Stuck on Automatic Repair, With Unreadable Sectors

The decode

Unreadable sectors is the drive naming its own fault. When a hard drive can no longer reliably read parts of its surface, those regions become unreadable sectors — and if they hold operating-system files, the machine cannot boot, which is precisely what sends Windows into an automatic-repair loop it can never complete. The loop is a symptom of physical degradation, not a software glitch to be cleared.

chkdsk on a failing drive is genuinely dangerous, and this is the important part. The command tries to repair the filesystem by writing corrections back to the disk — and forcing writes onto a surface that is already failing to read stresses the mechanism further, can move or destroy still-recoverable data, and sometimes turns a partial fault into a total one. It is designed for logical errors on healthy drives; on a physically dying drive it is close to the worst available first move. The neighbour meant well, but the drive needed imaging, not repairing.

"Detected but files not viewable" fits the picture exactly. The drive still identifies itself over the caddy — enumeration survives long after reliable reading fails — but the filesystem structures needed to list and open files sit partly in the damaged regions, so nothing shows. That is a recoverable state, provided the drive is not stressed further trying to read it repeatedly.

Every boot attempt and every repair pass has been spending the drive's life. A degrading mechanism has a limited number of reads left in it, and the repair loop plus the chkdsk plus the caddy attempts have each drawn on that budget. The single most protective action now costs nothing: stop powering it for access, and let controlled imaging take over.

Imaging inverts the risk that chkdsk embodied. Rather than writing to a failing drive, a hardware imager reads from it under strict control — capping retries, mapping what each head can still manage, banking the readable majority first and returning to the bad regions last. The readable bulk of his files is secured before the drive is asked to strain for the difficult remainder, which is the opposite of what a repair command does.

The honest ceiling is set by the surface, and named per file. Sectors that never read leave holes; files crossing them come back damaged or incomplete, and which files those are is geography. The result is an honest listing of what recovered cleanly against what the bad sectors claimed — determined from the image, reported before the customer is surprised by it.

On the bench

The drive's health was assessed on PC-3000 UDMA, reading the sector-error and reallocation counts behind the repair loop. It was then imaged under a hardware write-blocker on DeepSpar Disk Imager with per-head mapping, tight retry limits and skip-and-return ordering, the readable bulk secured first and the degraded regions revisited last. Filesystem structures were rebuilt from the image in R-Studio, damaged areas carved, and the recovered files validated by rendering.

The outcome

Drive health assessed, imaged under retry control with the readable majority banked first, and files rebuilt from the image with an honest per-file listing. 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: the repair loop and the chkdsk were the drive telling you its surface is failing, and the command run against it was the wrong medicine. Imaging reads from a dying drive under control instead of writing to it — which is what keeps your files reachable.

Our case files are written up from genuine enquiries our lab has handled for customers across Sheffield and South Yorkshire, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery approach our engineers apply to that fault, using the equipment listed.