The Issue

A little over a year back, I had over 200 pictures stored on Lexar Memory Stick Pro Duo 4GB (MagicGate) card. After a while the card stopped working. It gave an error C:13:01 on the Sony Cybershot I tried it on (Reinsert the Memory Stick). I even tried it on another camera and that also didn’t work. There are a lot of good pictures on that card so I still need to get them off. Most were from a Australia Vacation, with pictures of Koalas and Kangaroos and all sorts of neat pics, but I never got to see them! I tried several different data recovery efforts with no success. My thoughts now lead back to the card which may be damaged. Initially after doing a bit of research I found out that the file allocation table (FAT) can become corrupted if you switch cards between different cameras, which was exactly what happened before the problem came up. However, if this were the case, then the data would still be there and a data recovery software could read the raw data on the card and rebuild the JPEGs, however the card itself can never be read.

Disassembly

Of course when there is a problem with an electronic device it means its time to disassemble it and troubleshoot. The Memory Stick is a fairly small device. It doesn’t have any screws to separate the chassis. Instead it consists of a few plastic components wedged and fused together with a multi-layered PCB containing the components. The PCB’s components are all sealed up in what looks like a solid block of plastic on top. Doesn’t look very easy to troubleshoot compared to a much larger circuit. Its are approximately 20mm x 31mm x 1.6mm. Considering how small it was, I took high resolution photographs to see it up close. Even though I cleaned the contacts, after removing the plastic chassis, the contacts looked extremely dirty.

Even after cleaning the contacts at least one of them looked partially corroded. This may have happened from small amounts of salt water leaking into the camera from splashing on the outside. I figured if that edge connection was severely corroded then it must be disconnected from the trace that leads to the rest of the circuit.

So using the high res pics, I followed the trace up the board a little, scratched some of the enamel off using a razor to expose the conductor. Then I did a continuity test from the edge connector to the exposed trace, and that test passed without any problem.

I did several other continuity tests too based on information regarding the pinouts shown below, and all those tests passed. I tested for continuity between Pin1, Pin10, and Pin6, all of which should be connected to VSS. If it was a simple problem such as the continuity of a trace being broken, then a simple jumper from one end of the trace to the appropriate edge connector would solve it. And data would be able to be read from the memory, however none of that is working.

Memory Stick Pinouts

Memory Stick Terminal Functions

Test Points

One thing which got me looking closely is the 5 test points on the board. They are significantly larger in diameter than any width of any trace on the board, so they were meant to be worked with by humans. However what can be done with them? Nothing is labeled at all. The only obvious test point is the one on the top corner which is connected to the largest trace on the board, VSS.

Looking at the Block Diagram of the Memory Stick Pro from the Memory Stick Pro standard, you can see that its modularized into different components, its not just a flash memory chip on the board.

I’m thinking that test points would be useful on a device that is a prototype since when you’re in the process of electronically engineering the device and testing the prototype, you can run tests through the test points. However this device is in production, so of what use are test points? The answer must be data recovery. Its a possibility that the test points lead directly to the flash memory chip bypassing all the other modules allowing you to extract data in the case of a failure. This still doesn’t tell me the pinouts of the test points though.

I have one far-fetched idea that could work. When performing data recovery on a hard drive with a broken PCB, an identical PCB can be swapped out allowing access to the data stored on the disks. Likewise, if I am able to get an identical Lexar Memory Stick with working modules, I can bypass any broken modules on the current memory stick to get to the flash memory chip on the card which contains the actual pictures.

The actual far-fetched idea consists of connecting all test points and edge connector pins in parallel and plugging that into a memory card reader. In theory, I’m thinking this could work. If you think of something simpler such as a nand gate chip connected to an LED or two powered up by whatever DC power source. If you put another identical nand gate chip right on top so all pins are connected to all pins in parallel, the LED should still function with the same output as normal right? Likewise, even if the original nand gate was malfunctioning, the second one could probably handle the I/O by itself.

The only problem I could imagine happening with this far-fetched setup is if some kind of transistor or other component had somehow fused in a way that it became a jumper, or if somehow something got bridged where it shouldn’t have been bridged, this could inhibit the new modules being added to the circuit.

Still I’m confident that the test points are there for data recovery which is the ultimate goal here. If anyone has any information on how to get the information off the card, post a comment. I’m thinking of going to buy an equivalent Memory Stick to use in the data recovery experiment.

Tags: featured

• Digg
• Stumble it!
• Reddit

• Sony Cybershot W120 Teardown and LCD Replacement
• How to Hook up an N-Channel MOSFET
• LCD Screen No Power, No Signal