsed the volume out of the image, and a spread of proof-of-delivery scans and job files from across the last three years was opened to prove the recovery before it was written to fresh media.
Everything came back and the office was working from the recovered copy inside the week. The drive itself was not returned to service, which was advice rather than a technical necessity.
A single drive is £300 + VAT regardless of its capacity. The fault was firmware, which counts as logical, so no fix, no fee applied and nothing was payable until the files had been confirmed. Had the diagnostic pointed at the board instead, the terms would have been the other ones, and they would have been quoted before the work rather than after it.
Stop testing once you have tried a second machine. A drive that is not detected anywhere has a fault you cannot reach from a computer, and each further attempt is another power cycle spent on nothing, which matters a great deal if the eventual answer turns out to be mechanical. Do not initialise it if Windows offers to, and do not let anything write a new partition table to a disk that has simply failed to introduce itself. If it is a shared office drive, get the business onto its most recent backup while the recovery runs, rather than waiting: the diagnostic takes two working days from arrival and a finished single-drive recovery is two to four working days after that.
These files are written up from genuine enquiries taken from households and businesses in Bracknell, across Berkshire and Surrey and into London, with names, firms and anything else identifying stripped out. Each one sets down how the fault was reasoned through, the method it actually needed and the equipment it was worked on. One of them is a job that was turned away, which belongs here as much as the rest of them.
A drive taken out and replaced can still be read as a data device once it is out of the machine it was failing in
Heads fly closer to the platter than a speck of dust is wide, so the drop does not have to be dramatic
Where the flash is soldered flat to the mainboard there is no unit to send, and saying so at the enquiry costs nothing
Examining it is free and takes 2 working days from arrival, the figure goes to you in writing and does not move afterwards, and a logical fault carries no fix, no fee. Start online or ring the freephone.