David said:
Yep, I'd go with that as well.
Only certain kinds of capacitors are prone to chemical (corrosion)
failure. Not all electronics need be that flaky. For example,
ceramic caps have no liquid in them, so no pH to worry about.
http://en.wikipedia.org/wiki/Capacitor_plague
In the example here, the casing type is meant to look like a
Polymer cap. This is to fool the observer into thinking the caps
are not the bad type. Yet, the pressure relief stamping in the top
says they're electrolytic. And if any leakage is noted, that
proves they have the bad electrolyte in them. A good electrolytic
can last for fifteen years before the rubber plug in the bottom
cracks and the electrolyte dries out. If the caps don't have the
pressure relief seams, they're more likely to be polymer type.
Or, they could be a ticking time bomb
The pressure relief
seam is for safety, when a liquid electrolyte is used instead
of a polymer. Electrolytics could explode from overpressure, if
it wasn't for that seam. So if a manufacturer wished to hide
the internal composition, the manufacturer could remove the
seam in the top. Depends on the ethics of the manufacturer in China
as to whether this would happen or not. If you've ever seen
an electrolytic explode (even with the seam in place!), you'll
get some idea what a mess it makes. You can make an electrolytic
explode, simply by reversing the polarity - a trick done by
students to one another in university electronics labs :-(
http://upload.wikimedia.org/wikiped...d.jpg/567px-Blown_capacitor_on_video_card.jpg
You could take the bottom cover off the keyboard and do a visual
inspection. My current keyboard makes that easy. I also have an
old Apple keyboard, with steel rivets used to hold portions of
it together, and you cannot do much with that. Your keyboard
could be somewhere in between those two extremes. At most,
there would be one electrolytic in it, to reduce battery
impedance. Or, there might be none at all, shooting this
"failed capacitor" theory down.
*******
It's possible the communications on the keyboard are bidirectional.
It transits and receives. On the transmit side, the electronics
are clever enough to power down after a character is sent. On
the receive side, if the keyboard is designed to receive things
such a "turn LED on" commands, the receiver could be burning power
if the environment around the keyboard is filled with 3GHz modulation.
To test for that, you'd need a "controlled environment" for the
keyboard and USB keyboard dongle on the other end. And that's just
a little too hard to do properly.
If you recently purchased a new wireless router or other wireless
hardware, a new long range Bluetooth device, try turning those
off and retesting keyboard battery life.
You could accelerate the testing process, if you could monitor
battery current flow in the keyboard. If you had that capability,
it would be easier to walk around the room, turn off Wifi stuff,
then see if the battery current flow in the keyboard dropped or not.
That's faster than spending days seeing if the battery drain has stopped.
A cheap home multimeter can measure current flow (uses a series connected
shunt), so that's one way to do some monitoring. The shunt can be
left in place and the meter turned off and the keyboard would still
work. You only need to turn on the multimeter when making a reading.
Paul
