|> On Fri, 04 Apr 2008 06:38:55 GMT
[email protected] wrote:
|>
|> | The electronic one has an AAA battery and a bistable relay; it
|> | doesn't draw any current through the lamp and only draws significant
|> | current from the battery twice a day.
|>
|> I wonder how hard it would be to design one that stores energy in
|> a rechargeable battery, charges it only when the light is on via the
|> current going through to power the light, and runs the electronic
|> timer from that battery when the light is off.
|
| This is how many older setback thermostats worked; they drew current
| through the fan relay coil in the furnance (relay off) to operate the
| clock motor and charge a small NiCd battery. When the fan relay was
| on, the clock motor was powered by the battery.
I'm suggesting that the battery be charged during the _on_ times, and
the clock run from the battery during _off_ times. This would not work
for a motor clock. But a CMOS clock could do it.
| One problem with using a rechargeable battery in this application, even
| in a circuit that is designed for it, is that some chemistries (NiMH in
| particular) have a rather high self-discharge rate, on the order of 1%
| of remaining capacity a day. If the lamp was off for a long period of
| time - manually shut off by the user, or burned out - the battery could
| discharge fairly quickly.
Or in the case I'm suggesting, the device under control not be turned on
for a very long time. It would need to have a manual-on button _and_ do
the charge during manual on. Then you could simply turn the device on
via the manual on button (closes the circuit long enough to get enough
current to turn the solid state switch on).