A friend of mine is just getting started in footling around with hardware, and he's trying to upgrade the amount of memory in his (commercial) .mp3 player. Said player is flash-based, so he figures that it's a simple matter of soldering an identical memory chip atop the existing one, so that all signals are duplicated to and from both chips.

While I'm quietly scoffing to myself, my dad informs me that this sort of thing actually used to work, and that he has personally seen it in practice. This, of course, has blown my mind with befuddlement.

First question: can anyone here corroborate his claim? Everything I know about electronics says that it shouldn't work, but there it is. Further, my electronics skills are typically dismal, so I'm inclined to believe that there's a loophole somewhere that I'm not seeing.

Second question: what should my friend do? Assuming that just doubling the number of storage chips isn't going to work (nand Flash is my guess; I'll be finding out tomorrow afternoon), how can he upgrade the storage in his player? My best ideas, in order of preference, are 1)use a simple switch to power on each chip at a time; 2) just use tristate buffers to determine which chip is reading/writing; 3) use multiplexer logic to determine which chip is reading/writing; 4) add a selector latch to control which chip is active in conjunction with option (3), and hack the firmware to use it.

None of these are very good, I know, but I can't just shut him down immediately. 'Sides, it's an interesting problem. Second question: what are your opinions on the feasibility of this upgrade, and what routes would you reflexively investigate for this upgrade? (No, "just buy a better one" isn't on the list; thank you for playing.)

Thanks,
Bob Robertson and co.