Sep 172013

This has been making the rounds on the Internets in the past few days: a modular mobile phone concept, with swappable parts.

Except that (with apologies to its inventor and supporters) I don’t think it will ever work. And no, not because conspiring corporations will torpedo it. (For what it’s worth, I am a free agent: I am not on the payroll of any conspiring corporations.)

The first reason is mechanical. For the phone to be robust, the backboard would have to be really strong and bulky. The connectors would have to be rock solid. Yes, it can be done, but only by using expensive materials, and the backboard itself will be half as thick already as a modern phone like a Samsung Galaxy.

The second reason is power and signaling. The placement of components on a modern phone mainboard is not accidental. Signal paths matter when things run off a multigigahertz clock. Power matters when some components can momentarily draw significant current. The placement of antennas matters, to maximize efficiency and minimize interference from the phone’s own components.

Third, the design will inevitably prove too constraining. Take modern PCs as an analogy. Yes, they are modular (it is much easier, of course, to make a desktop PC modular.) But only to a point. Try shoving an old ISA extension card into a modern PC. Even if it were perfectly functional (e.g., an old modem, serial/parallel or low-speed communication card that never needed more than ISA speeds) you can’t use it anymore, as no modern motherboard supports ISA slots. Many modern motherboards don’t even support PCI slots. Processor sockets change. Memory module standards change. Even power supply standards changed a surprising number of times. (You’d think there are only so many ways to supply 12VDC, 5VDC, and maybe 3.3VDC, but you’d be wrong.)

Still, Phonebloks is a neat idea. In fact, it’s one of those ideas that may never work as intended, but may still inspire other useful inventions.

 Posted by at 7:27 pm