Slashdot Mirror
Memory Card Torture Tests
saikatguha266 writes "BBC is reporting that five types of memory cards were dipped into cola, put through a washing machine, dunked in coffee, trampled by a skateboard, run over by a child's toy car, given to a six-year-old boy to destroy, smashed by a sledgehammer and nailed to a tree. It was still possible to retrieve photos from the xD and Smartmedia cards while the others didn't survive just the last two tests. "
To avoid damage to your retinas, click here
Although this only really applies to the nailed-to-a-tree test, where the nail goes through will matter. In DIL ICs, most of the area is taken up by connections to the pins. If these cards have the core close to the edge with the connectors, and a nail is put through the center, it could miss the core entirely. And if the nail went through the bridges, a data recovery person could wire a reader the the connectors inside the package.
The Tunsten has something of a reptutation for destroying memory cards. Check out this list of problems encountered using different cards in different machines.
Note that in the Tungsten T3 only 1 out of 7 tested cards actually survived the ordeal! Of course, this particular test isn't scientific, but it is based on real user feedback really using them for real things, so it's probably worth a read.
no idea, perhaps the floating gate of the flash cell is affected (shifting bandgap,ect), and a read will accidently write ?
ok, you would need a few tesla, but the point is that if you have moving electrons, you cant just say magnetic field cant do anything...
HI O WISE PRINCE. WHT TOOK U SO DAM LONG?
It may have been the USB controller that was damaged. The way that these drives were designed was that the connector was attached directly to the circuit board, without anything else holding it in place. Or, perhaps it may have been the flash memory that was faulty, like in your circumstance.
When you move though a magnetic field, that induces a voltage which could mess up the memory card. That is, don't bring your camera to a MRI machine (Nuclear Magnetic Resonance describes it better but people freaked out about the word "Nuclear"!)
It is not the magnetic field that does the damage. It is always the rate of change of the magnetic field that is the problem (its gradient). This is how power plants make the electrons flow from the wall socket :)
So... as you can see, with a strong enough electric field, you may surely be able to move the charge across the barriers (e.g. break-through of the insulation or maybe tunneling of the electrons when you skew the potentials alot).
Now the magnetic thing:
Because the electrons in the floating gate move (they must do that in every case because of quantum uncertainty, but the greater effect here is the temperature movement!), they have a speed and as you may know, moving electric charges in magnetic fields feel the lorentz force... BUT this force does not change the energy of the particles, i.e. they do not get faster when one applies a magnetic field. So, yes, maybe there're weird changes in the bandgap. I don't know. Try it out, would be a nice experiment.
Post a link to a graph here that shows the total number bit errors over magnetic field strength
Static RAM mainly consists of two MOS inverter structures wired together on the chip to form a flip-flop. Static RAM needs a small bit of current (because of inevitable leakage currents) to keep it's state.
At least, that's how I read it some time ago for a seminar in the semiconductor book from the creator of these devices (S.M. Sze).