Camera cards · bench notes · NTG-2025-2061
The Break Stopped Short of the Memory.
A year of watching forecasts, repaid inside one week: two clear nights, no moon, and stars right down to the horizon
. None of it had been copied anywhere yet. His reader then took hold of the card and declined to give it up. He is an astrophotographer, in Worksop, and a screwdriver went into the slot, at which point the card came off worse: part of the casing snapped away next to the gold contacts
. The piece disappeared beneath a desk and stayed there. Since then, the readers all behave as if the slot were empty
.
Yours doing the same? Say what it is up to — looking is free.
0115 8220606
What went wrong, and why.
Very little of an SD card is card at all. What sits in your hand is mostly moulded plastic around one block, inside which controller and memory are sealed together in resin. Fine traces run from that block out to the gold pads, straight down the corner a screwdriver had been working on. Break that corner and you break wiring, never storage. It is a shop whose shutter has jammed: the stock is all inside, and the front will not open. The missing fragment cost us nothing, then. Magnification shows how far a break has run, and this one came nowhere near the die.
The bench work this one took.
How the work runs →| Kit on the bench | What it did on this job | Why we reach for it |
|---|---|---|
| Rusolut Visual NAND Reconstructor | Error correction applied, and the dumped pages put back into their real order | A raw NAND dump put back together: ECC, XOR mask, page order, then files |
| BGA rework station & IR preheat | Controlled heat, gently and briefly, only to take the coating off the pads | Hot enough to lift the chip and reball its pads, never hot enough to damage what is stored on it |
| PC-3000 Flash | Identified the monolith against our library and mapped its pinout, unpowered | Reads the flash at chip level, bypassing the controller, matched to a maker-ID library we keep current |
What happened on the bench.
The break explained itself
With no fragment left to examine, the card had to supply the answer, and under a stereo microscope that took about a minute. A section of the contact block had gone, the traces stopped where it used to be, and the split ran out a long way short of the die. Seen once in that state, a card stops looking like something finished and starts looking like an ordinary chip-level job.
The route the factory leaves open
A row of service pads runs underneath a monolith, and that row is how the part gets tested on the production line. The layout is published by nobody, and makers do not agree with one another on it, so probing went on until we had the map. Then a wire finer than a hair went onto a handful of pads that count, and back down that wire, along a route the break had come nowhere near, the memory replied.
Raw pages come first; pictures are a long way after that
Pages leave a chip in the order its controller chose to put them there, and in no other order. Error correction has to run before anything else can happen. Off comes the XOR mask next, the pages are returned to their real sequence, and a folder name only appears once a filesystem has been built across the lot. That gap is where the work is, and no reader on sale anywhere is going to close it for you.
What was returned.
Both nights survived intact. Every frame opened, every frame checked, calibration files as well. The card itself is in a bin somewhere, and nobody misses it.
Start here if the same thing is happening to yours.
More card jobs on the log.
Sounds like yours? Switch it off before anything else.
£250 + VAT covers a card or a USB stick; a single drive is £300 + VAT; RAID and NAS start from £500 + VAT. Nothing is charged for the first look, and a fixed figure reaches you in writing before any chargeable work begins. Switch the device off, leave it alone, and post it in. On most jobs, no data back means no fee.