Data Recovery Case File · Desktop Externals & Aging Drives · The Number Is the Diagnosis
When a Drive Reports the Wrong Size
His enquiry contained a figure that does almost all of the diagnostic work. A 3TB Seagate drive "showing on disk manager as unknown, unallocated 3.86GB. Unable to initialise due to I/O error. Heads not stuck. Was a USB external drive but now out of case." That capacity is not a rounding error or a display quirk — a three-terabyte drive is reporting under four gigabytes, roughly a thousandth of its size. A drive reporting the wrong capacity is one of the most specific symptoms in this trade, and it points somewhere precise: the drive cannot read its own internal records, and has fallen back on a default.
| Media | 3TB Seagate hard drive removed from an external enclosure — enumerating and reporting a capacity of approximately 3.8GB; initialisation failing with an input/output error |
| Reported situation | Drive listed as unknown and unallocated · reported capacity approximately one thousandth of actual · initialisation attempted and failing with an I/O error · heads reported free · drive removed from its enclosure |
| Fault class | Service area unreadable — drive falling back to a default capacity; firmware modules requiring assessment in technological mode |
| Equipment used | Assessed on a native connection · PC-3000 Express technological-mode service-area assessment · firmware modules repaired until true capacity was reported · Data Extractor imaging under strict timeouts |
The decode: where a drive keeps its own size, and what a default means
How a drive knows how big it is: not from anything printed on the label. Every drive holds a reserved region — the service area — outside the space users can address, containing its firmware, its defect maps, its adaptive calibration data and the records describing its own geometry and capacity. On power-up the drive reads that region before it can present itself, and what it reports to the computer comes from there.
What reporting a tiny capacity means: the drive could not read those records and has fallen back on a default — a minimal placeholder built into the firmware for exactly this circumstance. So the number he is seeing is not a measurement of anything on his drive; it is the drive saying I cannot read my own description. That is a very narrow finding, and it is entirely consistent with the rest of what he reports.
Why the I/O error on initialising fits: and why it is fortunate. Windows, faced with what it believes is a small uninitialised disk, offered to write a partition structure — and the write failed, because the drive cannot service real operations while in that state. Had it succeeded, it would have written a fresh structure at the start of the disk. The failure protected him.
Why his own observation is useful: he reports the heads are not stuck, which rules out the audible mechanical faults and matches the picture — a drive that spins, identifies itself and reports something has completed the mechanical part of its start-up. The failure is in reading the reserved region, which is below the filesystem and above the mechanism.
Why the data is untouched: the service area is separate from the user area. His files sit in the space the drive currently cannot describe, entirely unaffected. Once the firmware modules are read and repaired in the drive's own technological modes, it reports its true capacity, addresses its media normally, and is imaged conventionally. This is a well-understood fault with a well-understood route.
One note on the drive being out of its case: worth checking rather than assuming. Some external enclosures perform hardware encryption or present a different sector size, in which case a bare drive can behave oddly for reasons unrelated to the fault — so the original enclosure board is worth keeping and testing rather than discarding.
On the bench
The drive was assessed on a native connection rather than through an enclosure, and no further initialisation was attempted — the failed attempt having already protected the start of the disk. Access ran through the PC-3000 Express in the drive's technological modes, with the service area read directly and its damaged firmware modules assessed and repaired until the drive reported its true capacity rather than a fallback default. Imaging then ran under Data Extractor with strict timeouts, and the contents were verified by opening.
The outcome
The service area repaired, true capacity restored and the contents imaged and verified. Free assessment, one fixed written figure including VAT; where a drive has to be opened, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode, for anyone whose drive reports the wrong size: that number is not a measurement but a fallback — every drive keeps its firmware, defect maps and capacity records in a reserved service area, and a drive that cannot read them reports a built-in default instead; so a 3TB drive showing a few gigabytes is telling you it cannot read its own description, which sits below the filesystem and above the mechanism, and leaves your files entirely untouched. The failed initialise protected you.
Drive showing a capacity that's obviously wrong
Take the number seriously — it's one of the most specific symptoms there is. Every drive keeps its firmware, its defect maps and the records describing its own size in a reserved area that users can't see, and it reads that region before it can present itself to a computer. A drive reporting a tiny fraction of its real capacity has failed to read those records and fallen back on a built-in placeholder, so the figure isn't measuring anything — it's the drive saying it can't read its own description. That sits below the filesystem and above the mechanism, and it leaves your data completely untouched. If an attempt to initialise failed with an error, that was lucky: had it succeeded it would have written over the start of the disk. Don't try again, and keep the original enclosure board.
That's the service area — call Newcastle Data Recovery on 0191 406 1051; assessed on a native connection, firmware modules repaired in technological mode until true capacity returns, then imaged out.
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.