I have this sneaking suspicion (and that is all it is at the moment), that at least one manufacturer,
<brand name here>, may be programming their products to 'fail' at some point. Sounds paranoid, I know. But how do you explain six, well-cared-for-and-gently-used wireless optical mice (no moving parts except for switches and centre-wheel) failing for no apparent reason after about the same period of time? I can't, so I conducted an experiment: I purchased two identical <brand name here> wireless mice but used only one of them. The other I used as a 'control'.
Installed fresh batteries from the same lot in both mice. Turned them on at the same time, but left the control mouse simply sitting there on my desk, unused. Didn't touch the switches nor the codewheel, didn't move it around until after the other mouse failed. The control lasted a little longer than the other, identical mouse. Only my <brand name here> mice 'fail' after a short time like this.
My suspicion is this: the internal processor (some custom ARM or PIC variant, prolly) contains a counter which updates an internal Flash register so that the counter's value is remembered should the mouse be turned off or lose battery power.
The counter value is compared against a second value which is pre-set at the factory. This second value is randomized a bit about some central value which corresponds the median 'lifetime' the manufacturer desires of their product, say, four months plus or minus as much as three weeks, or something like that.
When the counter value equals or exceeds the preset value, a bit in Flash memory is set which drives the processor into a permanent HALT state and the mouse no longer works, in spite of the fact that nothing is actually wrong with it.
The result? The user replaces the mouse with a new, prolly identical, prolly <brand name here> mouse, as <brand name here> knows, like any corporation on top of their game, that most consumers tend to stay with what they know even if it is crap.
<brand name here> moves product out the door to the applause of the shareholders. More toxic scrap for our landfills, as users typically don't remove the batteries from failed products like these.
More often than not the innards of such embedded processors ('processors' here to mean any sort of 'intelligent' or computing or logic circuit, however implemented) are not accessible from the Outside via the pins and so whatever is going on here cannot be determined simply by plugging the thing into a logic analyser. Some other way is needed to determine what is going on, one which non-destructively reveals what is happening on the chip itself, usually potted or embedded in protective epoxy. Were the chip exposed (and circuit-side up, not flip-chipped in the IBM sense), the chip can be examined via electron microscope while powered.
But is there a way, say, using X-rays or charged particles or whatever to probe the chip while it is operating and without removing the protective coating?
How to find out? If it can be determined that the internal processor remembers any kind of state, something fishy is going on. Computer mice don't need to remember a damn thing.
Your thoughts?
Programmed Failure?
Re: Programmed Failure?
