Little A - 2xCR2 light - long

HarryN

Flashlight Enthusiast
Joined
Jan 22, 2004
Messages
3,976
Location
Pleasanton (Bay Area), CA, USA
Little A (very long)

This light is actually part of a journey I had never intended to make, not a finished product, but I decided it was time to post it. Perhaps the most important thing in this post is not the light, but thanking everyone on CPF who has contributed to my understanding of LEDs, electronics, and optics, as well as the encouragement needed to go on through some very difficult technical problems. Some names that come to mind are Gadget Lover, Doug Owen, Doug S, Evanxxxx, Newbie, Mr. Al, PaulB, Georges80, McGizmo, and my brother (EdN) who rarely posts due to a very slow internet connection. Sorry for the ones I missed on the list.

I apologize that I am having trouble with pictures today, so I will try to add them in at the appropriate places next week.

This particular light is just a "demonstration mule" that was built to double check some concepts for a light that is coming down the road, most importantly, the OPTICS, which is perhaps my contribution to CPF.

It started life as a 2xAA MM which I borrowed from my daughter. If I had realized that the purple ones were so rare, I might have just purchased a new light instead, as now I have to actually finish it. This one had been regularly used by a 9 year old girl and her 7 year old brother, so it had scratches and was not working. Since it is now her light again, I gave it her family nickname "Little A".

Gadget Lover bored it out for me using his lathe from AA size to CR2 size. Due to the original planned length, he had to bore it from both sides. The tail cap threads are just a tad small, so they had to be shaved but are intact. He managed to insert the boring bar through that small head end as well to bore from the "head end", but I eventually just cut if shorter using a manual tube cutter. The boring bar did not quite reach from both ends, and it appeared to have flexed in its journey through the tube, so additional material was removed by mounting a dremel drum sanding paper on a doll rod driven by a drill.

(Picture of battery tube will be here)


Heat Sink

I did not want to take the time to have a heat sink machined for this one off, so instead I made my own. I took a copper rivet and spun it in the dremel tool on some sandpaper to the right diameter, then flattened the face on sandpaper the same way. The rivet was then cut to the final length (about 10mm). The Lux III slug was attached to the rivet using silver filled epoxy, as well as the cathode (ground) of the LED. Technically, you are supposed to isolate these electrically, but I experimented with this setup on the bench, and it worked fine. I suspect that it is more of an issue if the slug and ground path are at different voltages, such as when you are using a constant current regulator.


The second part of the heat sink was made from aluminum foil, a trick posted by another CPFr. I made a cylinder of foil by first making a "cup" shape and pushing it part way into the battery tube. Additional foil as rolled up and folded so that the end would fit into the "cup", and stick out about 250mm (9 inches). This entire setup was compressed down using a doll rod and a hammer until it formed a perfectly fitting cylinder the size of the "head end" of the battery tube x 15mm long. Remarkably, if you really compress the foil down, it is a very convenient starting point for further work. I trimmed this Al slug down to 13mm, cut a wireway slit in the side and did some trimming work on it with a dremel tool and an exacto knife. A hole was drilled down the center to hold the copper rivet (described above) using a normal power drill. You do have to pay attention to keep it from unrolling, but after 3 tries, I managed to make a decent one the right size.

The silver filled epoxy was cured in an oven first preheated to 250 F (its lowest setting), then turned off, cooled for 10 minutes, then the assembly was put in for 3 hours. (covered with Al foil). It is important that the Luxeon does not see high temps, and radiant heating from the electrical elements cycling would put it over the edge. I have used this setup before for some bench tested / abused Lux mounting, and the attachment is remarkably tough. In that test, the joint survived multiple 50mph impacts to solid concrete.

(I will try to post a picture of this heat sink here next week.)


I also tried using Al tape to make this section of the heat sink, but the gummy layer was getting in the way of post processing. Basic Al foil worked quite well, and as Gadget lover will attest, was quite effective as a heat path to the mm tube body.


Electronics

I am partial to simplistic electronics, so I set this one up as 2 x CR2 and a resistor. The "circuit board" for this light was cut with a dremel from a R/S PCB with pre-punched holes and copper surrounding each hole on one side. The anode from the LED is soldered to a wire from an alligator clip, run down the side of the heat sink, and mounts through a pcb hole at the center. A 1206 surface mount resistor was soldered to the wire, then soldered to a spring for the battery anode contact. I like using a spring on both ends of the batteries – it makes up for small fit errors, and adds additional shock resistance.

This particular spring was from the original MM tailcap. The resistor is a 10 ohm, 50 watt, 1206 size resistor I purchased for the original project (not this one). Its power rating comes in part from being made on a high thermal conductivity substrate. My soldering iron is actually an old Weller "Gun" far to large to solder SMT parts properly, but I was able to heat the entire resistor from the center and solder both ends thanks to the thermal conductivity of the resistor.

In some ways, this was the most difficult part of the project – not the soldering, but using a 10 ohm resistor. I had planned on using a 2 ohm resistor, which would provide something like 700-800ma and LOTS OF LIGHT, but my daughter (the original Little A) told me that "Dad, your lights are too bright. Could you make mine just a normal light". Being the softy I am for her, I bench tested various current levels on an identical LED / Optics setup for her to choose – 200ma was about right, so 10 ohms became the plan.

In the end, I managed to short out the 10 ohm resistor when soldering on the spring, but did not realize it until I had permanently mounted the entire assembly in the tube, meaning – it was now in direct drive. This mounting operation also slightly damaged the LED dome, so it is not quite as nice as the bench test beam, but still acceptable for the use.

Gadget Lover came to the rescue with his history of developing 2 stage tail caps. He mounted a 20ish ohm, 1/8 watt axial resistor in the Kroll, allowing it to run in spec, and have approx 100ma of current. (Thanks Dan) Perhaps he will post on exactly how this is done.

In the meantime, we had a chance to run the light for short periods in DD (approx 1.7A) through a Lux III TWOK. It was able to outshine his SF L4, my "target beam" light. It gets warm fast, and DD is not an acceptable setup, but it was fun to see.

OPTICS

This is the area where this light is actually unique. Most lights use a reflector and head which is considerably larger than the battery tube. In my case, I have a light under development where the space for the optical package was designed too small for even a 17mm reflector, in fact, it will only handle a diameter of 12.5mm (1/2 inch) and approximately the same length including the Lux III. There is very little available in this size range, and believe me, I have tested anything I could get my hands on.

This was clearly a mistake / lack of knowledge based error, but I was too ignorant at the time to understand just how hard it would be to obtain a decent beam from this small of a volume. (The size of a 1/3 N cell) Over a 12 month period, I tested just about every commercial or privately made optic, ball lens, aspherical, and reflector I could find. If I were an optical engineer, it might have helped save some money.

Finally, I was looking at my car one day, and noticed that it contained both a reflector and a Fresnel lens. For those that have not had the pleasure of exploring optics while desperate, a ½ ball lens is a great simple lens for taking light from a point source, and focusing it to be relatively collimated beam. An asphirical lens is similar to a ½ ball lens, except some reality corrections have been shaped into it. Both of these setups are quite effective, and due to high volume, large, high quality asphericals are pretty inexpensive. Some show up on the front of common flashlights and appear as a "bulge" on the front, such as some TIR setups.

For reasons that still do not make sense to me, only the very front of an aspherical lens actually does the focusing, so the bulk of the material is just a source of loss. This was recognized a long time ago, and several optics experts from centuries ago figured out how to cut this down to an array of lenses on a plane. This was largely optimized by a guy named "Fresnel", and he made a fortune selling these to light houses.

A Fresnel lens by itself still does nothing to capture light which has not reached it (such as the light coming out the side of the Lux III), so occasionally, optical setups are designed to include a reflector to "get the light going the right general direction" and the "Fresnel" to focus it. It turns out that the concept is much simpler than the actual implementation, and most optical designers will correctly tell you that it is such a PITA, that it cannot be done properly. They are correct on both accounts – it is a PITA, and it cannot be done properly under most circumstances, but it can be done well enough approximately to make it worthwhile.

I will save all of you the bother of attempting to find the combination of an existing reflector and Fresnel lens which will work for a Lux III – there is not one. I went through the considerable challenge to have a set custom made, and now have one for 12mm, and one under development for larger sizes.

Below, you will see the picture of the reflector (when I get the pics working) in the 12mm size, and the one I used for this project, which was made LARGER to fit the CR2 tube by wrapping thin strips of Al tape around it to grow it concentrically.

(Upcoming picture of custom reflector)




Beam Measurements

There are a lot of ways to optimize a beam. Some people like small hot spots, some like flood, etc. My particular favorite is the SF L4. It is quite bright, can light up a room, yet is very effective for distances in the range of 100 – 200 ft. This general beam shape became by target beam shape, and I think it came out pretty close.

(Pics of beam will be coming here)


I took the light out for a walk the other night, and was actually fairly impressed. Even on 100ma, the light easily lights up trees at the park, and while it will not overpower a street light, it will light up dark areas on the other side of the street.

The beam is nice and smooth in the central area, and has a low level corona. I will post more beam info later.

(BTW, my Little A likes her new light)

Thanks for reading this long post, and sorry for the short term lack of pics.

HarryN
 

greenLED

Flashaholic
Joined
Mar 26, 2004
Messages
13,263
Location
La Tiquicia
/ubbthreads/images/graemlins/kewlpics.gif /ubbthreads/images/graemlins/smile.gif sounds like a neat project (and optics)! waiting for pics.
 

greenLED

Flashaholic
Joined
Mar 26, 2004
Messages
13,263
Location
La Tiquicia
/ubbthreads/images/graemlins/kewlpics.gif /ubbthreads/images/graemlins/smile.gif sounds like a neat project (and optics)! waiting for pics.
 

Neg2LED

Flashlight Enthusiast
Joined
Apr 25, 2004
Messages
1,127
Location
'straya, mate!
NICE! as has been said before,

BRING ON THE 'This Thread Is Worthless Without Pics' SMILEY!

neg
 

gadget_lover

Flashaholic
Joined
Oct 7, 2003
Messages
7,148
Location
Near Silicon Valley (too near)
I was really suprised to see that this little light, with the optics inside the battery tube, was brighter than my Surefire L4. That was a T bin Lux III running over 1 amp, but still, I was impressed.

I'll be giving Harry a hand posting his pictures.

Daniel
 
Top