Data Recovery Case File · Desktop Drives · Mechanical Failure

Clicking hard drive recovery: a 1TB Seagate Barracuda that overheated — then survived four internet remedies

This case file is published with some affection for its customer, who told us everything he'd done with complete honesty. That honesty shaped the job. It's also the clearest illustration we can offer of why the first hour after a drive starts clicking matters more than anything a lab does later.

The enquiry

“1TB Seagate Barracuda — overheated. Really loud clicking at approx 1–2Hz on powering, then settled to normal drive noises. The drive initially mounted fine and I got a little data off. On about the third attempt it wouldn't mount at all. I tried the freezer trick — no benefit. Tried tapping the drive sharply on concrete. Tried opening the drive and moving the heads to a neutral position.”

What actually happened to this drive, in order

The original fault was heat. Sustained overheating degrades read/write heads — the clicking he heard at 1–2Hz was the head-stack failing to calibrate on start-up and retrying, and the fact that it then "settled" and mounted tells us the heads were marginal rather than dead. At that moment, this was a modest job: a weakening drive that would have imaged, carefully, probably in a day.

Each remedy then made a specific contribution. Repeated mounting attempts made a marginal head-stack do the one thing it couldn't afford: more work. The freezer added condensation to sub-micron head–platter clearances — moisture inside a drive is how surface corrosion and head sticking begin. The tap on concrete subjected a precision mechanism to exactly the shock loading that turns weak heads into platter-contacting ones. And opening the lid on a desk placed household dust — boulders, at the scale a head flies — onto the data surfaces, while moving the heads by hand risked slider and suspension damage no drive tolerates.

We describe all this not to embarrass anyone but because every step came from a confident search result. The internet's advice for clicking drives is a museum of things that occasionally worked on 1990s hardware.

The recovery

The drive was opened where it should have been the first time — under laminar airflow. Platter inspection came first, because it decides whether there is a job at all: the surfaces carried contamination but, decisively, no scoring. The platters were carefully cleaned of introduced particulate, the manually-displaced head-stack was inspected under magnification, found unfit, and retired. A donor head-stack from a matching Barracuda — same family and firmware generation, compatible head configuration — was transplanted on the bench.

On the PC-3000 Express, the drive's serial terminal gave us the F3 diagnostics to confirm the drive came ready with sane firmware, and background housekeeping was disabled before imaging so the drive couldn't spend its remaining life reorganising itself. Data Extractor then ran the image head-by-head, prioritising his data partition — the customer had told us which partition mattered, and a file-level map of that volume let us bank the important material in the earliest passes while the transplanted heads were freshest. Remaining passes worked the harder regions with tuned timeouts and retries.

Outcome

The data partition came back at essentially full coverage; the losses were confined to scattered sectors in regions the original heads had already damaged, and to files nobody wanted. Which is the uncomfortable truth of the case: this result was still available after four DIY attempts only because the platters got lucky. The same sequence with a head-crash instead of contamination ends as an unrecoverable drive, and no lab in the world changes that.

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.