When a computer quietly stops listing a drive, that drive is normally still holding every file ever put on it. Not the lucky outcome, the usual one. Parcels arrive at the bench from Bracknell town centre and the Lexicon, from the office parks strung out along the A329M, from Wokingham, Crowthorne, Sandhurst and Binfield, and from Reading and Ascot on either side of them. No money changes hands until the fault has a name and a single fixed figure has been put in writing in front of you.
Anything sent in for hard drive 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.
The arm swings out hunting for servo bursts, locks onto none of them, resets and goes again. What you are listening to is a head that has stopped reading. Every pass takes a small toll on the coating, so the sensible number of power-ups after the first click is none.
Something is now in contact with a surface whose outer edge is moving at motorway speed. Coating comes away as dust, and that dust then works as an abrasive on everything it passes. This is the one symptom where minutes genuinely count. Pull the plug and leave it out.
A hand on the casing feels nothing at all. Three unrelated faults produce that silence: a spindle seized in its bearing, a head stack resting where it should never rest, and a board no longer passing power through. They need three different repairs, and swapping leads at a desk will not tell you which one you have.
The motor reaches speed, the controller goes looking for its firmware modules on the platters, fails to get a clean copy of them, and stops the drive to protect what is left. Mechanically there may be nothing wrong whatsoever. The repair happens through a service terminal.
It draws power and answers, then identifies itself as a row of symbols or as nothing at all. That is the firmware zone talking rather than the recording surfaces. No downloadable utility has a route into a firmware zone, and none ever has.
The translator, which converts the addresses your computer asks for into physical positions on the platters, has collapsed. So the drive answers honestly and says it holds nothing. Rebuild the translator on the bench and the real size comes back, contents included.
The surface is decaying and the drive is quietly shifting what it can onto spare tracks. Leave it powered and it carries on decaying. So it comes off power here, gets imaged once with the difficult regions deferred to the end, and every later step is performed on the copy.
The heads have reached a band they cannot read and the drive is retrying, at several seconds a sector, for as long as anybody allows. Ordinary file copying has no way to give up gracefully. Imaging hardware does, and that single difference accounts for most of what a lab recovers.
Platters turn at 5,400 or 7,200 revolutions a minute with the heads flying a few nanometres clear. A shove is plenty for contact. Switching it back on afterwards to gauge the damage is how a head replacement turns into a surface problem.
Mains spikes, ageing consumer units and the cheapest sort of desktop rebuild finish off drive boards across Berkshire most weeks. Boards are repairable, but the serial flash holding that drive's own calibration has to be carried over to whatever replaces it. Electronic work is quoted at 50% upfront and is one of the stated exclusions from no fix, no fee.
Boards carry a small suppressor across the supply rails whose entire purpose is to short and take the hit when the wrong voltage arrives. It performs exactly as designed, the drive then looks completely dead, and the mechanism behind it has not been harmed at all. Ten minutes at a rework bench, in the right hands.
Each board carries adaptive figures measured against the particular head stack and preamplifier it left the factory with. Transfer the board without transferring those figures and you get a drive that spins beautifully and returns noise. The chip travels with the drive or the job goes nowhere.
A disk stored since a house move in Warfield, or an office clear-out in Reading, comes back with the heads welded down by old lubricant and surface tension. The motor cannot break that bond and must not be asked to try. The stack is lifted clear by hand, in filtered air.
Notebook drives park their heads on a plastic ramp beside the platters. Interrupt the parking sequence with a flat battery or a snatched cable and the stack settles on the surface instead. Removing it without dragging it is clean-bench work, and no amount of shaking substitutes.
One preamplifier rides on the actuator and serves the whole stack. When it fails, all the heads fall silent together and the drive behaves as though it has none, which electrically is now true. Only a matched donor stack, preamplifier included, answers that.
Sealed high-capacity drives fly their heads in helium because it is thinner than air. A slow leak changes the flying height, error rates climb, and the drive gets measurably worse at reading itself every month it stays in service. Image it early, and let nobody open it outside a lab.
Fluid dynamic bearings give notice before they go. A hum under load first, then a wobble the servo starts fighting, then a seizure. The window for a clean image sits between the hum and the wobble, and it shuts sooner than most owners expect.
Failing sectors are retired into spares, and the list that records those retirements is a fixed size. Once it is full the drive hangs during start-up, disappears from the bus, or reports itself permanently busy. It looks terminal. The platters are frequently in decent order.
The partition structure has gone and the file data has not. Accept that offer and an empty structure is written over the top of the old one, converting half an hour of work into several days of it. Cancel the box, shut the machine down, and take the drive out.
A quick format rewrites the index and leaves the contents where they lie. A full format on any recent version of Windows writes zeros across the whole disk, and zeros are final. Which of the two ran, and how much has been saved since, decides how much comes back.
Deleting marks the space as available rather than clearing it. The harm is done by everything written afterwards: updates, downloads, browser caches, a backup job that runs overnight. Stopping is worth more than any software. The sooner the drive is idle, the more of it comes home.
Aiming a repair tool at a drive full of bad sectors is a slow method of making matters worse. It rebuilds directory structures out of data it cannot read dependably, on a mechanism already in trouble. Repairs are safe on an image and close to irreversible on the original.
Laptop firmware menus will lock a drive at the interface, and plenty of people set one in 2013 and thought no more about it. The mechanism is sound and the data is untouched, but nothing reads it until the lock is cleared in drive firmware. That is bench work, never software.
Worth being blunt about. Where the recovery key exists, in a Microsoft account, on a printed sheet, in Active Directory, or saved onto a second drive, a BitLocker disk recovers exactly like any other and the key is applied to the finished image. Where the key is genuinely gone, nobody decrypts it, us included. Track down the key first and worry about the hardware second.
ST4000DM004 and its relatives write in overlapping tracks and shuffle data out of a fast cache in the background. Interrupt a long write and the drive can stop replying for minutes at a time. Most are queueing rather than dying, and they image cleanly at a pace shingled recording will tolerate.
That Seagate generation had a firmware fault which parks a healthy drive in a busy state, or has it insist it holds nothing, and no operating system gets a word out of it afterwards. It is repaired through the diagnostic port with the lid still screwed down. Firmware repair of that kind runs under no fix, no fee.
The WD20EARS generation parked its heads every few seconds to save a watt, and in a machine that never left it alone the load and unload count climbed past a million inside two or three years. Ramp wear and head wear follow in that order. They arrive slow, then noisy, then silent.
MQ04ABF100 and WD5000LPCX units live inside machines that get carried about, and the head stack absorbs whatever the chassis does not. The running order never varies: donor heads matched by family and factory code, filtered air, then an unhurried image.
Ultrastar and Exos units come out of the small servers sitting under desks on the business parks along the A329(M), most of them running continuously since the day they were installed. Better engineering postpones the ending rather than cancelling it, and when it arrives it tends to arrive quickly.
The lid came off on a kitchen worktop with a Torx set from the shed. It happens more often than you would think, and it is not automatically fatal. Dust that has settled is a problem; dust dragged under a moving head is a much larger one. Send it exactly as it is, lid loose or not, and let the diagnostic report what the real damage amounts to.
Glass platters turn in there at several thousand revolutions a minute, and the read heads ride a cushion of moving air a few nanometres clear of the coating. Robust enough while they are flying. Catastrophically delicate the instant one settles down. That steady tick callers describe is the actuator parking and sweeping out again because the servo bursts it steers by have stopped resolving, and every sweep drags a damaged head back across the ground your photographs occupy. Which makes the least technical line on this page the one worth acting on: cut the power and leave it off. Anyone who arrived here at one in the morning after searching HDD Bracknell on a phone can stop reading and pick it up after breakfast. Bench work runs at the opposite pace. Filtered air, a matched donor head stack fitted by hand, a seized spindle freed where the bearing has gone, then a slow read that banks the cooperative regions first and returns to the awkward ones last. None of that alters the figure. Whatever the diagnosis comes back saying, one hard drive is £300 + VAT, and where the failure is mechanical or electronic, half is payable before work starts, since donor parts get purchased for your specific drive and not for drives at large.
Seagate recovery splits three ways on this bench, and the split is worth understanding before you post anything. Group one is the printed circuit board. A spike travelling down a tired supply kills the controller and leaves the recording surfaces entirely untouched, and putting that right means a donor PCB with the original ROM chip lifted across by hand. It does not mean a board ordered from a marketplace and bolted on hopefully, which is how a depressing number of drives arrive. Group two is firmware, the quiet one. Translator tables and defect lists stop agreeing with each other, and the drive sits there spinning contentedly while telling the host nothing at all. Group three is heads and motors, which is the noisy one everybody recognises. A good share of the hard drive recovery Seagate owners actually need falls into the first two groups, where no platter is ever touched. Shingled models want patience instead of speed, since they lay their tracks in overlapping bands and image slowly by design: Seagate's ST2000DM008 and ST2000LM007, the WD Blue WD10SPZX, Toshiba's MQ04ABF100. None of this trade is local, either. A boxed drive posts perfectly well, so Seagate data recovery UK enquiries land here from postcodes nobody in this building has ever driven through, and a parcel handed in anywhere in the country reaches the bench the next working day.
Badges matter here for one reason only, which is that spare parts follow them. IBM sold its disk business to Hitachi in 2003, the operation traded for years as HGST, and Western Digital bought the whole thing in 2012, so a drive wearing a Hitachi label today is really a Western Digital product under a name that outlasted the company. Two families cover nearly all the Hitachi hard drive recovery booked in. Deskstars, which are the 3.5-inch line, arrive out of towers assembled between roughly 2005 and 2013 and left in a loft ever since. Travelstars, the 2.5-inch line, come out of laptops of the same vintage, and there are three part numbers this office writes down again and again: HTS541010A9E680, HTS721010A9E630, HDS721010CLA332. Neither family goes out with a bang. One patch of one surface turns unreliable, reads across that patch slow to a crawl while the rest of the drive behaves impeccably, and the ticking only starts later, once the head covering that surface fails outright. The entire game is played in the gap between those two events. A Hitachi still reading in fits and starts has its best odds on the day the limp begins and considerably worse ones a month afterwards. Donor stock for both lines is racked in this building rather than ordered in when needed, because sourcing it gets harder every year, and anything that has lain still for ten years is inspected on the clean bench before it is offered power.
Spinning disks and flash storage break along quite different lines, which is why the same equipment is not used on both. A solid state drive has nothing rotating in it, no head assembly to be replaced, and no reason to be opened under filtered air at all. Its failures belong to the controller, or to the firmware table that keeps a record of which physical piece of memory each part of your data was actually written into. The routes in are the manufacturer's service mode or, where the controller will not answer, detaching the flash packages and rebuilding the translation with software. Where what has failed is an M.2 stick, an NVMe module or a 2.5-inch SSD instead of a mechanical disk, this site has a page given over to solid state work, naming the Samsung, Crucial, Kingston and WD models individually and describing how each one tends to give out. Cost is identical on either bench. One hard drive, or one SSD, comes to £300 + VAT with NVMe included, and on both sides the diagnostic is free and finished two working days after the parcel is signed for. Owners of flash storage get one piece of advice that disk owners never require: leave it unpowered. TRIM only performs its housekeeping while current is flowing, so each hour the machine spends switched on is another hour spent throwing away exactly what you are trying to recover.
Nothing on this list was bought to sit on a shelf. It is the equipment a failing drive actually passes through, roughly in the order it meets it, and the free diagnostic decides which parts of it a particular job needs. That diagnostic closes two working days after the drive is booked in at Guildford:
Nothing is opened anywhere else. A head flies closer to the platter than a particle of household dust is wide, which makes an opened drive on a desk a drive with a countdown attached to it.
The firmware half of the trade: service mode, translator rebuilds, defect list maintenance, locked modules read back out of the service area. None of that exists in software anybody can download, which is why so many drives get written off before a lab ever sees them.
Copies are taken one head at a time with the current draw watched from start to finish. Timeouts and retry behaviour are enforced in hardware, per sector, so an obstinate band costs a few minutes instead of the rest of the drive.
Tooling for lifting a head stack, or an entire platter pack, out of one chassis and into another. No recording surface is touched at any stage, and there is no version of this job in which touching one is acceptable.
Racked by model, firmware revision and factory code, because a head stack that is nearly right is no use to anyone. Most jobs find their match already on the shelf rather than waiting a week for one to arrive in the post.
Burnt components replaced, lifted tracks rebuilt, and the drive's own adaptive data moved across onto any board fitted in place of the original. Electronic work takes 50% upfront and is one of the stated exclusions from no fix, no fee.
Every port your media touches has a hardware blocker in front of it. From the moment it is booked in to the moment it goes home, nothing on this side of that blocker is capable of writing to your drive.
Read that as a record of arrivals rather than a catalogue. WD40EZAZ and ST4000DM004 disks head it because those two went into more Berkshire desktops and budget externals than anything else of their generation. IronWolf ST8000VN004 and WD60EFAX units arrive from four-bay boxes in home offices around Wokingham and Crowthorne. SkyHawk ST4000VX016 disks come out of recorders that have been writing without a pause since the installer packed up, and recorder work is one of the stated exclusions from no fix, no fee. Travelstar 7K1000s and elderly Deskstars surface during loft clearances and read perfectly well once their heads are freed. A single drive is £300 + VAT, the diagnostic ahead of it costs nothing and closes two working days after the drive is booked in, and a finished single-drive job runs 2 to 4 working days. Logical faults are no fix, no fee. Mechanical and electronic work is 50% upfront. PATA is still supported, should the drive in question have been in a garage since 2004.
The colour on the label tells you where a WD drive has spent its life. Blue in the family tower, Red in the NAS, Black in a workstation, Purple in a recorder, Gold in a rack. Blue desktop disks and 2.5-inch WD5000LPCX units make up the bulk of the WD workload here. Two patterns dominate what arrives. The first is a service area silting up with relocation records until a perfectly sound drive reads at walking pace; putting WD's own housekeeping routines through the diagnostic port clears the backlog, and the platters then image as though nothing had ever been the matter. The second is physical and is nearly always a drop, which means matched donor heads and filtered air. The older Green units earn a line to themselves, having parked their heads every few seconds for years and worn the ramps and the heads together. Whatever the fault turns out to be, the sequence is unchanged: image first, then work on the copy.
More Seagates reach the booking sheet than any other badge, and they fail in three recognisable ways. Shingled BarraCudas stop answering under sustained writing and give a convincing impression of death while remaining perfectly alive underneath. Firmware seizures leave a sound mechanism sitting at busy or claiming to hold nothing, and those are cleared through the diagnostic port without the lid coming off, which puts them squarely inside no fix, no fee. Then there is head failure, which sounds like clockwork and needs a matched donor stack on the clean bench. Around those three sit the specialist families: IronWolf disks pulled out of NAS enclosures, SkyHawk disks pulled out of CCTV recorders, and Exos units off small servers that have never once been switched off. The donor rack is stocked by firmware revision rather than by capacity, because on a Seagate the revision is what actually has to match.
There is a rhythm to which Hitachi arrives and what it wants. Travelstar 7K1000 drives out of mid-2000s laptops lead comfortably, with Deskstar 7K3000 desktop disks of a similar vintage behind them. Leave either one standing unused for a few years and stiction sets in, the heads bonding onto the surface with the motor no longer able to break them free. Both come apart on the clean bench without drama. Ultrastar units off small servers arrive with hour counts in six figures, and as long as the seal has held they image beautifully. Whichever turns up, the method is the same: one surface at a time, bank whatever the healthy heads can still reach, and only then ask a donor stack to cover the remainder. Spares for these families are held here rather than ordered when the job comes in, because they get harder to find with every year that passes. That shelf is the reason a Deskstar from 2004 is handled exactly as a disk pulled out of a machine last month.
Until the handover to Hitachi in 2003 the Deskstar and Travelstar names belonged to IBM, and one model from that period collected a nickname which then stuck unfairly to the whole family. Survivors from the IDE years keep coming out of retired towers and lock-ups, usually holding the last copy of a set of accounts nobody has thought about for a decade. Twenty years of standing still seizes spindles, stiffens lubricant and bonds the heads onto the platter. A disk of that age is inspected on the clean bench before it is given power rather than afterwards, because one hopeful power-up is how a stiction case becomes a scoring case. The legacy shelf carries period donors and PATA adapters permanently, for exactly this reason. No badge has yet proved old enough to condemn what was written underneath it.
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.
The drive has to come out of the machine before it travels, because a bare drive is all we take in and stripping computers is not work carried out here. On most desktops that is four screws and two plugs. Send it tracked and insured to Guildford Data Recovery, Building 2, Ground Floor, Guildford Business Park, Guildford, GU2 8XH. From Bracknell it is roughly 18 miles, about 40 minutes on the A322 and then the A3, so if you would sooner hand it over yourself the reception accepts drop-offs Monday to Friday, 9:00am to 5:30pm. A courier you book independently is equally welcome. Nobody in this network collects and there is no Bracknell office to call at, so getting the drive to Guildford is your half of the arrangement. Ring 0800 689 0668 inside those hours if you want the packing looked over first. One limit is worth stating up front: storage soldered onto a mainboard falls outside what we handle, as do phones and tablets.
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.
Missing a letter in Disk Management, or missing from the BIOS
Six noises, what each one means, and why the freezer never helps
Silent, invisible or stopping halfway: four faults sharing one word
A broken index and a dying surface want opposite treatment
Corrosion begins as it dries, so rice and hairdryers cost files
Ten towns across the county, one bench, and nothing collected
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.