A drive the computer has stopped listing is usually still holding every file that was ever written to it. That is the normal result of this work, not the lucky one. Hard drive recovery Liverpool sends us arrives by post from Bootle and Crosby, from the estates out past Speke and the airport, from the Wirral side of the water and from Ormskirk, St Helens and Widnes. Nobody is asked for money until the fault has a name and one fixed figure is sitting in writing in front of them.
Every hard drive job is diagnosed free. One fixed figure follows in writing, agreed before a screwdriver comes out of the drawer.
No fix, no fee all jobs except electronic and mechanical failures, chip level work, DVR and Forensic jobs. Every band is listed on the data recovery cost page.
Matching the symptom to the fault underneath is the first job on any hard drive, and after twenty-odd years these thirty account for very nearly everything that arrives.
Each tick is the head stack reaching for servo marks it can no longer read, giving up, and starting the sweep over. Head failure announcing itself. The recording surface pays for every one of those attempts.
Three separate faults hide behind that silence. A spindle that has seized, a head stack sitting where it should never sit, or electronics that have died. Telling them apart safely needs a bench, not one more power-up.
It draws power and never reports a model name. That points at the firmware zone or the controller board rather than at the platters, and a service terminal is exactly what firmware zones are for. A downloaded utility on a USB lead will not touch it.
Nothing inside a hard drive is built to beep. What you can hear is the spindle motor fighting heads that have bonded to the surface, or a bearing seized solid. Each attempt tightens the knot.
Surface decay. The file browser freezes, copies crawl, and the count grows for as long as the drive has power. So it does not get power here. It gets imaged once, and the recovery runs on the image.
When the internal address map falls over, the drive either parks at busy or declares itself to hold zero bytes. Seagate earned the headlines for it. Every manufacturer has the same problem. A service terminal is how the map gets rebuilt, and firmware work of that kind runs under no fix, no fee.
At several thousand revolutions a minute, heads and platters settle the argument in well under a second. Powering it up again does not appeal the verdict, it only widens the damage.
Ageing wiring, tired consumer units and bargain power supplies retire drive boards across Merseyside week in, week out. Board repair carries the original ROM across onto the replacement. Like all board-level work it takes 50% upfront rather than running under no fix, no fee.
A quick format barely disturbs the contents. A full format is a different matter entirely. The outcome turns on how much has been written to the disk since it happened, which makes stopping the single most useful thing an owner can do.
The index has gone and the files have not. Partition structures come back readily enough. Say yes to that dialogue and you turn a straightforward job into a bad one, so decline it and pull the lead.
Servo patterns go on at the factory and no lab can write them again. Where a band of them has worn, the heads lose position mid-track and the drive resets, yet it reads cleanly on both sides of the fault. Those bands get mapped and the imaging steps around them.
Calibration figures matched to that drive's own heads and preamp live in a small chip on the board. Move a board across and leave the chip behind, and you get a drive that spins, sounds healthy and returns nothing but noise.
In a 2.5-inch drive the heads park on a plastic ramp beside the platters. Catch the parking sequence with a power cut, or wear the ramp out, and the stack ends up on the surface and stuck there. Lid off, filtered air, slow hands.
Laptop BIOS menus and disk utilities can set one, and plenty of people did. Mechanically the drive is fine. Nothing reads it until the lock is dealt with in firmware, which happens on a bench and never in software.
Sectors that start to fail get quietly moved to spare tracks, and the table holding that record is a fixed size. Fill it and the drive hangs, falls off the bus, or never finishes starting. Behaviour that alarming usually hides data in decent condition.
A drive this slow is failing gradually rather than all at once, and gradual gives us something to work with. Image it today, leave the awkward zones until last, and stop asking it to finish a copy it cannot finish.
The ST3000DM001 built the reputation of head-stack failure and still lands here most months, generally pulled out of a machine that has had no backup since the day it was bought. Its donor shelf has never been allowed to empty.
Feed sustained writes to a shingled ST2000DM008 or ST1000DM010 and it stops responding convincingly enough to fool anyone. Most of the time it is queueing rather than dying. Read it at a pace shingled recording tolerates and the image comes out.
The drive went into a box during a house move and is wanted this week. The heads have stuck fast where they were left sitting. Parting them takes patience and filtered air. Nobody here hits anything, whatever a forum told you.
The WD10SPZX and the MQ01ABD100 spend their lives inside machines that get dropped, and the head stack absorbs it. The order is fixed: matched donor heads, filtered air, then an unhurried clone.
Relocation entries accumulate inside Western Digital's own tables until a perfectly sound WD10EZEX reads like cold treacle. Clear them through the diagnostic port and the platters turn out to have been healthy throughout.
The HDWD110 goes into self-built desktops by the thousand, so head wear on it is a weekly booking rather than a curiosity. Its place on the donor shelf does not stay empty for long.
One small amplifier rides on the arm and handles every head in the stack. Age or a static hit finishes it, all the heads fall silent together, and the drive looks completely dead. Nothing short of a full donor stack answers that.
Every time a clicking drive is switched on to see whether it has improved, the heads score arcs into the coating. Data either side of the arcs still images. What sat under them has gone, and no laboratory anywhere gets it back.
Exos and Ultrastar units come off small servers with five-figure power-on hours behind them. Heavier engineering postpones the ending rather than cancelling it, and when it does arrive it arrives quickly.
Sealed units fly their heads in helium because it is thinner than air. Let some out and the fly height shifts, errors multiply, and the drive grows steadily worse at reading its own surfaces. Image it early, and let nobody open the lid outside a lab.
Fluid bearings resign slowly. A hum under load first, then wobble, then the servo losing its footing. While the wobble is still the smaller of the two troubles there is a window, and that window is when the image gets taken.
Bumpers and gaskets moulded in the early 2000s break down and leave a varnish across the heads and the landing zone. Drives out of lofts, garages and dockside lock-ups specialise in it. The filtered bench takes it in its stride.
A shingled drive moving data out of its media cache does not appreciate being interrupted. It can lose the thread entirely: stuck at busy, or writes acknowledged that never landed anywhere. Untangling that happens on a full image, never on the drive itself.
The lid comes off in a kitchen, a thumb lands on the top surface, and the drive reaches us in the post with an apologetic note attached. Frequently still recoverable. Every mark left behind narrows the odds, and the truthful answer waits for the diagnostic.
Three or four glass platters turn in there at speed. The read heads fly about five nanometres above them, held up by moving air. In normal use none of that is fragile. The moment a head touches down it is very fragile indeed. The tick people describe on the phone is the arm parking and swinging out again because the servo marks it needs have stopped being readable, and every one of those attempts drags a head back over the region your photographs live in. So the least technical instruction on this page is also the most useful one: switch it off and leave it off. Anyone who found this page by typing HDD Liverpool into a phone at midnight can stop reading there and pick it up in the morning. Bench work is deliberately unhurried. A matched donor head stack comes off the shelf and goes in under filtered air. A seized spindle is freed if the bearing has gone. Then the drive is read slowly, banking the healthy regions first and leaving the difficult ones until last. None of that changes the number. Any single hard drive is £300 + VAT whatever the diagnosis turns out to be, and where the failure is mechanical or electronic, half of that is taken before work starts, because donor parts have to be bought in for your exact drive rather than for drives in general.
A good share of dead drives have nothing at all wrong with the mechanism. A spike coming down an old supply will kill the controller board and leave the platters untouched. Fixing that properly means a donor PCB with your drive's own ROM lifted across by hand. It does not mean a board bought off a marketplace and bolted on in hope, which is where a lot of drives arrive from. Firmware is the other quiet killer. Translator tables and defect lists stop adding up, and the drive sits there spinning happily while telling the computer nothing. Shingled models want handling gently rather than briskly, because they lay their tracks in overlapping bands and copy slowly by design: the Seagate ST2000DM008 and ST2000LM007, the WD Blue WD10SPZX, the Toshiba MQ04ABF100. Most months, mechanical faults make up the bulk of the hard drive recovery Liverpool generates, with deleted partitions, accidental formats and reinstalls that went too far filling in the rest. A boxed drive travels perfectly well, so this doubles as an HDD recovery service for the whole country. Something packed properly and posted from any postcode lands on the bench the next working day, which is why the phone rings from counties nobody here has ever driven through.
The badge history matters when it comes to parts. Hitachi picked up IBM's disk division in 2003, traded it as HGST, then sold the whole operation on to Western Digital in 2012, so a drive carrying a Hitachi label now is really a Western Digital product under a name that stuck. Almost all the Hitachi hard drive recovery handled here involves one of two families. The 3.5-inch Deskstar comes out of towers assembled somewhere between 2005 and 2013 that have since spent a decade in a loft. The 2.5-inch Travelstar comes out of laptops from the same period, and three part numbers keep reappearing on the booking sheet: HTS541010A9E680, HTS721010A9E630 and HDS721010CLA332. Neither family fails dramatically. What happens is that one region of one surface becomes unreliable, reads over that region crawl while everything else responds perfectly normally, and the ticking only arrives later, once the head covering that surface gives out completely. Everything useful happens in the gap between those two stages. A Hitachi still reading in fits and starts stands its best chance on the day the limp begins, and a distinctly worse one a month afterwards. Donor stock for both lines is kept racked in this building rather than ordered as needed, since sourcing it gets more awkward with every passing year. And anything that has sat motionless for ten years goes onto the clean bench to be inspected before it is offered any power at all.
The two kinds of storage break in completely different ways, so they are not worked on with the same tools. Nothing spins inside a solid state drive, there are no heads to replace, and no reason to open it under filtered air. What goes wrong is the controller, or the firmware table that keeps a record of where on the memory each piece of your data was actually put. Getting back in means either the manufacturer's service mode or, when that route stays shut, lifting the NAND packages off and rebuilding the translation in software. If what has stopped is a 2.5-inch SSD, an M.2 stick or an NVMe module rather than a mechanical disk, this site has a page devoted to SSD data recovery which goes through the Samsung, Crucial, Kingston and WD models by name along with the way each tends to fail. Neither bench charges more than the other. Any one hard drive or SSD is £300 + VAT, NVMe included, and the diagnostic is free on both sides, completed inside two working days of the parcel being signed for. Owners of solid state drives get one instruction that disk owners never need: keep it unpowered. TRIM does its housekeeping only while there is power going in, so every hour left switched on is an hour spent quietly discarding the very thing you want back.
What a lab can do is decided by what it owns, so here is the inventory. Every hard drive booked in meets some part of it, and the free diagnostic that settles which part reports back two working days after the drive lands:
The firmware bench. Service mode, translator rebuilds, defect table housekeeping, locked firmware zones read back out. No downloadable tool goes near any of that, which is why drives get written off long before they reach a lab.
Failing mechanisms get copied one head at a time with the current draw under watch the whole way. Sector-level timeouts and retries handled in hardware mean an obstinate region costs a few minutes rather than the remaining life of the drive.
No lid comes off outside filtered air. A speck of ordinary room dust stands taller than the gap a head flies in. There is no arguing with that number.
Combs, jigs and lifters to move a head stack, or a complete platter pack, from one chassis into the next. Nothing is ever allowed to touch a recording surface. The rule holds without exception.
Sorted on the shelf by model, firmware family and factory code, since a head stack that almost matches is no use to anybody. Most jobs find a match already racked here instead of waiting a week for one to arrive.
Board repair under the microscope, with the drive's own ROM carried onto whatever board replaces it. Skip that and the adaptive values go, and the data follows them. Board-level work takes 50% upfront rather than running under no fix, no fee.
A one-way valve in hardware, sitting between our equipment and your media. From booking in to going home, no equipment on this side of the blocker is able to write to your media.
Take that as a record of what comes in, not a product range. WD10EZEX and ST1000DM010 disks lead it because more Merseyside machines were built around those two than around anything else. WD40EFPX and IronWolf ST4000VN006 units come out of home and small-office NAS boxes from Bootle to Ormskirk. SkyHawk ST2000VX015s arrive from recorders that have been writing without a break since the installer left. The ST3000DM001 renews its reputation with a fresh example most months. Travelstar 5K1000s and elderly Deskstars turn up from loft and office clearances, and they read perfectly well. One drive is £300 + VAT, the diagnostic ahead of it is free, and the report lands two working days after the drive does. Logical faults run under no fix, no fee. Anything mechanical or electronic takes 50% upfront. IDE is still supported on the legacy shelf, should you be thinking about the drive in the garage.
Western Digital colour-codes our week. Blue in the family tower, Red Plus alongside the NAS, WD_BLACK in gaming machines, Gold in the rack. Nothing arrives here more often than a WD10EZEX, with the WD20EZBX just behind it and the WD10SPZX coming out of laptops in similar numbers. The slow WD failure is a service area gradually filling with relocation records. The drive still shows up. It has simply become glacial. Running the housekeeping routines inside WD's own diagnostic mode clears the backlog out, after which the platters image as though nothing had ever been wrong. The remainder of the WD workload is knocks and drops. Matched donor heads and a filtered bench deal with those in the ordinary way.
Seagate appears on the booking sheet more often than any other badge. Shingled BarraCudas, the ST2000DM008 and ST1000DM010 in particular, put on a convincing death scene under sustained writing while remaining, in most cases, perfectly alive. Firmware seizures in the same family park a sound mechanism at busy or leave it reporting nothing whatsoever, and those come back through the service terminal without the lid ever being lifted. Firmware repair of that kind runs under no fix, no fee. The ST3000DM001, whose appetite for its own heads is legendary, still occupies a dedicated donor shelf ten years after the last one was sold. Clicking after a knock completes the picture, and that means filtered air, matched donors and an unhurried image. Behind all of it sit IronWolf ST4000VN006 units lifted from NAS boxes, and SkyHawk ST2000VX015 units lifted from recorders.
There is a pattern to which Hitachi turns up and what it needs. Travelstar 5K1000 drives out of mid-2000s laptops lead by a distance, with 7K1000 Deskstars from desktops of the same era behind them. Leave either standing for a few years and stiction sets in: the heads bond down onto the surface and the motor has no chance of freeing them. Both come apart on the clean bench without any drama at all. Ultrastar units off small servers show up with six-figure hour counts on the clock, and provided the seal is intact they image beautifully. Whichever of them arrives, the method is one surface at a time, banking whatever the good heads can still reach, and only then asking a donor stack to cover the rest. Spares for these families are racked here rather than ordered when needed, because ordering them gets harder every year, and that stock is why a twenty-year-old Deskstar is treated no differently from a drive bought last spring.
The Deskstar and Travelstar names belonged to IBM until the 2003 handover, and one model of that period earned a nickname the industry has never quite dropped. Survivors from the IDE years keep coming out of retired towers, usually carrying the only remaining copy of somebody's old accounts. Twenty years of sitting still seizes spindles, stiffens the lubricant and bonds the heads down onto the platter. A drive of that vintage gets inspected on the clean bench first and given power afterwards, never the other way about. Period donors and PATA adapters keep a permanent place on the legacy shelf, and no old badge has yet condemned the data underneath it.
Get the drive out of the machine first, because the bare drive is all that should reach us. Not sure how it comes out? Ring 0800 689 0668 and someone will talk you through it, or any computer shop will do it for very little.
A drive still bolted into a PC, a laptop or a server has to come out before it is sent. Bare drives are all we take, and dismantling machines is not work carried out here. Four screws and a plug covers most of them. Send it tracked and insured to Manchester Data Recovery, Peter House, Oxford Street, Manchester M1 5AN, about 35 miles down the M62 and next working day from any Merseyside post office. Handing it in at the lab reception works too, as does a courier you book yourself. No part of this network operates a collection service. Getting the drive to Manchester is therefore down to you. Ring 0800 689 0668, Monday to Friday between 9:00am and 5:30pm, if you want the packing checked over first. There is one hard limit and it is stated plainly: storage soldered onto a logic board falls outside what we take on, and so do phones and tablets.
Almost every job on this bench came in by tracked, insured post. It is the calmest way to move a drive that is already struggling, and a parcel handed in anywhere on Merseyside usually reaches the bench the following working day.
Still screwed inside a laptop, tower, MacBook, iMac, server or a CCTV recorder? Take the drive out and post that on its own. We do not strip machines here, and any computer shop will do it in a few minutes. The one job nobody can take on is flash soldered straight to a logic board, as on Apple Silicon Macs and a handful of very slim laptops: if the storage will not unbolt, there is nothing to send.
↓ Print the shipping & booking-in form (PDF)
Address it to Manchester Data Recovery. It is roughly 35 miles from Liverpool along the M62 if you fancy the run, and next working day by tracked post if you do not. We ring or email the moment it is booked in, and the free diagnostic closes 2 working days after that.
Not sure what belongs in the parcel? Call 0800 689 0668 before you tape it shut, or step through the free online diagnostic first.
Diagnosis free, one figure written down, and £300 + VAT covers a single drive or SSD. Start online or ring.