LEDs are current driven devices. this means you can supply them with whatever voltage you want (just about),.. as long as you control the current going to them with resistors. since each LED needed 30mah to produce its brightest light, this meant each "light head" needed to be provided with 360 mah,. (30 mah x 12 lights = 360 mah) I soldered them all onto a proto board in parallel and decided on a 3 AAA power source that would give me a running voltage of ~3.6v (was actually closer to 4.2v, batteries tend to be a bit more powerful than there rated values) I get most of my batteries from http://www.all-battery.com/ for no other reason than they tend to run a lot of sales and have decent prices, that and there product seems to work as advertised. I ordered 6, 1000mah AAA batteries so that I could make two battery packs.

With the batteries at 1000mah, and the drain of both heads together at about 720mah I figured I would get about  an hours worth of time out of each battery pack before the batteries started to weaken and lose power.

About this time I started to realize the error of my ways,.. To get maximum efficiency out of LEDs you need to provide them with a set current based off of there peak performance,.. The current is set with resistors based off of the voltage of the power supply (using ohms' law). The higher the voltage the larger the resistors that are needed. however a battery is not a set power source, Ni-Mh and Ni-Cad batteries have a curve, as you apply load, voltage drops, and as you drain power, the voltage drops even more. This caused a problem, if I set the resistors for the batteries peak of 4.2v then 5-10 minuets into use when the battery has drained a little bit to lets say 4v, now the resistors in the circuit are to large for the new lower voltage and the LEDs are less bright than they could be. This degradation of power to light and thus exposure would continue thought the entire cycle of the charged battery and I would only get "peak" performance for a few minuets at the very start. It would work, but I would have to constantly be re checking my exposure and adjusting to less and less light while I used the lights.

The solution is simple in concept, regulate the power supply to a fixed voltage for as long as possible. this is done with a voltage regulator a simple IC that takes your supply voltage and drops it down to a pre set regulated voltage. I looked around for 3.8v regulators as this is the balancing point of the LEDs I was using, where if they are feed this voltage they need no resistors. The closest I could find where 3.3v it would work but I would never get peak light from the LEDs,. I also tested adjustable regulators, but they seemed to float a little to much when the supply voltage slipped downwards making them (or at least the ones I tested) a little to unstable to work with for maximum efficiency. I did more searching and decided on fixed 5v regulators as my regulated power supply, this was mostly due to the fact that many electronic devices use 5v for there logic level so there are an abundance of 5v regulators out on the market to fit just about any application.

This of course caused additional problems, my power supply was a battery pack that delivered 4.2v max. To use a regulator you need to provide it a voltage source greater than the voltage its regulating to. If I was going to regulate to 5v, then I needed a power supply greater than 5v. Of course there was no real way to pack 5 AAA into my design which would have given me a voltage of ~ 6v, further more even if I could, 6v is really close to 5v and as load was applied and the battery drained a bit I would be back in the same boat of fading power about half way through the batteries cycle life.

I did two things to combat this problem. First I decided to switch to Li-po batteries. with them I would easily be able to pack 7.4v @ 1250 mah into a tiny space within the base of the light. In addition Li-po packs tend to provide slightly more stable power throughout most of there charge only really dipping down harshly towards there drain point. Next I searched for a 5v regulator that was stable at 5v even when the source was as low as 5.5v. On Digi-key.com I found part "497-1404-5-ND". This regulator, if set up properly, should give me a stable output of ~ 5v while the battery is providing a source voltage of 7.4v-5.5v, a much larger window of stable power.

Of course this opened up a new problem, Li-po batteries are a bit tricky, they pack a lot of power into a little package, but they have fickle charging characteristics. Li-po packs don't like to be drained all the way down, most mass produced devices have built in battery monitors that cut of battery power when the pack is drained down to just above its danger zone. My device being hand made and original designed for Ni-Mh which actually benefit from being drained all the way down had no such circuit. I decided to go with a basic voltage monitor that would let me know when the power dropped below a certain point. I decided to have two indicators. one yellow LED that would indicate when the battery dropped below 5.5v this would tell me when the voltage regulator would start to fail and provide less than peak output. and another red LED indicator that would signal when the battery dropped below 5v telling me when it was time to switch batteries or risk damaging the pack.

I accomplished the battery pack monitoring circuit buy following a simple diagram at http://www.free-circuit-diagrams.com/ duplicating this circuit twice (once for each indicator) was about the simplest solution I could come up with.

Parts are on order, this project will continue at a later date.