Under a desk in most houses sits a tower holding every photograph the family owns, and in a fair number of small firms the same box holds the books as well. When one stops, the assumption is that the disk went with it, and nine times in ten only the computer did. Getting from a dead PC in Syston, Coalville or Market Harborough to a padded envelope in the post is a side panel, four screws and two plugs. After that it stops being your problem.
Every desktop computer that arrives here is examined at no charge. The figure quoted afterwards is put in writing and settled before anybody reaches for a screwdriver: £300 + VAT on any one hard drive or SSD, an NVMe blade counted the same.
Logical recoveries carry no fix, no fee. The named exceptions to it are electronic and mechanical failures, chip-level work, DVR jobs and forensic jobs; physical work takes 50% up front. The five bands are set out in full on the data recovery cost page.
Tracing a symptom back to the failure underneath it is where the real work begins, and these thirty account for all but a handful of the boxes opened at the Cambridge bench. A fault missing from the list is not an unfamiliar one — describe it on the telephone and you will get a straight view of the odds before you spend anything on postage.
A tower that powers up and stops short is telling you the firmware found nothing it could boot from. That is a boot record, a partition entry or a disk that has stopped answering, and in nine of ten cases the files are all still present. The machine looking dead and the disk being dead are two different findings.
Sound carries badly inside a steel case, so this gets blamed on a fan for weeks before anybody puts a hand on the drive. It is the actuator resetting because the heads have stopped reading. Once it starts, further power-ups take more of the coating off with each pass.
The disk is not answering the bus. Common causes are a failed board component, a shorted protection diode, a seized spindle or a head stack resting on the platters. They need four different repairs and no amount of cable swapping distinguishes between them.
Mains spikes come down the power lead and out through whatever is connected. The drive board takes the hit, often at the protection diode, which shorts itself deliberately to save everything behind it. That is the best possible outcome from a surge and it is repaired at a rework bench in a short session.
Corrosion continues under the surface for weeks after everything looks dry. A drive that has been wet is cleaned properly before any current goes near it, because the first thing a shorted rail does is destroy components that were undamaged. Do not dry it out and try it.
The new installation writes a fresh file system across the front of the disk and the old one stops being addressable. Underneath, the previous files are largely intact, because an installer takes up a small fraction of a modern disk. The clock starts the moment the machine goes back into daily use.
Workstation mirrors fail quietly. A member drops out, nobody sees the warning, and the machine writes to the survivor for months before that one goes too. Now there are two versions with a long gap between them. Both get imaged, and the newest good copy of every file is assembled from the pair.
The boot sector is unreadable or the partition table has been damaged. Both occupy a tiny amount of space at the very front of the disk and neither says anything about the data behind them. Rebuilding the structures from an image is routine work.
The motor reaches speed, the controller cannot read a clean copy of its own firmware from the platters, and it stops to limit the damage. Mechanically the drive may be perfect. The route back in is a service terminal, and there is no software equivalent.
All-in-ones put the whole computer behind the screen and are considerably harder to open than a tower, but the drive inside is a standard 2.5 or 3.5-inch unit. If the panel defeats you, ring before you force anything, because a cracked screen adds nothing to a recovery.
A dying supply can deliver the wrong voltage on its way out rather than simply stopping, and everything downstream sees it. The drive board is usually the casualty. Replacing the supply and hoping is how a repairable board becomes an unrepairable one, so the drive comes out first.
Damp, cold, and a disk that has not turned since 2015. Lubricant stiffens, heads can bond to the platters, and one hopeful power-up turns stiction into scoring. Drives of that age are inspected on the clean bench before they are offered power, and the legacy shelf carries period donors and PATA adapters permanently.
The prompt means the file system header is unreadable, not that the disk is blank. The photographs are all still there. Accept and a fresh empty structure gets written over the top. Decline, shut the machine down, and take the disk out.
This is worth being blunt about because CHKDSK is the first thing Windows offers. On a physically failing disk it writes corrections onto weak sectors, drags the heads across damaged ground for hours, and frequently moves recoverable files into found.000 as numbered fragments with their names stripped off. If the disk is making a noise it should not, do not run it.
The partition is visible and the file system inside it is unrecognised. Behind that word sits a damaged boot sector, a corrupted master file table, or a disk returning bad reads at exactly the wrong offset. All three are worked from an image and none of them benefits from a repair utility.
Windows is proposing to write a new signature and an empty partition scheme. Saying yes adds a fresh problem on top of the existing one. The disk being visible is good news, because the electronics and the firmware area are answering. What is missing sits further in and gets rebuilt.
Flash stops without any warning noise. The module goes silent on the bus and the memory behind the controller is untouched. Where the module unclips it travels here on its own in a padded envelope, and the rest of the machine stays where it is.
Read retries and sector relocation on nearly every request. The disk is working hard to hand you correct data and just about managing. That makes it the best possible moment for a full image and the worst possible moment to carry on using the machine for ordinary work.
On a spinning disk, deletion clears directory entries and leaves the contents until something writes over them. On an SSD, TRIM starts genuinely erasing within minutes. Which of the two you have decides how urgent this is, and in the second case the only useful action is shutting the machine down at once.
Drives that have run continuously for years are often carrying wear that only becomes apparent when they are stopped and restarted. Moving a machine, cleaning it out, or a long power cut can all be the last event. That is not superstition, it is the restart exposing a disk that was already marginal.
Feature updates occasionally change drive letters, hide volumes, or fail to bring a secondary disk back. Nothing has been deleted. This gets established quickly and is one of the more cheerful conversations in a week, because it usually ends without a recovery being needed at all.
Files encrypted where they lie on a physically healthy disk. What can be done depends on the variant, on whether shadow copies survived, and on what has been written since. It is priced as ordinary media, 300 pounds + VAT for a single drive and from 500 pounds + VAT for an array, and it is not forensic work.
Installing a second operating system alongside an existing one is where partition tables get rewritten, and getting the order wrong removes the entry for a volume that is otherwise complete. The files are untouched and the table is rebuilt from the structures further in.
Sustained writes, high case temperatures and a disk mounted where it gets no airflow. Corruption that appears under load and clears after a reboot is the classic sequence, and it means the drive is on its way rather than that the game is at fault. Image it before it gets worse.
The reset removes the file system and, on most machines, writes enough over the front of the disk to make the old volume unaddressable. The data further in is generally intact. What decides the outcome is how much has been done on the machine since the reset finished.
Database files from accounting software are frequently held open and written to constantly, which makes them the first casualty of a bad shutdown. Older complete copies of the same file often survive in unallocated space alongside the damaged one, and both get looked for as a matter of course.
A head that works cold and stops working warm, or a board component drifting with temperature. Each cycle wears the pattern down further. Imaging in short cooled passes recovers far more than repeated attempts at getting one long copy to finish.
A damaged journal, a lost superblock, or a resize that was interrupted. Backup superblocks and the journal frequently allow the volume to be reconstructed from an image with everything in place. Linux volumes are handled here as ordinary work rather than as an exception.
A mainboard change, a firmware update or a reset can bring up the recovery screen on a machine whose owner never knowingly enabled encryption. The key is usually escrowed somewhere, most often against a Microsoft account. Encrypted drive work is 400 pounds + VAT rather than the standard single-drive figure.
An old tower under a desk holding twenty years of photographs, replaced by something newer, and nobody knows how to get the contents out. No fault, no drama. The disk comes out, it gets read, and the files go back on fresh media. It is the least expensive job here and the one most often put off.
A tower is the easiest machine in this trade to get a disk out of, and that is worth knowing because the disk is the only thing that needs to travel. Side panel off, four screws, two plugs, and the drive is in your hand. Nothing else in the machine is wanted at this end and posting the whole computer means paying to protect a steel box around a small object. The same is broadly true of all-in-ones and the small boxes that hang behind a monitor, though both are fiddlier and both are worth a phone call before force is applied to a plastic clip. If the machine held two disks, or a small SSD alongside a larger drive, send both and label them, because on a number of models the two were tied together as a cache and neither is interpretable on its own. Put a sheet of paper in the box with your name, a mobile number and a line about how the fault started. That sheet saves more time here than anything else in the parcel.
Three stories account for most of the towers on this bench. The first is a surge, which comes down the mains and out through whatever is connected, and which usually stops at the small protection diode on the drive's board. That component is designed to short itself out and take the hit, so a drive killed by a storm is frequently a ten-minute repair at a rework bench with the mechanism behind it entirely undamaged. The second is a repair shop reinstalling Windows to fix a machine that had slowed down, which writes a new file system across the front of the disk and leaves the previous one unaddressable while the old files sit intact further in. What decides that outcome is how long the new installation has been in daily use since. The third is a workstation mirror where one disk dropped out months ago, nobody noticed, and the survivor has now failed too. Both get imaged and the newest good copy of every file is assembled from the pair, which is a job that goes considerably better than most owners expect.
Windows offers three things at exactly the wrong moment and all three make matters worse. The first is an offer to format a disk that is full of finished work, which appears because the file system header is unreadable rather than because the disk is empty, and accepting it writes a fresh empty structure over the top of the old one. The second is CHKDSK, offered on the next restart, which is a repair tool for a healthy disk with an untidy file system and is actively destructive on a disk that is failing physically: it writes onto weak sectors, keeps the heads working across damaged ground for hours, and moves recoverable files into found.000 as numbered fragments with the names stripped off. The third is Disk Management offering to initialise a disk that has years of work on it, which writes a fresh signature and an empty partition scheme. Decline all three, shut the machine down, and take the disk out. Everything those prompts would have destroyed is recoverable while it is left alone.
What arrives from a tower is a bare disk, and from that point it is treated like any other piece of storage in the building. The benches below cover the range from a straightforward image through to a head stack that has to be replaced under filtered air.
Head replacements, platter transfers and freeing a stack that has settled on the surface, all done with the drive levelled and supported in filtered air. A single airborne particle is larger than the gap the heads fly in, which is why this bench exists rather than a clean desk.
Firmware modules read back out of the service area, translators rebuilt, defect lists maintained and drives brought back out of a hung boot. This half of the trade has no downloadable equivalent, and it is the reason so many desktop disks get declared dead prematurely.
Copies taken one head at a time with retry limits enforced in hardware and current draw watched from start to finish. A stubborn region costs a few minutes of the schedule instead of the rest of the disk.
Head stacks, boards and complete mechanisms racked by model, firmware revision and factory code. A stack that is nearly right is no use to anybody, so the matching matters more than the capacity does, and most jobs find their match on the shelf.
Burnt components lifted and replaced, damaged tracks rebuilt, and the drive's own adaptive data moved across onto any board fitted in place of the original. Electronic work takes 50% up front and is one of the published exclusions from no fix, no fee.
Workstation mirrors and small striped sets are put back together from read-only images of every member rather than from the original disks. Nothing is tested against your hardware, and a set that has one stale member is reconciled file by file rather than by trusting the newest disk.
A tower is the easiest machine in this trade to get a disk out of, and that is worth knowing because the disk is all that travels. Four screws and two plugs on most models, and the machine stays with you. A single hard drive or SSD is 300 pounds + VAT, the assessment costs nothing and closes two working days after the disk is booked in, and a finished single-drive job runs two to four working days. Two mirrored disks out of a workstation are still priced as single drives if only one has failed, but a striped set is an array and starts at 500 pounds + VAT. Logical faults are no fix, no fee. Mechanical and electronic work is 50% up front and sits outside the guarantee. PATA is still supported, which matters more often than you would think, because a good number of the machines that come out of Leicestershire lofts predate SATA entirely.
Dell OptiPlex and Lenovo ThinkCentre towers dominate the business side of this bench, largely because so many of them were bought in batches and are now failing in batches. They are fitted with whatever the supplier had in volume that quarter, which in practice means WD Blue, Seagate BarraCuda or a Toshiba P300. Home builds are less predictable and frequently more interesting, because the person who assembled them chose the disk deliberately and often chose well. Either way the badge on the case tells you nothing useful about the storage inside, which is why the diagnosis begins with the disk in hand rather than with the model number over the phone.
Not every desktop opens with a side panel. All-in-one machines put the computer behind the screen and want a careful hand and a plastic tool. The very small boxes that sit behind a monitor are usually built around a single M.2 module, which is the easiest thing on this page to remove and post once you have the lid off. Where a machine looks like it wants force, stop and ring, because a cracked panel or a snapped clip adds cost to your side of the job and nothing at all to the recovery. Ten minutes on the phone has saved a good many people an expensive afternoon.
Separate the storage from the machine before you make the parcel up, since a bare drive is all that ought to arrive at Cambridge. If the screws win, telephone 0800 689 0668 and somebody will talk you through it; otherwise any repair counter in Leicester will do it for very little.
Take the disk out and send that, not the computer. A bare drive is the only thing booked in here, machines are not stripped at this end, and posting a tower means paying to protect a heavy steel box around a small light object. On a standard desktop it is a side panel, four screws and two plugs, and there is no need to remove anything else. If it defeats you, ring 0800 689 0668 between 9:00am and 5:30pm and it can be talked through, or any repair counter in Leicester will do it in ten minutes. Send it tracked and insured to Cambridge Data Recovery, Compass House, Vision Park, Chivers Way, Cambridge CB24 9AD. That is about seventy miles, the M1 south to Junction 19 and then the A14 east, and the lab sits two minutes off the A14 at Junction 32 with parking outside the door. Reception takes drop-offs in person Monday to Friday, 9:00am to 5:30pm. There is no collection service anywhere in this network and no Leicester counter, so the journey is yours to arrange. Put your name and a mobile number in the box, and if the machine held two disks, label them.
The post office does most of the work of getting a job here. A drive that is already unwell travels better boxed and insured than rattling around a car for a day of errands, and something dropped into a Leicestershire postbox this afternoon is generally logged in at Cambridge tomorrow.
The general rule is the drive travels and the machine stays behind — out of the laptop, out of the tower, out of the iMac, out of the recorder under the counter. This bench does not dismantle equipment, and a repair shop will do it while you wait. Three things are the other way round, and getting them wrong costs you the recovery: an external drive stays sealed in its own case, a NAS comes as a complete unit, and a WD My Passport or My Book travels whole with its cable, because on those the encryption key is held on the bridge board rather than on the disk — separate the two and the data becomes unreadable even to us. A Fusion Mac needs both of its drives, each labelled. The one thing nobody can work round is flash soldered onto the mainboard, as on Apple Silicon machines: if it will not come off, there is nothing to post.
↓ Print the shipping & booking-in form (PDF)
Address it to Cambridge Data Recovery. It is about seventy miles from Leicester if you fancy driving it — M1 south to Junction 19, then the A14 east — and the lab is two minutes off Junction 32 with parking at the door. Posting costs you a stamp and a day instead. Whichever you choose, you hear from us the moment it is booked in, and the free diagnostic closes two working days after that.
Not certain what belongs in the box? Ring 0800 689 0668 before you tape it up, or let the free online diagnostic ask the questions for you.
The Windows.old folder, and the reinstall paths that leave nothing at all
Bin emptied, folder gone, share tidied: three starts, three sets of odds
What can be steadied long enough to read, and what cannot be
The repair prompt Windows offers, and the damage it does when accepted
The examination is free, one written figure follows it, and the band covering this page is £300 + VAT on any one hard drive or SSD, an NVMe blade counted the same.