Henry_Best:
I've emergency lighting in my home. It's just a car battery charger, a car battery, a 12V relay and a 20A power diode.
It's wired so that, if the mains fails, the battery charger stops powering the relay coil, the relay drops out and the battery powers up some 12V lights. The diode is needed to prevent the battery powering the relay coil. The battery is on trickle charge 24/7. I've added a push to break button in the 12V line to simulate a power cut and check that it's working, as all the components are hidden in my attic.
What a simple and cool configuration. However, I'd be concerned to have it on the attic with a relay coil energized 24/7. Anyways, I think you've look into this and made sure it's all good. Thanks for sharing!
Paul__B:
The rule is, you put it all in a proper steel equipment case.
Apart from the (double insulated) battery charger, its all 12V. The charger is fused on both the 240V input (3A) and the 12V output (15A). It's in a dry location with plenty of ventilation. I don't see any need for a steel case.
I have several devices that can detect power without contact but even though they detect power, it may not be usable. Brownouts or the loss on a single phase. The bigger problem is trying to detect power in open air has to be sensitive enough to detect, it would probably detect the arduino also. Just looking for an RF signal to drop out does not constitute a power outage. There are other factors that could cause a loss of signal. When I first read your post my immediate thought was look at the clocks on your appliance and see if they were flashing.
Hack the output section of a cheap "non-contact" voltage stick and use the pulse/no-pulse via perhaps a transistor to match voltage requirements to activate your arduino.
Paul__B:
12 V at less than 15 A is plenty to start a fire.
If you are not worried, fine. If he is worried, put it in a steel case.
15A is only drawn when the battery is fully discharged. Most of the time there's only 500mA (trickle charge to the battery and the relay coil) in the circuit.
