Two Fujitsus, Thirty Years in a Drawer

Reading the fault

Nothing fades off a platter. The belief that old data rots is persistent and wrong on any timescale that matters here. Magnetic recording on a sealed drive is stable for decades, and the two drives were as legible in their good regions as they had been the day they were put away. What ages is everything around the recording: the lubricant, the adhesives, the rubber, the electronics.

Thirty years of stillness is the actual risk. A drive at rest has its heads sitting on or very near the platter, and over a long enough period the two can bond. The motor then draws current, produces torque and gets no rotation, which is what the second drive did. Prising that apart by repeated power cycles is how the head suspensions get sheared off, and a sheared head is a much longer job than a stuck one.

The interface is a solved problem, not an obstacle. Drives of this generation speak a protocol modern equipment no longer bothers with, but bench hardware still supports it, so connecting them is routine. What is less routine is being disciplined about capacity limits and translation modes from that era, where guessing wrongly produces an image that looks plausible and is offset from the real layout.

What comes off needs interpreting as well as recovering. A period filesystem stores short names in a form modern tools show accurately but people no longer recognise, and drawing files from that decade are often written by software that no longer exists. Recovering them is the bench's job; being told plainly which formats will and will not open on a current machine is part of the same conversation, and it is better had before the invoice than after.

Bench work

Each drive was handled separately, because two drives from the same office are two different pieces of hardware with two different histories. Both were connected over their period interface on PC-3000 with legacy support, behind a write blocker. The first spun up, readied and read at its original speed with a defect list that had barely grown; it imaged on DeepSpar in one pass and needed nothing else.

The second drew current without turning. One profiled start on a current-limited supply confirmed a stalled spindle rather than a dead one, and it was opened under laminar flow. The heads had bonded to the surface and were eased clear, after which the mechanism turned freely by hand and the surfaces showed no marking under the microscope. It imaged with a small number of slow regions revisited at reduced speed. R-Studio parsed the period filesystems from both images and the recovered files were listed by type before anything was written to fresh media.

What went back

Both drives gave up their contents. The period operating systems were there as expected, and so were the drawings, in two formats of which one opens in current software and one needs a converter the shop has since obtained. Everything went back on a single modern disk as an ordinary folder tree.

Each drive is quoted on its own at £300 + VAT rather than as a discounted pair, because each is a separate piece of work. The one that read straight off needed nothing opened. The seized one had to be, which is mechanical work outside the published guarantee, so 50% of that drive's figure was settled upfront and the balance on a result.

Each of these is a real enquiry, taken from a household or a business somewhere in Leicester or the surrounding county, with the name of the customer, the firm and anything else that would identify them taken out. What is left is the reasoning: what turned up, what had actually failed inside it, the method that fault required and the equipment it was worked on. One of the thirty-six is a job that was refused, and it belongs here every bit as much as the successes do.