SSD Repair: What Can Be Fixed, and What Can’t

People look for SSD repair when a drive has stopped being detected, has dropped to a fraction of its size, or has gone read-only overnight. Exactly what SSD file recovery costs is set out service level by service level. This page is about which job you need.

The distinction matters because of how these drives are built. A mechanical hard drive stores your files in one place and its own housekeeping information in another, so a mechanical fault and your data are largely separate problems. On a solid-state drive they are not. The controller keeps a constantly changing map of where every block of your data physically sits, and that map is itself written to the same memory. When the controller or the map is damaged, the drive can be electrically perfect and still be unable to tell anyone what it holds.

So “is it repairable” is usually the wrong question. The useful one is whether the flash memory still holds your files. If it does, there is a path to them. If the memory itself has degraded past the point of reading, no amount of board work brings back what is no longer there.

The drive starts working against you the moment it has power. Solid-state drives run housekeeping of their own in the background. Those routines permanently wipe any block the drive thinks is no longer needed. It is the same mechanism behind why deleted files vanish on these drives. A drive that has lost track of where your files are can believe almost all of them are no longer needed. You do not have to run a single program for this to happen. Leaving it plugged in is enough. This is the part almost nobody knows, and it is why “I left it connected overnight while I read up on it” is one of the saddest sentences we hear.

Every tool that says “repair” means “write”. Repair software, partition fixers, CHKDSK, disk utilities, firmware updaters. The friendly word on the button is not what the software does to the drive. It writes. Nobody has ever shipped a utility called Overwrite Everything, though one or two have earned the name. On a drive whose map is already damaged, writing risks more than the one file. It can overwrite the only remaining record of where everything is. One click, and a full case becomes an empty one.

You get no warning, because there is nothing to hear. A dying mechanical disk clicks, grinds and buzzes, and people react to that. A solid-state drive fails in complete silence, often from full working order to nothing between one boot and the next. There is no noise to tell you that the window is closing, so people take their time, try things, and only call once it has closed.

Read this part twice. A mistake here does not make the job more expensive. It makes it impossible. A case that would have taken three days and cost under $450 can become one that nobody can do. Not us, not anyone, at any price. That is the real thing you are weighing when you decide whether to have a go yourself. Not money. Whether the files still exist when you finish.

So the honest advice is uncomfortable and simple. If you are not certain, do nothing. Unplug it, leave it alone, and let someone look at it who can tell the difference before anything is written. Unplugging it writes nothing to it. Almost every other option on this page does, and writing is what takes the files away for good. The one exception is liquid damage, which keeps working while the drive sits, so those go straight in the post.

Most cases that arrive described as corruption are the first kind. Files open as garbage, filenames turn into symbols, folders show the wrong sizes, or the volume asks to be formatted. All of those are symptoms of a translation layer that has lost its place, not of files that have been destroyed. Once the map is rebuilt, the contents are usually intact and ordinary.

The second kind is harder to accept, and worth knowing early. On a solid-state drive, deleted and trimmed data really is gone, quickly and for good, in a way it never was on a mechanical disk. That is a design decision, not a fault. It is also the single biggest reason a drive that is still working can come back with less than its owner expected.

There is one message worth recognising on sight. When Windows or macOS offers to format a drive it cannot read, that offer is not a repair. It is the operating system giving up and proposing to start again on blank ground. Say no. The volume it could not read is usually still sitting there, and formatting is one of the few actions on this page that writes directly over the area holding the answer.

The rush fee buys position, not certainty. It speeds up when we start, and it does not guarantee data is recoverable. The evaluation always comes first, you always see what was found before you commit, and No Data Recovered, No Data Recovery Fee means exactly what it says.

Some of these are model-specific rather than general. Read-only lockouts, and health readings that collapse without real wear, are both documented behaviours on particular drives and particular firmware versions, which is why we keep a separate write-up of Samsung’s known firmware faults on the 870 EVO and 990 PRO. If your drive is one of those, the fix can be far smaller than the symptom suggests.

The common thread is that all five involve writing to a drive that should only be read. Physical damage follows the same rule. If a drive has been dropped, crushed, burned or soaked, do not power it on to check. What a physically damaged drive needs is different work in a different order, and the first power-up after the damage is the moment most often responsible for the final loss.

Those checks are worth doing properly, because a surprising share of “dead” drives are not dead. A portable drive is two devices in one shell: the solid-state drive itself, and a small bridge board that translates it to USB. The bridge is the cheaper, hotter, more fragile half, and when it fails the drive behind it is usually untouched. Taking the drive out and connecting it directly separates those two possibilities in about five minutes.

The same applies to interfaces. An M.2 drive can fail to appear because it is in a slot wired for a different protocol, because a BIOS setting changed, or because a slot shares its lanes with another device. Trying a second machine rules out an entire class of problem that has nothing to do with the drive at all.

What is not safe is everything that comes next in most guides. Formatting, initializing, partition-repair tools, CHKDSK, firmware updaters and secure-erase are all writes. On a drive whose map is already damaged, a write risks more than the file you are looking at. It can wipe the information needed to find any of them. That is the difference between a case that takes a few days and one that cannot be done at all. If you are weighing up whether your particular situation counts, the common questions cover it.

Physical damage on these drives is less about the memory and more about everything around it. The flash packages themselves are surprisingly durable. They have no moving parts, and they survive impacts that would destroy a mechanical disk. What breaks is the board they sit on: connectors sheared off, power components shorted, tracks cracked, water bridging contacts that should never touch. In a great many of those cases the data is entirely intact and just unreachable through the drive’s own connector.

Liquid is the one to move fastest on. Corrosion continues for as long as the residue is there, so a drive that is recoverable this week may not be next month. Do not dry it with heat, and do not put it in rice. Rice is for dinner. It has never recovered a byte. Do not and do not power it to see whether it still works. Send it while the damage is still only what the accident caused.

Work at this level is done to open a window, not to restore a product. The window may be one power-on long. Plan around that. Copy everything off, check the copy opens, and only then think about the hardware. That is what turns a fragile situation into a finished one.

It also changes what you should ask for when you get a quote. The useful question is not “can you fix it” but “what will I have when this is finished, and on what media”. A drive that has been to the lab and come home is on parole, not cleared. A returned drive that works today and stops next week has solved nothing. A verified copy of your files on new media has solved the actual problem.

It is also the right answer when the drive is under warranty and empty, or when the fault is outside the drive. A failed enclosure, a bad cable, a machine that will not power it. We handle every form factor and interface these drives ship in, from 2.5-inch and M.2 to soldered and portable designs. Every drive type we handle is listed in one place.

You send the drive, we evaluate it at no charge, and you are told what is on it and what it will cost before anything else happens. Nothing is attempted until you say so. You see the list of what was found before you pay. If it cannot be done, that fee promise applies. That order is deliberate. Evaluate, report, then decide. Nobody should pay to find out the answer is no.

Recover-SSD was founded in 2011 by Bruce Cullen, who has worked in data recovery since 1999. Drives that arrive untouched come back fuller than drives that have been worked on, and that gap only widens the longer a failing drive stays powered. If your drive is showing any of the symptoms above, the useful next step is to stop using it and request a free evaluation. The evaluation costs nothing, and it tells you which of the two jobs on this page you are actually facing.