Data Recovery Case File · Desktop Drives · The DIY Escalation

Three boards, one smell of smoke: the Hitachi whose heads couldn't be judged until its ROM came home

Few enquiries arrive with this much evidence. After a power fault killed his 1TB Hitachi Deskstar, the owner had bought two donor drives and a brand-new PCB, tried each board in turn, and filmed the results: the original board producing a worrying smell and smoke, and both replacements leaving the drive clicking. His conclusion — "looks like the head needs replacing" — was reasonable, widely shared on every forum he'd have consulted, and premature in a way that this page exists to explain.

The enquiry

“It stopped working after a power shortage, after which it was making a clicking sound. I bought two donor drives to help repair it, and a new PCB board. Unfortunately it did not work — it looks like the disc head needs replacing. I've made videos: with the original PCB there was a funny smell and smoke; with the donor board it's clicking; with the new PCB it's also clicking.”

Why clicking on a donor board proves nothing yet

The gap in the home experiment was eight pins wide. As other cases in this archive explain from different directions, every drive of this era carries its individual factory calibration — the adaptives — in a small ROM chip on its board, tuned to its specific heads and platters. A donor board runs donor adaptives; a brand-new board runs nobody's. Ask either to fly this drive's heads and the heads fail their opening reads — and a drive whose heads fail their opening reads clicks, in perfect health. So his three videos, read at the bench, said only this much for certain: the original board was electrically destroyed (the smoke settled that), and the drive clicks on boards that were never going to work. Whether the power fault had also reached through the electronics into the heads — entirely possible with a surge — was a verdict that had to wait until the drive could be tried with its own calibration aboard. Order of operations is the whole discipline: ROM first, judgement second.

The recovery — his purchases, redeemed

The original board's ROM chip had survived its board's dramatic death — they usually do — and was transferred under the microscope to the best of his donor boards. On the PC-3000, with its own adaptives home again, the drive got its fair trial: and this time the click was a verdict. Even correctly calibrated, the heads would not read — the surge, as feared, had reached the preamplifier and beyond, and the smoke-scented board had taken company down with it. Which is where his other purchase came good: one of his donor drives, checked at the bench for genuine family-and-generation compatibility, gave up its head-stack under laminar airflow. Rebuilt on transplanted heads and its own ROM, the drive came ready at last; Data Extractor imaged it in the usual patient order, and the volume came off the image intact and verified.

Outcome

Full recovery — reached through a two-stage fault his experiments had half-mapped, using parts he'd half-selected, in a sequence that made all the difference. Credit where due: he stopped at the right boundary, filmed instead of forcing, and his donor-buying instinct saved sourcing time even though his diagnostic conclusion jumped a step. The step, for everyone following his path at home: a drive clicking on a foreign board hasn't been diagnosed — it's been asked the wrong question. And a board that has ever smoked should never see power again, however curious you are about video number four.

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.