Memory cards · bench record · MHD-2025-6911
The SD Card Lost Its Corner.
Twelve months of watching the weather forecast finally produced two clear nights under dark skies in Northumberland
, and when the card came home, every frame was on it. But nothing could make it come out of a tight reader in Slough. A screwdriver was inserted and the card split before the reader did, and the plastic broke away beside the gold contacts
. That piece went under the workbench and was never seen again. After that time no reader would even see a card in the slot
.
Much the same fault? Call us.
0800 6890668
In everyday terms.
Most of an SD card is just packaging. The active portion is a single moulded component that contains both the memory and the controller, and they are connected to the gold contact areas using traces that run very close to where he snapped the corner off. Usually, breaking that corner causes damage to your wiring as opposed to losing the stored data, just like how tearing off a door handle doesn't destroy anything in the room behind it. He took the loss of the small part badly, but it didn't affect the work at all. Also, a break leaves evidence that you can look at under magnification, and this fracture had already used up its energy long before it got near enough to the die to cause any damage.
What the bench used here.
How a job is handled →| Tool | Why we used it | What it brings |
|---|---|---|
| Rusolut Visual NAND Reconstructor | Put the raw dump into page order and corrected the ECC | Turns a raw NAND dump into readable files: ECC, XOR, page order, reassembly |
| IR preheater & BGA rework station | Controlled low heat to lift the coating from the test pads without harming the memory | Lifts the NAND off the board and reballs it, keeping the heat away from the stored data |
| PC-3000 Flash | Identified the monolith in the library and settled the pinout before powering it | Reads the NAND chips directly, matched against a maker-ID library that is kept up to date |
The lab work.
Read the break before doing anything
Since the broken piece was now missing, the rest of the card had to provide all of the evidence needed; thus, we looked at it under the stereo microscope, and it told us clearly what had happened: part of the contact block had been bitten out; the traces were severed wherever that block had been; and there was a crack that had stopped far short of where the die was located. If you view it this way, an ominous-looking card became just another common chip job.
Take the service pads as the way in
Underneath a monolith sits a series of service pads. These pads are intended for use by the factory during manufacturing and have no other purpose. No two manufacturers lay these pads out alike, and none of the layouts are published, therefore we created a map of these pads by probing. Tiny wires thinner than a hair were soldered onto those few pads that mattered, and the memory responded through a path that the fracture had never touched.
Make files out of what the chip returns
The photographs do not come out of the chip as photographs, but as pages and pages of data, in whatever order the card's controller used. First, error corrections occur. Next, the XOR (exclusive-or) mask has to be removed. Then you have to reconstruct the page sequence, and the folder and file names will only appear when a file system is rebuilt over these corrected pages. Closing this gap is engineering work done at the bench, and a card reader can't do this type of work.
How the job closed.
Each of the light frames captured during the two nights was opened and viewed, along with the calibration frames. The card itself is finished, but he doesn't need to worry about it any more.
More of the jobs we've closed.
Filed on the Memory cards shelf.
Does any of that match your case?
Until the diagnosis is received there isn't anything to decide: turn the unit off, get a quote, then send it in, and allow the findings to make the determination.