Data Recovery Case File · Desktop Drives · Electronics & Firmware

Hard drive PCB swap not working: a Hitachi Deskstar 7K1000, a matching donor board, and a lost ROM chip

This customer did almost everything right. The drive wouldn't power up, he identified a board-level fault, sourced a donor PCB with matching part numbers, and made a clean swap. The drive then spun up, appeared in the BIOS — and showed nothing at all. One 8-pin chip, mislaid during the swap, was the whole story.

The enquiry

“The original PCB failed — no power — so I swapped it with a matching donor board. During the swap, the 8-pin ROM chip got mixed up or lost from the original board. Now the drive spins up normally and is detected in the BIOS, but nothing is accessible — no partitions, files, or readable data. I believe the adaptive data may need to be rebuilt or extracted from the platters.”

He'd diagnosed his own situation correctly, which is rarer than you'd think. What he was missing was a way out of it.

Why a matching board isn't a matching board

Every hard drive leaves the factory individually calibrated. Head fly-height parameters, preamplifier gain settings, per-head tuning values — the adaptives — are measured for that one drive's specific heads and platters and written into non-volatile memory on the PCB, on this Hitachi family an 8-pin serial chip. Two boards with identical part numbers and identical main controller ICs still carry different adaptive data, because they were calibrated against different mechanics. A donor board with its own donor adaptives will happily power the spindle and answer an identify command — the behaviour this customer saw — but the moment the drive tries to position its heads and read the service area using another drive's calibration values, it finds noise where its firmware should be.

On most PCB failures the fix is routine: move the original 8-pin chip to the donor board, and the drive gets its electronics back without losing its identity. Here that chip was gone. What remained was the harder version of the job: give the drive a working board and reconstruct adaptive data it could actually fly on.

Rebuilding what was lost

The drive went onto the PC-3000 Express in the HGST utility, running through the donor board in technological mode. The critical fact that makes this recoverable at all is that Hitachi drives of this generation do not keep everything on the PCB — substantial firmware content, including copies of calibration and configuration data, lives in the service area on the platters themselves. The problem is circular: you need workable adaptives to read the service area, and the adaptives you want are in the service area.

The way through is iterative. Starting from the donor board's own NVRAM contents as a baseline, our engineers adjusted the adaptive parameter set step by step, testing service-area readability at each stage, until the heads began returning clean reads on individual SA copies. From there the surviving on-platter data was used to refine the reconstruction — effectively teaching the donor electronics to fly this drive's heads rather than its own — until the drive loaded its firmware overlays, reported its true capacity and came ready with a stable translator.

Once the drive was ready, no chances were taken with it. It went straight to Data Extractor for a full sector-level image — a drive running on reconstructed adaptives is a working drive, not a trustworthy one — and all recovery work thereafter happened against the image, never the hardware. The NTFS volume mounted from the image, and the customer's file set was extracted and verified.

Outcome

A drive that was two separate failures deep — dead original electronics, then lost calibration data — came back to a full image and a verified copy of its contents. It's worth being clear about the order of luck involved: the platters and heads were healthy throughout, which is what made an electronics-and-firmware recovery possible. Had the customer kept powering the drive with mismatched adaptives for weeks, heads flying at the wrong height over good media could have turned this into a far more serious job.

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.