A solid-state drive (SSD) gives you no warning of failure. There are no sounds; there's no slow down over time; there's nothing that alerts you. Just fine at five o'clock and dead in your lap by nine the next morning. That's an entirely different trade from disk work. No platters, no heads. It runs every day for households, for both universities, and for survey and design firms where all their daily output lives on one laptop.
All jobs for SSDs are diagnosed free. The fixed figure will be provided by written communication prior to any tool usage.
In most cases you only pay if we recover your data. The exceptions are electronic/mechanical failures, chip level work, DVR and forensic jobs. All pricing information can be found on the data recovery cost page.
The initial diagnosis of the problem of the SSD is based on how well the symptom matches the fault. In twenty-odd years, these twenty-five cover almost all of the ones which reach this bench.
The management chip fails and the drive loses access to the bus. All files are retained within the NAND. The only item which has failed is the item which speaks for that memory, and a firmware bench takes over that job.
No drive is actually called that. This is simply a Phison controller stuck in its own boot loader. Your files exist behind a wrong name and remain undamaged; service mode allows us to walk your files out of there.
Failed translator. The drive provides a polite response, however, the drive indicates a size that belongs to no product. We rebuild it using service mode, take a complete image of the drive and only after this process can we request further activity from the drive.
Boot code failure, power rail failure, or a failed controller. Each of these produces the same blank-screen effect. They each require a bench to separate; it is the first answer provided by the free diagnostic.
Write protection has been engaged at end of life, to prevent loss of existing data. While frustrating if you want to use the drive, it is helpful when you have to recover data, since a frozen drive cannot harm itself during recovery.
Flash provides very limited warning. As such, so many calls begin with someone describing how the device functioned flawlessly the last time it was used. It is not carelessness; this lack of warning is due to how the technology fails.
When a write is interrupted, the tables which translate logical addresses to physical locations become jumbled. When there is a true surge, it also damages or destroys the power stages of the drive. Both of these occur frequently and both require firmware work to fix.
Worn-out NAND will sometimes return different data for each successive read, while SMART reports minor issues in the background. Because we cannot trust subsequent images, the initial full image is what is relied upon; therefore, it is taken once and taken correctly.
Any time the unit has been connected to power, the background erase process for freed blocks will run. Once the unit loses power, the erase process will stop at whatever point it was when the power went off. If you leave your drive connected to power while you decide what to do, the thinking may cost you files.
Thin boards have failed because of screwdrivers, attempted upgrades, and stiff envelopes. A "clean" break near the edge of the board is likely going to be a track repair job using a microscope. Anything more extensive will be performed by way of chip-off. Chip-off repairs are slower but get back the same files.
Samsung's MZ-76E500 is an incredibly popular drive in both UK desktops and laptops, so every version of this issue passes through here as well. This particular "signature" issue simply shows up with no warning, will leave the SATA bus during a read/write operation and when it comes back it reports that it can see virtually no space on the drive or sees nothing at all; and neither switching ports/leads nor using different ports makes any difference. The translator has left the controller. Normally, the V-NAND underneath is untouched. Reverting to service mode and creating a complete image is typically how this issue gets recovered. Owners get the best results if they pull the power cord immediately after it starts acting wonky.
MZ-77E1T0 units from the first production runs were found to rapidly reallocate sectors according to documentation. Due to this rapid sector reallocation, Samsung released a firmware update. There are still units arriving without the patch. Even though the memory reads fine, there is damage to the mapping table that needs to be rebuilt.
Most owners are worried when the pending sector count rises for Crucial's CT500MX500SSD1, a known habit, but there really isn't much to worry about until genuine read errors occur alongside it. Telling those two apart is bench work, not guesswork.
WD's SATA Blue exhibits a documented behaviour; this drive simply leaves the bus and doesn't come back until some firmware attention gets applied. This is the usual patient with a usual route. The NAND is very rarely responsible for this issue.
A400 SA400S37 units and their relatives reduce themselves down to single-digit megabyte amounts, or disappear completely, after the translator has corrupted. These are clearly service-mode issues. In general, the flash underneath these drives has been fine.
These units go dark during a write operation and refuse to continue to talk to you. Some of them can be revived using service mode. The others have to be read at chip level (which reaches the same files by a longer path).
Fitted right through the 2013 to 2017 machines, the blades of both the SSUBX and SSPOLARIS families will eventually fail, just like all other solid state devices, and these blades can be imaged on native adapters here. No donor Mac is used at any point in this process, and that usually surprises many of the customers we speak to.
The DRAM cache helps level the wear on each SSD block, so without it, the wear-levelling is very primitive; therefore, any BX500 SSD model put into heavy use will age prematurely. The packages come off the board, and then the data comes off those packages.
Marketplace counterfeits loop writes back into a tiny sliver of genuine storage space. The data that was written to the real amount of storage will remain. Any data that was written after that was never stored and cannot be retrieved by anyone.
In a routine swap, the plastic tongue breaks off and takes some pads with it. To fix it at this point, board-level repair puts everything back together again. After that, the drive acts as if nothing ever happened.
After years of intense use (writing), all of the rated life of the drive has been consumed. At that point, the drive either becomes locked or it begins reading data erratically. At most, there is enough margin remaining for a single image operation. It gets planned as such, not started in hope.
A vendor utility had to stop mid-stream and left the drive unable to determine which drive it was. Based upon what the controller can still be persuaded to tell us about itself, we'll reflash it here or bypass it completely.
Repeatedly reading one block affects the charge stored in its neighbouring blocks. In drives filled with archive files that continually get read but never rewritten, eventually the errors begin building up and become too large for the ECC (error correction code) circuitry to cover. Slowly reading the same areas multiple times results in recovering significantly more data than would occur from one rapid pass.
Cells lose their charge when powered down. A device last turned on in 2019 could possibly show up as "noise". Performing an image of the drive via repetitive reads with plenty of room to allow the ECC to perform its function restores a significant amount of information that a single rapid pass would deem unreadable.
Just about all modern SSDs encrypt data as it comes into the system, regardless of whether you have specifically asked for encryption to be used or not, and they do this by storing the encryption keys in the controller itself. Therefore, if the key storage area is damaged, the memory reads perfectly and is essentially meaningless. Instead of providing false hope we provide an honest description of what is going on.
An SSD doesn't store your files at a physical address the computer has. Wear levelling distributes them across the NAND packages where the controller likes, and that distribution is stored in just one location; the controller's firmware contains a mapping table. If you damage this table, or if you lose the controller, the device will stop responding on the bus but all of its bytes will still be fine. There are two ways to recover from this type of failure. The first way is through service mode; the controller is accessible through the manufacturer's own diagnostics interface. The controller can then recreate/reload the lost information. The second route is called chip-off. Remove the NAND from the board and read each individual chip directly and then create the mapping table within a piece of software. This method is slower than service mode and the end result isn't dependent upon the drive being able to communicate with the system. Consumer utilities manage neither, because all of them need a drive the computer can already see.
The 860 EVO receives a whole paragraph. This model is the most prevalent model to fail here and the failures have a recognisable appearance. A previously functioning drive will drop off the SATA bus mid-session. At next boot, the BIOS reports no drives, or reports the failed 250GB, 500GB, 1TB or 4TB drive as being only a couple of MB in size. That usually indicates firmware corruption. The controller boots, performs self-checks, fails them and thus never presents the user area. Corrupted firmware signatures in the service area (or a corrupt controller that makes the translator un-readable) look identical from the outside. Important to remember is that the V-NAND underneath is virtually always good. The memory is good. The device describing the memory is bad. Route in: service-mode access using the vendor-provided diagnostic tool; repair/reload damaged firmware modules; then a complete dump of the entire image, taken immediately (before attempting anything else). Writes occur in areas of the service area where the firmware resides, but not anywhere close to your data. When service mode is denied, the packages are removed and read directly. Practical note: vendor provided firmware updates are intended to be performed on a healthy drive. Attempting to do so when the drive is failing may result in another failure which may be your last.
Samsung, Crucial, Kingston, SanDisk, WD, Intel, Solidigm, Kioxia, SK hynix and Apple's removable blades come in the three typical sizes: 2.5-inch SATA, mSATA and M.2. The common ones are just what were commonly used in the trade for that particular year. The common Samsung are 850 EVO, 860 EVO, 870 EVO and the QVO variants. The common Crucial are MX500 and BX500. Common Kingstons are A400, KC600 and the old SSDNow V300. Common WDs are Blue SA510 and the Green that came before it. Then there are the SanDisks, Ultra 3D and SSD PLUS. After these are the SK hynix and Kioxia units pulled from prebuilt laptops. There are two types of devices that do not qualify for this service. Hearing about them early beats having to pay postage to find out. The first type is when your storage device is soldered onto the motherboard of your computer. If you're not sure which kind you have, please ring us first. All phones and all tablet computers fall into the second category; they are outside this service completely. Both aggressive TRIM and hardware encryption will also make a job harder than it would appear when you initially look at it. This is something we tell you during our initial assessment, not something we print on an invoice weeks later. SSD data restoration Scotland wide takes place on the exact same bench, and a bare drive in a padded envelope is next-working-day post from anywhere in the country.
Solid-state recovery is typically firmware and chip level repair work from day one. This lab is designed for both of these processes:
The bench wakes the silent drive in its maker's service mode, performs the necessary firmware repair, rebuilds the translator and makes a full image copy of the contents before anything further is asked of the drive.
When drives stall they receive hardware timeouts, with retry behaviour set based upon exactly what device is currently being tested. Default settings that were intended for different models are worse than useless.
Each removable solid-state media type has native sockets available, including Apple blades and short 2230 modules. Forced cooling for any device that begins to malfunction when warm is standard procedure for our bench. There are no USB caddies in the entire chain.
In cases where no service mode will activate the controller, individual packages are removed from the board and each package is read individually. While this process may take significantly longer, ultimately the same files will arrive at the end.
Working from the raw dumps, the interleave, XOR and ECC are solved until a valid mapping is present in software again. Once this occurs the volume will then be rebuilt on top of it.
First to fix are burnt power stages and cracked boards. A drive that browns out during an image transfer tells you only one thing: the board needed fixing.
There is one model number to this page, and that model number is Samsung 860 EVO. The MZ-76E500 was an upgrade that nearly every single user installed when they were replacing their hard drive(s) with an SSD, thus there will be an entire generation of these upgrades all failing around the same time. There is no need to interpret what is the most commonly known failure point of this model. Disappears off the bus without warning, comes back as a sliver of its reported capacity, or disappears completely. Failure of the translator, sitting above healthy V-NAND. This unit does recover. The 870 EVO does the same thing. Around them are the A400 SA400S37 losing its mapping, the MX500 worrying about pending sector issues, the SA510 disappearing silently, the BX500 failing early because of work it was never sold to do, the 870 QVO stalling after using up its fast cache, and the NV2 showing up with whatever controller and flash happened to be going that quarter. In very few words, if it un-plugs, we can work on it. Whether you purchased a 2.5-inch, mSATA or M.2 (read-only lockout also included). If it's soldered onto a motherboard, it is not taken on (if you're not sure which type you have, ring us first). And neither are phones and tablets.
Get a quote first by telephone or use our free online diagnostic. Most of what you see on this bench will then be delivered via tracked, insured post. This is one of the safest ways to move a failing drive and packages handed in at a North-East post office are normally delivered here the next working day.
Still attached to a laptop, desktop, MacBook, iMac, server or CCTV/DVR unit? Remove the hard drive or SSD first. The bare drive travels on its own; when removing disks from a server remove all the disks and label each disk by its bay. This lab does not take machines apart. Any storage that has been soldered onto a motherboard (as in T2 and Apple Silicon Macs and a couple of thin laptops) will not be worked on: if it will not unbolt, we cannot work on it. If you're not sure what type of machine you have, ring us first. Phones and tablets are also not taken on.
↓ Print the shipping & booking-in form (PDF)
Your parcel should be addressed to Edinburgh Data Recovery, approximately two and a half hours (and 130 miles) from Aberdeen via the A90; alternatively it can reach us the next working day by tracked post if this journey time / distance is prohibitive to travel. As soon as your drive has been scheduled in we will contact you.
Still unsure about what should be placed inside of the packaging? Please call us first on 0800 689 0668, or use our free online diagnostic tool to assist you in deciding which items belong inside of the packaging before sealing.
Free diagnosis, one fixed price quoted in writing, and in most cases you only pay if we recover your data. Start online or phone it in.