Board-level faults · bench notes · HUL-2025-0603
It Broke When the Power Came Back.
A digger on the site next door went through a supply cable, and a garden centre out on the A1079 between Hull and Beverley lost its power for a morning. It came back, and the office desktop stopped at Default boot device missing
. The BIOS was shorter about it: no drive found on either channel
. On the machine behind that screen: every photograph the website uses, the plant database, and supplier accounts going back nine years.
Sounds like yours? Ring us.
0800 6890668
What causes it.
Look at the board and you can see what happened. One small part across the incoming supply is there to be destroyed — that is its entire job, to go before anything expensive goes — and it had done exactly that. Restoring a supply is a rough business, rougher than switching it off, so kit that sat through a dark morning quite happily can be finished in the first second of coming back. None of it gets near the recording. A hard disk keeps files as the direction that magnetic grains point in. There is no charge in a direction, and nothing there for a spike to attack. So a drive that has gone missing after an outage is normally a sound archive locked behind a dead board.
The tools this one needed.
The steps we follow →| The gear | What this one did | Why we have it |
|---|---|---|
| ACE Lab PC-3000 Express | Proved out the replacement board — model string, SMART log and defect tables all answering | Reaches the vendor service commands an ordinary ATA connection cannot |
| PC-3000 Data Extractor | Imaged the surfaces afterwards, with the drive’s own defect maps loaded in first | Copying is run at firmware level, with the PC-3000 in charge |
| R-Studio Technician | Mounted the NTFS volume inside the image and worked through every folder | Handles almost any file system, and puts RAID sets back together |
The stages.
Work out where the energy really ended up
Under magnification a blown sacrificial component is easy enough to spot, and spotting it settles almost nothing. Where the energy travelled afterwards is the real question. A fair amount of it often gets past that part — the motor driver and the interface chip are both downstream — so each of those was tested before the board was passed as sound. Miss that step and the drive dies partway through the copy.
Move the original ROM onto the donor board
A donor board of the right revision gave us electronics that worked. It could not give us the drive’s own memory; that had to be recovered from the dead board and moved over. Inside the ROM are calibration figures taken at the factory, and they describe these heads over these surfaces and no others. Put a board on without them and you have a drive that starts, sounds right, and reads gibberish.
Put the defect map in before reading a sector
Once the donor was on, the Barracuda gave its own name and its full 2TB. Its defect tables went into the imager’s setup first, and only then did the read begin. Not one sector needed a second attempt. That is what a surface looks like when nothing worse than a board has failed.
How it finished.
All of it was checked and returned on a new drive before the week was out. A UPS now stands between the mains and both machines, till and office, and the supplier accounts are written to a second disk every night.
What people read after this.
More of the power jobs.
Is yours doing the same?
Turn it off. Send it to us. The diagnosis comes first. It says which files are still readable and which are not.