Case record · Desktop externals · SDR-2025-0334
One Disk Clicks, the Other Fades Out.
Two drives arrived in one parcel with nothing else in common. The first had gone off a shelf during a house move: a 2Tb drive that now clicks and won't show up
, which its owner put down to mechanical failure
, correctly. The second was quieter, growing slower and slower and then disappearing from Explorer
mid-copy, which he guessed was the chip or some software thing
.
Seeing the same thing yourself?
0800 6890668
The translation.
He was able to get them correct but did not know what the terminology for the technical aspects was. A click after the fall indicates a head stack which cannot recognise or process the servo burst (this is hardware), and only parts will be able to repair it. On the other hand, slowing down and eventually "disappearing" when under load indicates that it takes longer for the reads than the controller waits (the time-out expires), so Windows drops the device instead of waiting. There are two very different things here. One requires surgery and the other requires increased speed. If you treat both as if they were the same problem, you will destroy what little remains of the dying one while the clicking one sits idle.
Kit used on this job.
What happens in a case →| Which platform | Its role in this case | Why we use it |
|---|---|---|
| DeepSpar Disk Imager 4 | Head map and staged reading on the dropped drive once donor heads were in | Head-map-driven imaging, with per-head timeouts, resets and power-cycle control |
| PC-3000 Data Extractor | Ran the Iomega's read pass with the timeout control Windows never gives you | The PC-3000 platform's imaging unit, commanded by its firmware unit |
| R-Studio Technician | Rebuilt both filesystems from their images and put the data back together | Wide filesystem support, with RAID rebuilds you can trust |
In the lab.
Split them apart and decide what goes first
Each of the drives was evaluated independently and put into a priority list based upon which one would be next to have data extracted from it. Since the click drive was just sitting there doing nothing (unplugged on a shelf), it was stable simply by virtue of that fact and could wait without getting worse. On the other hand, the Iomega drive was losing ground every time power was applied to it, therefore it was run first, since its "window" was the one shrinking.
Copy the fading one before it finishes fading
The failing drive was processed using a very specific limit on how long an individual read operation can take, which is exactly what Windows will not do. Where the machine had timed out and dropped the drive off the bus, the image process simply logged the slower sector and came back to this point later. Once these stalls no longer wrecked the whole transfer, most of the disk was read at a good speed.
Then the mechanical one, slowly and properly
After the dropped drive was given a matched donor stack and a full head map, we made an image of it and worked the stronger heads first. One surface had been scored where the head came down; the rest were sound, and most of the data was retrieved thanks to patience.
The result.
The recovery of these two drives worked out well, but did not recover equal amounts. We got almost everything off the Iomega and lost part of one surface and the client folders on it on the other disk. Two brand-new disks ended up being used in their place and they contained nearly all of the information from the previous two failing disks.
Similar cases on the index.
Also from Desktop externals.
Does this sound like your own drive?
The rule works just like in all of the other cases above: turn off your device, and let a free diagnosis be run before you make any decisions.