LED Lantern with LM2576 Switcher
This project was put together in the spirit of "I want to be prepared". Last November when we had hurricane Sandy, we came so close to losing power. A lot of utility poles were snapped like they were matchsticks and were lying all over the place in our neighborhood afterwards. Luckily, the power didn't go out but we lost our Time Warner internet service for nine days, at the time when I was working online on getting the MIT 600.2x done. Now there's another storm, Nemo, barreling down on us and I don't want to take any chances. You only get to appreciate something when that something is not available.
My wife bought something on Amazon.com and was given a LED lantern (see below) as a throw-in. It has two separate LED panels that can be folded up and rotated 360 degrees, independently. There is a nifty carrying handle and a three position switch; OFF, one panel ON and both panels ON. Looks like this feature was implemented to save power; when you really don't need the full output, you can click the switch until there is only one panel lit. The lantern is to be powered by hefty three D-size dry cell batteries, connected in series. The batteries are housed in the triangular black plastic box under the LED panels. The three heavy D-size batteries should significantly increase the stability of the lantern because if empty, it hardly weighs anything.
To insert the batteries, there is an access door/tab on the bottom of the lantern. Notice the cracks around the door as the plastic molding job leaves a lot to be desired. That's kind of funny because the little paper booklet that came with the lantern tries to pass it off as a piece of waterproof equipment, a storm lantern as it calls it! I personally don't think it's going to work in any kind of storm. But we got it as a freebie!
When it comes to spending money on large batteries that otherwise have no use in our household, I'm not your man. On the other hand, I have quite a few friends that have been giving me their somewhat old but fully operational Lithium Polymer batteries or Li-Po's for short, such as the one shown below on the left.
The one I decided to employ here is a respectable 12 Volt, 2200 mAH unit, kindly donated by my friend Tony. The 20C discharge rating doesn't matter here because in this application, I'll be drawing energy from it very little at the time.
The trouble is, if I were to connect it to the lantern directly, the battery voltage is simply too high. Thus a voltage regulator is needed.
To conserve energy, I decided to use a SMPS-type regulator. I happened to have a LM2576, made by the National Semiconductor, in my junk box and it found a new home right away. You can download its datasheet from their web site or here.
After a little thinking the circuit below came together. The LM2576 is used here in it's most traditional way. In order to see what's in the belly of the beast I opened up one of the LED panels. It's actually pretty easy if you are not afraid. There is a flat piece of plastic that covers the circuitry behind the panel and it is actually both glued and screwed to the face. I had no choice but to use a little bit of old fashioned elbow grease and a flat-bladed knife to pray the back off. When it finally did fly off, it did it with such a crackling noise that made me a little scared.
The LED's inside were soldered to strips of PCB board and all connected in parallel. There were no markings whatsoever as to the "identity" of the LED's. But they definitely must be of some "more powerful" variety because they were pressed against a thin metal foil that doubles as a reflector. Or should I say reflector that doubles as a head sink. Anyway, I guess that you, my reader, are already getting the idea. I am not really impressed with the quality of the workmanship here.
Each panel also had one 1.5 Ohm 2 Watt ballast resistor. When I tested the panel under full voltage (I decided to raise the voltage slightly, up to 5 Volts) it was found that resistor dropped about 1.2 Volts. That left the rest, 3.6 Volts, on the LED's. Both panels together drew about 1.6 Amps of current, thus drawing ~8 Watts at 5 Volts. Typically you can't parallel diodes like this because they are never the same and one specimen always starts drawing more then the rest of the pack, thus becoming the candidate for dying. But here I think the diodes were from the same batch and/or perhaps specifically manufactured for such an application.
The way the three big D-cell batteries were housed was that the two were placed side by side and the third was underneath them, deeper in the plastic housing of the lantern.
Since there was now room left over, I decided to squeeze the quick-and-dirty SMPS board right in there as you can see in the picture on the left. The object on the right is the Deans connector. There's really no need for it because of the current draw requirements but the battery already had one on it so I soldered the male kind here to the wires.
Here is the finished "product", lighting up what looks like a snow cave or igloo :-). No worries, we don't live in an igloo, at least not yet. It's my wife's photography arrangement, a bath towel thrown over two nearby pillows to form a cavity between them and simulate the snow.
When tested, the completed circuit drew 788 mA at 12.4 Volt fresh battery. This gives us 9.77 Watt input. I had already calculated the output power, at ~8 Watt so the efficiency works out to about 81.8 percent. Not too bad!
And another shot with the panels slightly unfolded.
Official storm testing is Astoria Park. Looks like it's working all right.
Let there always be light, 73 de Brian.