Yet another mod: UKE 2AAA

Steelwolf

Flashlight Enthusiast
Joined
Feb 6, 2001
Messages
1,208
Location
Perth, Western Australia
I wish I had a digital camera. Sorry, no pics for this till I can borrow a camera.

The essence of this mod is to create a small pocket LED flashlight that is waterproof, not just water resistant.

When I gave some of the Arc-AAA away to friends, I had to warn them that it was basically a Maglight Solitaire but with more constant output and whiter light. And though the "sun" mode burn time is about the same as the Solitaire's burn time, it can still carry on producing dim light for long after the Solitaire has given up the ghost.

However, the down side was that the construction of the Arc-AAA only renders it water resistant. Sure its fine in heavy rain, but what if you go swimming or surfing alot, or like me, get drenched by heavy swells while fishing. Some friends of mine prefer the UKE-2AAA to the Solitaire because of its water-tightness (and slightly improved output).

I looked around and found the Trek/C.Crane and Pelican stuff, but the small ones were not all that small and required either some lithiums or 3xAAA. The 3xAAA one would probably be a little bulky in the pocket as I tried out by taping 3 AAA batteries together. Also, Craig had complained about the difficulties of changing the batteries in his review on his website. The Pelican light has a rubber switch behind, which in my experience, would quickly dry out and crack in this humid weather.

So I took my friend's old favourite UKE-2AAA and drilled out an old bulb assembly to accommodate the 5mm 5.6cd white Nichia LED. I know there are brighter ones available, but that was all I had for the moment. I'll be visiting my local electronics mart (Sim Lim) when the shops open after the Chinese New Year. To reduce light lost to the side and rear, I painted the sides and back with gold coloured paint. I chose gold because the light was already quite blue and I hoped the reflected gold light might impart a slight hue to make the light more pleasant.

Behind the bulb assembly, I built a full SatCure circuit (1 resistor, 1 transistor, 2 capacitors, and a strangely wound inductor), instead of the simplified circuit which has no capacitors. I found that the capacitors visibly increased the light output. This was put together free standing, i.e. no circuit board, each component directly soldered to the next. This was done so as to allow the circuit to fit in a space similar to 1 AAA battery.

After testing to ensure that everything was working, I put a mould in place and encased the circuit with styrene resin (Diggers Casting and Embedding Resin). This wasn't easy as the mould leaked and I had resin flowing everywhere. Fortunately, I managed to get the mould to stop leaking and embed the circuit. This is needed to impart strength to the circuit assembly.

When completed, the entire circuit looks like a AAA battery has been glued on to the back of a standard UKE-2AAA reflector.

The negative terminal remains the brass ring around the reflector and the positive terminal is the bottom of the assembly. This allows the device to be inserted as though it were the second battery for the flashlight and to be activated in the usual manner (turning the bezel). The case was not modified in any way.

Another advantage of this system is that the UKE system replaces the entire reflector/bulb assembly when changing to a fresh bulb. So the user can decide if he wants a xenon bulb or LED. (More likely LED as they are almost equal in brightness and the LED runs longer using only 1 battery instead of the 2 required for the standard bulb.)

I was quite pleasantly surprised to discover that the light output from this circuit is similar to that given by the Arc-AAA with fresh batteries in both. Though of course there is no regulation so the output will dim as the batteries die, but the specs for the circuit claim that it can work to 0.7V, and that would take quite a while to reach.
 
Very, very cool!
cool.gif

Post pictures soon and if you could include a diagram of the circuit that would be great too. That way I could look all in one place to try to build this mod. I think I might try it using a Princeton Tec Blast body.

I don't understand electronics, but I can follow directions quite well. I once built a AC to DC converter this way.
smile.gif
 
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Steelwolf:

Behind the bulb assembly, I built a full SatCure circuit (1 resistor, 1 transistor, 2 capacitors, and a strangely wound inductor), instead of the simplified circuit which has no capacitors.
<HR></BLOCKQUOTE>

Please do post a diagram of this circuit that you used. I've always like the SatCure concept and I'm interested in any and all variations. Also, please list the actual transistors used and details of the inductor, core type and windings.
 
Sorry. Having trouble getting hold of the camera. But you can view the circuit h ere.

Recently went to the local electronics mart and found some electrolytic transistors that are about 1/3 the size of those used in my original circuit. Don't know how they did it, but this might make it possible for an even smaller set up, like the Pelican Mini Mity Lite (1AAA sized). We'll see.
grin.gif
 
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Steelwolf:
Sorry. Having trouble getting hold of the camera. But you can view the circuit h ere.
<HR></BLOCKQUOTE>

Thanks for the circuit drawing. Don't have that one in my collection. Do you know the source of that circuit. It looks like there was once more to the drawing. Any details of the toriod core used???

Will be nice to see pictures of your mod. Hope you can get ahold of a camera.

Good luck in all your endeavors.
-Mercator-
 
I have a question.
I always have used the original satcure design. I built the many other variations...Gadget's etc...and never really got much better performance. I did get less battery life however. My only change to the original design is the changing of the resistor to as high a value as the circuit will allow...for a nightlight. I never use an on off switch as the draw is so light. I do use a toggle to have max bright/max dim.
I can get the whole circuit to fit in a straw...seal with epoxy..snip it...and I have a circuit extremely small.

I compare the current through the LED and through the whole circuit and try and get these values as close to each other as I can.

My question...adding all the other components seems not to give more light etc but add to the total consumption of the circuit. Am I missing something?

I still would one day like to have the Arc circuit with 3 wires that was sort of promised...one day...sometime in the future.
ledcct.gif

Original Satcure location...
 
Mercator: There is more to the drawing. There was an article attached and another circuit. The article talked about the 2 circuits and the other circuit was a pulse-skipping circuit for when there was a fluctuating load. Not necessary for our purposes as the LEDs do not fluctuate in their demand that much to warrant another 6 or 7 extra components. I can put it up for scrutiny if anybody wants. The circuits and article were, I believe, presented by a guy called Z. Karparnik in some magazine about practical electronics.

The toriod is a small high-frequency ferrite bead about 5mm high, 4mm outer diameter, 2mm inner diameter. I used 0.125mm enameled wire. The windings are done in the following manner: take a length of enameled wire about 0.5m long (approx. finger to centre of chest) and double it over. Wrap the wire tightly around the ferrite core. Each wrap lays down 2 strands of wire. Make 20 such wraps, keeping the wires close together. When all the wraps are done, cut the end that is doubled over so that there are 2 ends. You will now have 4 ends total, the 2 from the start and the 2 exposed when you just made the cut. You need to join one of the ends from the start side and one of the ends from the cut side, but they must not be the same wire, so check for continuity and join the two ends that show no continuity. Remember that one end to be joined comes from the start of the winding while the other comes from the doubled end that you cut and they must not be the same wire.

That will give you the required flyback inductor (learnt this new term recently
smile.gif
). The joined ends go to the battery and the other 2 ends are interchangeable.

Wolfman: I'm trying to check out the battery life thing, but don't have a set up to measure light output versus time like Craig's yet, so I need some free time to set it running and just keep looking at it. I will check the current draw and minimum voltage in again to get an idea of run time. I found that the extra transistors help quite a bit in the quality of the light. The peak voltage output was very high without the 47uF transistor in place that the LED was more bluish and didn't look as bright. The other 0.47uF transistor is to help deliver the higher currents required when the inductor is charging. This reduces peak current demand on the battery and hence reduces the effect of the battery's internal resistance. (That's my take on the situation anyway.
smile.gif
) The diode is a necessary evil to prevent the 47uF capacitor discharging to the wrong components.

Since I was going for the brightest light possible, I used a 100ohm resistor instead of the setup of resistors shown in the diagram. I think I tried lower values once before and found that the light wasn't any brighter below 100R but consumption went up.

I also wait for the day when Gransee starts selling little hobby kits of the circuit he uses. I can't get the hang of soldering SMDs. Burnt out my MAX1676EUB which was part of a kit from **** Smith Electronics. It was supposed to have a feedback loop to regulate the output.
 
Steelwolf

Thanks for the picture of your mod and text on the Satcure ckt. that you used. I'll add your instructions to the rest of info for the Satcure that I've been collecting.

-Mercator-
 
Okay, I know this is a way old thread, but the idea of putting this step-up circuit into a battery casing is excellent.

Why doesn't anybody sell these? I just bought a ton of mini-mag knockoffs and something like this would work great! (see here: http://www.candlepowerforums.com/ubbthreads/showflat.php?Cat=&Board=UBB14&Number=342555&page=0&view=collapsed&sb=5&o=&fpart=1 )

I see so many people talking about SatCure circuits, step-up circuits, etc. but haven't found anyone that mass produces them. I don't know how much the individual parts cost, but there should be a way to mass produce them and stick them in a casing to give you a "dummy" aaa, aa, c or d battery.

I'd love to buy one of these circuits in the form of an AAA battery. If not available, then I'd like to find a place that sells the circuits... and if not that, then buy a kit with all the pieces.

Thanks all!
 
WRT the SatCure circuit, the individual parts are quite cheap and easily available. But putting it together is not that easy. Well, it could be, but you would need some quick and easy way of winding the flyback transformer. Perhaps it would be simpler if it were possible to buy tiny little ferrite rods. Just clamp them in a drill and run the 20 turns you need. But we mostly use toroid cores which are a little fiddly to wind, especially the tiny ones that are needed for the really small mods. The advantage of the toroid core over a straight rod is that there are less field losses. That's an important point when the power source is already so tiny.

WRT just any step-up circuit, complete circuit, there are a few out there. The various Dat2Zip sandwiches, the MicroPuck and PowerPuck... erm I think that's it. As you can see, there is no reason to take up the entire space of a dummy cell when you can use SMD parts and make it tiny enough to fit in to dead space.

As to just building the circuit into a dummy cell and popping it in place, I think it might be possible. The only things would be to ensure that the circuit can be completely cut off from the power source so that the battery doesn't run down (probably a tail cap switch in the case of Mags), and that there is no output capacitor, or that the output capacitor is connected permanently to the LED array, not the dummy cell with boost circuit. This is because the capacitor can blow the LED if it gets charged up while the LED is disconnected, and then gets connected to the LED.
 
Top