The plug came out of the wall while the machine was saving. The recorder shut down mid-write, and afterwards the drive inside it was thought to have a stuck head. The project on it was not backed up anywhere. Sudden power loss during a write can leave two separate marks, and both need looking at rather than assuming one or the other. The first is mechanical: a drive that loses power without warning does not park its heads the polite way, and on an older mechanism the heads can end up resting on the platter and bonding there. The second is logical: whatever was being written was left half finished, so the structure describing the recording is inconsistent. The recording itself is almost always fully present on the platters. What is broken is the mechanism's ability to start, and the bookkeeping around the last few minutes.
| Media | Maxtor hard drive fitted in a hard-disc recorder. Mains power removed mid-save; the machine shut down without parking. Head suspected stuck. |
| Reported situation | Hard-disc recorder · fitted with a Maxtor drive · mains plug knocked out mid-save · machine shut down mid-write · head thought stuck · project recording sought. |
| Fault class | Power loss during a write. The mechanism may be seized with heads resting on a platter, and the recorder's filing structure may be left half written. The recording is normally intact on the platters and is reached by freeing the mechanism under laminar flow and rebuilding the interrupted structure from an image. |
| Equipment used | Drive taken out of the recorder and put on a bench supply with the current limited and profiled, spin-up attempt and any stiction assessed by current draw · opened under laminar flow where the heads were resting on the surface, heads freed or the platters moved across to a matched donor base, alignment set afterwards · imaged on DeepSpar with PC-3000 UDMA driving the drive · the recorder's filing structure parsed and the half-written section rebuilt in R-Studio, the recording validated by playback. |
A hard drive parks its heads on a ramp at the edge of the platter, or on a landing zone, when it is shut down properly. Losing mains power removes that opportunity. Modern drives use the spinning platters as a flywheel to generate just enough current to retract the heads, but older mechanisms are less reliable about it, and a head left on the surface can bond to it over time. That is stiction, and it presents as a motor that will not turn.
The way to tell stiction from a dead motor is current draw. A seized mechanism pulls hard and stalls, and the drive gets warm without any rotation. That is a five-minute test on a bench supply with the current limited and it is worth doing before anything else, because the answer decides whether the casing has to come off.
The second mark is logical and much easier. Recorders write in their own way, often with a proprietary layout, and an interrupted write leaves the index describing the recording inconsistent with what is actually on the disk. The material is there. The bookkeeping is not. Parsing that from an image and rebuilding the missing section is ordinary work once the drive can be read.
The mistake to avoid is repeatedly powering the recorder to see whether it comes back. If the heads are bonded to the platter, the motor trying to break them free can tear the surface, and that damage is permanent and precisely where the last recording was written. One or two attempts are informative. A dozen are destructive.
The drive came out of the recorder and was put on a bench supply with the current limited, where a spin-up attempt and the current draw settled whether the mechanism was seized or simply not being asked properly. Where the heads were resting on the surface, the drive was opened under laminar flow and the heads freed, with the platters moved across onto a matched donor base where the bearing or the surface required it, and alignment set afterwards. It was imaged on DeepSpar with PC-3000 UDMA driving the drive. The recorder's filing structure was then parsed from the image in R-Studio, the half-written section rebuilt, and the recording validated by playing it rather than by trusting a file size.
The drive read out of the recorder, the mechanism freed, the interrupted structure rebuilt, and the project recovered onto fresh media. Assessment free, and one fixed written figure with VAT in it beforehand. Because the casing had to be opened under laminar flow, this ran as physical work with 50% due before that stage and the balance owed only on the result. Worth adding for anyone with a recorder of this sort in a rack: they are computers with hard drives in them and they deserve the same treatment, which means a mains lead that cannot be kicked out and a copy of anything that matters.
Do not keep switching it on. If the heads have settled onto the platter, each attempt is the motor trying to tear them loose, and that damage lands exactly where your most recent work was written. Take the drive out of the recorder and send that alone rather than posting the whole unit — machines are not stripped here, and a bare drive travels far better than a rack box does. Say clearly that power was lost during a write, because that changes both the first power-up approach and the way the filing structure is rebuilt afterwards. And fit a small uninterruptible supply if the machine matters; they cost less than one recovery and they exist for precisely this.
Each file below comes from a genuine enquiry taken from households and firms across Aberdeen, Aberdeenshire and the wider north-east, anonymised so nobody can be identified. Each one sets out how the fault was reasoned through, what was done about it, and which equipment did the work.
Twenty-odd years old and refusing to mount is a filesystem or drive fault, not an empty disk — and the old Mac structures still parse perfectly well
No warning means no long decline, and no long decline usually means the data is all still there behind whatever stopped the drive presenting itself
An IT firm that stops and refers a failing drive on has done the right thing — the drive is imaged in measured passes, not read until it dies
Assessment is free and takes 2 working days at most once it lands, the figure is fixed in writing, and logical work carries no fix no fee. Begin online, or lift the phone.