Advanced intervention on SSDs with electronic failure, firmware corruption or no access. Recovery via direct NAND read and data reconstruction.
An NVMe SSD is not just memory. It depends on a controller, internal firmware and a complex logical structure that manages the data translation (FTL).
When these elements fail, data access is lost even if the NAND memory remains intact.
When hardware-level work is needed
Standard tools (Recuva, R-Studio, PhotoRec) work when the SSD is alive and only software is missing. When it isn't detected, reports zero capacity or shows electrical faults, you need a lab.
You connect the SSD and it doesn't appear in BIOS, doesn't appear in Disk Management, or shows 0 GB capacity. The usual cause is a burned controller or corrupt firmware. The NAND remains intact. We work board-level to restore communication or, if not viable, do chip-off to read the NAND directly.
Voltage spikes from a storm, faulty PSU or spills. The usual damage is to voltage regulators on the SSD's PCB. Microsoldering on affected components, line reconstruction and controlled boot.
The SSD enters the manufacturer's protection mode: the system sees it but with 8 MB or 0 GB capacity and it can't be read. Typical cases in some Crucial, Samsung and SanDisk series with buggy firmware. Resuscitation via PC-3000 or equivalent with controller-specific profile.
The SSD works but reads take minutes per file. Usually indicates NAND blocks degraded to the limit and a controller working overtime. Sector-by-sector imaging with aggressive timeouts before the SSD dies entirely.
Why an SSD isn't a normal drive
An SSD is nothing like a hard drive. No platters, no heads: there is a board with NAND memory chips where data is stored, a controller that decides which physical cell each bit goes into, and a table called the FTL (Flash Translation Layer) that maps the logical addresses Windows or macOS sees to actual physical cells. Without that table, the data is there but unreadable, like a book with the index and bookmarks ripped out.
When an SSD fails, it's usually not the NAND that's damaged (memory chips are surprisingly robust), but the controller that has suffered electrical failure or a firmware bug. With no working controller, BIOS detects nothing and the FTL is inaccessible. Recovery requires either restoring controller communication with tools like PC-3000, or desoldering the NAND chips, reading them directly with a hardware programmer (chip-off) and rebuilding the FTL manually using knowledge of the manufacturer's algorithm.
On top of this is the complication of TRIM: when you delete a file on a working SSD, the OS tells the controller "this cell is no longer in use". The controller, during idle moments, physically wipes those cells so it doesn't lose write speed later. Result: on a powered SSD, deleted data physically disappears within hours. The golden rule for SSD data loss: power off the device and disconnect the disk. Every minute powered reduces the chances.
Before sending the SSD
On SSDs the clock starts from the moment you notice the loss. These are the moves that destroy recoverable information.
TRIM and garbage collection keep running. Every minute powered, the controller can wipe blocks that contained your data. The rule: power off the device and physically disconnect the SSD as soon as possible.
When Windows sees an SSD with a broken partition table it asks to initialise. If you say YES, it writes a new table and recovering the data becomes much more complex. Always say NO and send it in.
chkdsk, sfc, Recuva-style tools or marketplace software can write sectors on the SSD. On an SSD with problems, every write can destroy entire blocks that were going to be recovered.
A clean Windows or macOS install writes gigabytes onto the SSD and triggers TRIM massively. Previous data becomes unrecoverable. It's neither 'fast' nor 'cheap' as an option.
Low-quality SATA-USB or NVMe-USB adapters provide unstable voltages. On an already-damaged SSD they can finish off the controller. If you must test the SSD in another machine, use brand-name adapters and a stable connection.
Restarting twenty times, trying different SATA ports, or warming the device doesn't bring a burned controller back. The more you try, the more you stress the remaining electronics. After 2-3 attempts, send it in.
Compatibility
We work with SSDs of any brand and format, from consumer NVMe to enterprise U.2. Every controller has its own algorithm and we keep specific profiles for the most common models.
Free initial assessment. Recovery is charged only if successful; the diagnostic fee is €45 + VAT and administration costs if you decline the quote.
