Board-level faults · job log · PLY-2025-0603
It Failed at the Moment the Power Returned.
A digger on the site next door found the supply cable feeding a garden centre out on the A386 between Plymouth and Tavistock, which lost its electricity for a morning. Once the power was back, the office desktop would go no further than Default boot device missing
, and the BIOS added only no drive found on either channel
. That machine held nine years of supplier accounts, the plant database, and all the photography the website runs on.
Sounds like yours? Talk it through.
0800 6890668
Why it happens.
The board tells you what happened. A small part sits across the incoming supply for the sole purpose of being destroyed — it exists to fail before anything costly does — and that is precisely what it had done. Bringing a supply back is a rougher business than cutting it, so equipment that sat out a dark morning perfectly happily can be finished off in the first second of the lights returning. None of that reaches the recording. A hard disk stores files by pointing tiny magnetic grains one way or the other. A direction carries no charge, and a spike has nothing there to attack. So a drive that goes missing after an outage is usually a sound archive stuck behind dead electronics.
The kit this job called for.
How the job runs →| The kit | What it did here | Why we keep it |
|---|---|---|
| ACE Lab PC-3000 Express | Tested the replacement board through — model string, SMART log and defect tables each answering | Drops the disk into its maker's service mode, out of reach over plain ATA |
| PC-3000 Data Extractor | Loaded the drive’s own defect maps into the imager, and only then read the surfaces | The PC-3000 runs the copy through the drive's own firmware |
| R-Studio Technician | Opened the NTFS volume within the image and went through every folder | Reads nearly every file system, and rebuilds RAID sets from images |
How it ran.
Establish where the energy actually went
Under magnification a burnt sacrificial part shows up readily, and spotting it decides very little. The real question is where the energy travelled next. A good deal of it frequently gets past that part, and both the motor driver and the interface chip sit downstream, so each was tested before the board was declared sound. Skip that check and the disk expires halfway through a copy.
Transplant the original ROM onto the donor
The correct board revision was tracked down, and it supplied working electronics. What no donor can supply is this drive’s own memory chip, so that was recovered off the burnt board and moved across. Calibration figures written into the ROM at the factory describe these heads above these surfaces and nothing else. Without them you have a board that starts the drive, makes all the right noises, and returns rubbish.
Load the defect map before a single sector is read
The Barracuda, with the donor fitted, gave its name and its full 2TB. Defect tables went into the imager’s configuration ahead of any reading. Nothing had to be asked twice. A surface behaves like that when a board is the only thing that has died.
How it ended up.
Everything was checked and returned on a new disk inside the week. There is a UPS in front of both machines now, the till and the office one, and every night the supplier accounts are copied onto a second drive.
Pages people open after this.
Other power-fault cases.
Seeing the same thing?
Leave it switched off and post it in. Nothing happens until the diagnosis, which lists what can still be read and what cannot.