For most households the tower under the desk holds every photograph the family owns, and for a lot of small firms it holds the books as well. When one stops working, the natural conclusion is that the disk has gone with it, and in nine cases out of ten only the computer has. Getting from a dead PC in Binfield, Crowthorne or Warfield to a padded envelope in the post takes two screws and two cables. After that, the problem stops being yours.
Anything sent in for desktop computer recovery is diagnosed free. The written figure that follows is settled before a screwdriver leaves the drawer: £300 + VAT for any single hard drive or SSD, NVMe included.
No fix, no fee covers logical recoveries. Outside it sit electronic and mechanical failures, chip-level work, DVR jobs and forensic jobs, and physical work is 50% up front. Every band is published on the data recovery cost page.
Reading a symptom back to the failure underneath it is where the work genuinely starts, and this set of thirty accounts for all but a handful of the parcels that reach the Guildford bench. If yours is not on the list, it will still be recognised — describe it on the phone and you will get a straight answer about the odds before you post anything.
A dead power supply and a dead drive look identical from outside the case and are nowhere near each other in cost or consequence. Plenty of towers scrapped this year had entirely readable disks bolted into them. Establishing which component stopped is the first job, and it takes minutes rather than days.
That rhythm is the actuator sweeping across, failing to find the servo pattern it needs and resetting to try again. It is a head fault and restarting never improves it. Each power cycle costs a little more of the recording surface, permanently, and nobody would authorise that trade if somebody itemised it for them beforehand.
This is not an arm hunting for its position. It is metal in contact with the recording surface, and it is the one fault where the difference between switching off now and switching off in ten minutes can be measured in files. Pull the plug. Do not run it once more so that somebody else can listen to it.
Disks fall off the bus for several distinct reasons: firmware that will not load, dead electronics, a controller that has lost the plot, or a SATA cable that has finally had enough. Each leaves a different signature on the bench and they can be told apart in an hour. Swapping parts at home mostly adds a fifth cause to the list.
One surge travels through the power supply and out into whatever it was feeding. Both of those are electronics and neither of them is the recording surface, which is the part worth caring about. The repair is a donor board of the same revision with the adaptive ROM chip moved across from the burnt one, after which the drive answers normally. Board work is invasive, so it takes 50% of the quote upfront and has no place under no fix, no fee.
A fresh installation lands on top of the old data rather than through it, and what decides the outcome is entirely what happens next. Every additional hour of use writes over a little more of the old material. Switch the machine off and the odds stop sliding away from you.
Nothing is scrubbed by that key combination. The reference disappears and the contents sit on the platters until the space is claimed by something else. That is the whole mechanism, and it is exactly why today is a better day to act than next Saturday.
A bad sector is requested, the drive tries and tries to deliver it, and every program on the machine stops until something gives up. Bench hardware does not wait politely like that. It applies a short timeout, moves past the damage, records where it was, and returns to the awkward areas at the end, which is the difference between a capture that finishes and a disk that dies mid-job.
Ransomware work goes to an isolated bench with nothing plugged into a network at any stage. Shadow copies, temporary files, previous versions and unallocated space are all searched before anything else is discussed, and no ransom is ever paid or brokered from this building. It is priced as ordinary media: £300 + VAT for a drive, from £500 + VAT for an array.
Two disks striped across the motherboard controller double the number of ways the set can come apart, and there is no redundancy to fall back on. Both members are imaged in full first. Stripe size and member order are worked out afterwards, from the copies, where getting it wrong costs nothing but time.
Ribbon-cable IDE, the occasional SCSI card, and a wedding that exists in no other form anywhere. Interfaces obsolete for twenty years still connect on the legacy shelf, and mechanisms of that age are read gently, at a speed chosen to suit them rather than to suit a schedule.
The WD10EZEX gives up its PCB the moment the mains misbehaves, usually taking the protection diode with it in the process. A donor board with the ROM chip moved across brings it back, and the platters were never anywhere near the line of fire.
These are what self-builders reach for when the budget is tight, and their head-wear cases are common enough on this bench that a donor slot stays permanently allocated to them. New heads first, then a capture planned around whichever tracks read worst.
Behind the screen of an all-in-one is laptop hardware in a desktop-shaped shell, so what comes out is a 2.5-inch disk or an M.2 stick, packed tight and held in with adhesive. Getting it open is the difficult part of the job. Once the storage is in your hand it behaves like any other drive on this page.
Drives pulled at decommissioning and wanted again long afterwards frequently will not spin at all. Lubricant migrates, heads settle, and the motor cannot break them free. Those are treated as motor cases from the first minute rather than being discovered halfway through a failing capture.
Two habits account for most of these. Modular cables borrowed from a second power supply, which are not interchangeable between makes whatever the connector suggests, and a build done on a carpet with no earth strap in sight. Either will finish a drive off within days. The two causes look nothing alike under a microscope, and electronics work brings most of them back, somewhat chastened.
A disk with its mounting screws loose takes a knock during an upgrade and begins clicking a fortnight later, by which time nobody links the two events. New heads first, then an imaging order planned around the damaged track.
Nothing dramatic ever happens. Every component simply runs warm for thousands of hours, bearings age faster than they should, and then the accumulated wear falls due all at once. Which is precisely how it feels from outside the case, and precisely why it seems so sudden.
A tower that boots from flash gives up silently and skips the warning stage altogether. There is no click to hear and no gradual slowing to notice, only a machine that finds no boot device on Monday morning. That crosses the workshop to the flash bench, where the job is controller and firmware rather than heads and surfaces, and where switching the computer on another twenty times contributes nothing whatever.
Exos and WD_BLACK units in render nodes and CAD machines accumulate duty cycles a household drive never approaches. Once one begins to falter, the capture is planned from measurements: every head is tested, the strong ones do their work first, and the poorest surfaces are attempted last when there is nothing left to lose by trying.
A power supply on the way out will dip under load without tripping anything or announcing itself. Writes complete partially, the file system takes on damage in small instalments, and nobody suspects the PSU. Once the pattern is visible in the structure, the whole thing is rebuilt from a complete image.
Aggressive idle parking, a WD Green habit above all, retires the load ramp while the recording surfaces still have years of life in them. That counter travels in one direction only. Where SMART is reporting a figure with six digits in it, copy the drive rather than watch the number go up.
Ten years in one case with a lead nobody has touched since it was fitted, and then interface errors begin filling the log. Errors of that class point at the link rather than at the recording surface, and a tired connector explains them at least as often as a tired disk does. Checking a cable costs nothing, so it happens before anything expensive is suggested.
A copy living in the same case, on the same supply, in the same room, is not a backup. It is a second copy of one risk. Both disks are recovered together here often enough that the conversation afterwards writes itself. This lab sells no backup service and has nothing whatever to subscribe to, which is why the advice is worth taking at face value.
Out of the old case, onto the floor, never quite into the new one. No computer takes any part in this failure at all. It is still head damage, it still means filtered air and matched donors, and it still costs the same as any other mechanical job.
The connector is staggered on purpose so that ground makes contact before power does. Go in crooked on a live machine and that sequence breaks, which usually costs the drive its board in an instant. Transplanting the ROM onto a matching donor board is the routine answer, and it generally works at the first attempt.
A trading estate building swings a good fifteen degrees between Sunday night and Monday morning, and every swing lays condensation on the boards. The breather filter packs solid, damp settles where it should not, and eventually something shorts. What survives all of that, more often than anybody expects, is the recording surface itself.
A file system needs somewhere to put its own bookkeeping. Run a Windows volume down to zero and the master file table has to grow into whatever scraps remain, scattering its records across the disk. Add one unexpected power loss to that and the index is in pieces. All of it rebuilds from an image, and none of it would have happened had a few gigabytes been held in reserve.
A bargain from an online listing, sold as barely used, with the SMART hours either reset or simply lied about, and dead inside six months. There is nothing interesting about the failure itself. The damage is done by the confidence that came with it, because nobody keeps a spare copy of files held on a drive they believe is nearly new.
Windows can pool several disks into one virtual drive, and when the pool metadata is damaged every member becomes unreadable at the same moment. Each disk is imaged separately, the pool layout is reconstructed from those copies, and the volume is mounted in software. Nothing is written back to the originals at any point.
Of everything that arrives on this bench, a tower is the least trouble. Side panel off, two screws out, power and data leads unclipped, and the disk is sitting in your hand ready for a jiffy bag. All-in-ones are a different matter, being effectively a laptop stood upright behind a pane of glass, and several models resist being opened, so the phone call is better made before the levering starts than after it. Searches for computer solutions and data recovery treat the two as one trade, which they are not: a repair shop exists to get the machine running again, whereas this bench cares about nothing except what was written on the disk. Two local hazards are worth flagging. Much of the older housing round here, plus a fair number of the new-town builds put up in the 1950s and 1960s, still sits behind a consumer unit old enough to pass a good deal of a surge straight down to whatever is plugged in beyond it. And the low ground beside the Thames at Maidenhead and Windsor, along with the Loddon and the Blackwater, takes water more easily than residents tend to expect, which is why a flooded cellar office turns up as a booking most winters. Disks out of either scenario are recovered far more often than their owners assume, subject to one thing: nobody must have connected one to check whether it still works. That single decision destroys more flood-damaged drives than the water itself does.
A slow computer goes to a shop and comes home fast, immaculate and completely empty, the documents having departed with the previous installation. That call comes in most weeks and the situation is far less final than it feels at the time. An installer writes to a fairly small proportion of the total surface. Beneath the new system, the old contents are still sitting there, unreferenced by anything but otherwise entirely intact. The reinstall itself is not the threat. Everything that happens next is. Each hour of web browsing, each update downloaded, each application installed puts fresh data on top of material that could otherwise have been read back. Which is why desktop recovery in this situation comes down to a single decision: stop using the machine today rather than at the weekend, take the disk out, and let a carve retrieve whatever the installer never wrote over.
A tower can arrive carrying anything from a 40GB IDE disk to a Gen5 module, so the desktop bench keeps the full range to hand:
PC-3000 for spinning disks and a separate bench for solid state, covering any age, any capacity and any interface a desktop has ever been fitted with.
Atola and DeepSpar hardware reads weak drives under strict timeout control. Nothing is left running overnight on hopeful software while the disk quietly gets worse in the background.
Head assemblies for clicking tower drives are exchanged inside a laminar-flow bench, using catalogued donors matched on both model and firmware revision.
Controllers and cabling for IDE and PATA, for SCSI, and for the earliest SATA generations. Hardware that went out of production two decades ago is still readable in this building, which matters far more often than the shelf space suggests it would.
The capture is finished before anyone starts looking for a file. Everything afterwards happens against the duplicate, so your disk spins once for us and then goes back into its box.
Nothing in this room is able to write to a disk that belongs to a customer. The block is physical rather than a setting somebody might forget to switch on, and it stays in place from booking in through to despatch.
OptiPlex estates from the offices on Western Road, gaming rigs assembled on a kitchen table in Warfield, and the beige box in the corner that has done nothing but drive the office printer since 2014. Take a side panel off any of them and the same three or four mechanisms keep turning up: a WD10EZEX, an ST1000DM010, an HDWD110. Between them they are responsible for most of the clicking described further up, and for most of the surge damage too. A single tower disk is £300 + VAT, quoted in writing before a screwdriver is picked up, off the back of a free diagnostic that closes 2 working days after the drive is booked in, with a finished recovery running 2 to 4 working days after that. All-in-one machines belong on this page as well, although the parts behind the glass are laptop-sized and the case is a great deal harder to get into.
The iMedia and the Akoya were bought on the same trip as the weekly shop, and their disks arrive here by two routes. Either the tower has stopped starting and the drive is still inside it, or the drive has been lying loose in a kitchen drawer for years with photos, do not lose written across the label in biro. The classic self-inflicted injury on these machines is the factory restore, reached for as a cure for slowness, which lays a brand-new Windows across eight years of family photographs. Nearly all of that survives underneath and comes back off a single full sector image. The older IDE and early SATA mechanisms inside them go onto the legacy adapter shelf and are imaged gently, at whatever pace their age deserves.
Two feeders keep this part of the booking sheet busy, households and decommissioned offices. Pavilions come out of spare bedrooms one at a time, OptiPlex fleets come off business parks ten at a time, and a Presario of genuine antiquity still surfaces every few months. The domestic story is nearly always one of two things: a surge reaching the power supply and the drive electronics together in one instant, or a Windows logo that stalls against a patch of unreadable sectors. The office story runs differently. A disk is shelved during a clear-out, forgotten for four years, then wanted urgently because the old accounts are on it, and after that long standing still the motor cannot break the heads free. Third and most avoidable is the recovery-partition reinstall, reached for as a cure for slowness, which buries documents rather than deleting them and is reversed almost entirely by one full image. Whichever of those applies, the instruction is identical. Keep the tower, send the drive by itself, and it is copied, repaired where repair is called for, and returned on new media.
The storage has to come away from the machine before the parcel is made up, because a bare drive is the only thing that should reach Guildford. If the screws defeat you, ring 0800 689 0668 and it can be talked through on the phone; failing that, most repair counters charge next to nothing for ten minutes with a screwdriver.
One side panel, a data lead, a power lead and two screws, and the disk is free. Getting it out is your job rather than ours, and it saves you the price of posting a steel case halfway round Surrey. Bag the bare drive, add a note with your name and telephone number on it, and send it tracked and insured to Guildford Data Recovery, Building 2, Ground Floor, Guildford Business Park, Guildford GU2 8XH. Reception takes drop-offs in person between 9:00am and 5:30pm, Mon–Fri, and the run from Bracknell is around forty minutes on the A322 and the A3. No collection service exists, so the tower stays with you and only the disk makes the journey.
Nearly every job here arrived as a parcel. Tracked, insured post is the calmest way to move a drive that is already struggling, and something handed over in Berkshire, Surrey or London is normally on the Guildford bench the next working day.
Is the storage still bolted into a machine — laptop, tower, iMac, MacBook, rack server, a DVR under the till? Free it first and send the bare unit. Stripping hardware is not something this lab does, though it is ten minutes' work for any repair shop on your high street. There is a single case with no way round it: memory chips soldered flat onto the mainboard, which is how Apple Silicon machines and certain ultrabooks are built. Where the storage cannot be unbolted, there is no parcel to make up.
↓ Print the shipping & booking-in form (PDF)
The name on the parcel wants to be Guildford Data Recovery. Driving it over from Bracknell is roughly forty minutes on the A322 then the A3; posting it costs you a stamp and a day. Either way, a message goes out to you as soon as it is logged onto the system, and two working days later the diagnostic is finished.
Unsure whether something should go in the box? Ring 0800 689 0668 while the lid is still open, or work through the free online diagnostic and let it tell you.
Nothing to pay for the diagnosis, one written figure before any work begins, and the band on this page is £300 + VAT for any single hard drive or SSD, NVMe included.