Would this concept light be workable?

Owning a Mini 123 has reinforced the concept that convenience of size directly relates to real world usage in an edc light.

With size in mind, I can't get the following light idea out of my head for an edc based off of an 18650 or two sub "c" rechargeables. Some of the sub c batteries are advertising 5000mah!

My thought is to build a light that is just a little longer than the battery itself, utilizing a twisty action, and using a series of emitters( 7 or 8 maybe) that are situated around the circumference of the battery barrel. The emitters would not be placed at the end of the battery barrel, but maybe an inch or less from the end. The head of the light, in this design, would also house a good portion of the battery in its core. Since 7 or 8 emitters would be used, the outer circumference of the head could be kept quite small. The difference in the diameter of the battery tube and the diameter of the head would only need to accommodate little more than the width of each emitter itself.

Since we are dealing with multiple emitters, a standard reflector is not needed. Just polishing the exposed portion of the battery barrel as well as the inside surface of the head will work. The reason for this is that the beam signature of each individual emitter is not important. Since the emitters are placed around the circumference of the battery barrel, the beam profile from each emitter will be superimposed on top of one another while being rotated in 8 equal steps through a 360 degree range. In this fashion it would be difficult to not come up with a near perfect beam.

What I don't understand is how lumen output and efficiency come into play when working with multiple LEDs compared to driving a single led. In other words, to achieve 200 otf lumen's, what sort of difference in run time would you get by driving 8 LED's vs driving a single led to achieve identical lumen output?

I am certainly not an expert, but I think one of the reasons for the 'Pill' is to whisk heat away from the delicate LED and transfer it to the aluminum body.

I don't think that having a bunch of power LEDs right next to a battery seems like a great idea...

The idea is quite sensible. For example, if you use one REBEL LED at 500 mA current it means 130 lm lightbeam and 1.5 W power consumption with 1 W of it dissipating as pure heat. In case of six REBELs a 130 lm lightbeam is achieved with only 1 W of electrical power and only 600 mW lost in heat.

Vaska,

if we follow this pattern of heat production and efficiency further toward a goal of 200Lm or more, and increase the number emitters up to 8, then we may have an interesting idea.

Also, using multiple emitters like the above examples, could a single 10000mah D cell at 1.5V produce 200lm or more? What about one of those "F" size batteries?

Joe,

thanks for the insight. I did think about heat issues. The forward portion of the barrel could have a slight expansion built in to provide a small air sleeve separating the battery barrel from the battery. From Vaska's observations, increasing the number of emitters up to 8 may very well reduce heat production so that it is a non issue. The other plus is that what little heat is created is distributed evenly at 8 different points. All of these points are located right at the outer circumference of the head and body of the light for heat wicking, rather than one central internal point with single emitter designs.

The multiple emitter design also lends itself to focusing beam applications quite well. Single emitter designs combined with a moving reflector can make it quite difficult to maintain a smooth beam from flood to throw.

For further information about flood vs power have a look at this thread: https://www.candlepowerforums.com/posts/3115908#post3115908

I don't think, that one 1.5V cell is a good choice, because voltage drop of the switching elements of the power cirquitry is too much relative to low voltage supply. I prefer higher battery tension.

Other than location of the LEDs, this concept is very similar to the 3AAA/multi-LED showerheads.

Problem #1 - balance the current to each LED. Due to manufacturing variations the current to each LED is slightly different. The LED which get the most current will heat up faster causing the LED to draw more current. With 5 mm LEDs eventually one of the LEDs will draw more current than it can handle and go .
With power LEDs there is a very low chance of LED burnout as each LED can handle all the current but brightness will vary.

Problem #2 - To get each LED 1/8th the power each, you need 8 driver circuits. What you gain in removing the reflector you lose in adding that many circuits. Both space and cost. If you try to direct drive the LEDs each LED will try to draw full power. You can use 1 circuit to drive all 8 LEDs but see problem #1.

Problem #3 - Following vaska's link to get 200 lumens from 1 LED you need 500 mA. To get 200 lumens from 8 LEDs each LED needs 25 lumens or 60 mA, total 480 mA. Your improvement in runtime is 4%!!!

Problem #4 - Beam pattern - loss of throw. The reflector provides the hotspot. Take away the reflector and you have flood. Polishing the back of barrel/head would not help much. SIZE of reflector is important.
MTE '900 lumen' P7 left 350+ lumen OTF, 35 mm reflector; 6V lantern with 50 lumen TLE-1F bulb 35 lumen OTF, 100 mm reflector.


I think for the smallest lights look at the Zebralight headlamps. Just turn the LEDs so they face forward instead of sideways.

LEDninja,

I can't comment on items 1-3 that you listed. As for #4 or beam pattern; this design would not be a shower head configuration. The led's would be recessed back as much as an inch on the battery barrel. The width gap between the battery barrel and the head would be small. Just enough room to accommodate the emitter. All interior surfaces would be polished, not just the back surface. Throw would be a function of gap width as well as rearward placement distance.

From someone who helped test the Lupine Betty beam angles($1000 bike light system), throw is highly over rated for real world applications in my opinion.

If the circuitry really is a deal breaker then thats that. Perhaps someone else can verify your points or offer a different perspective. I do have a few of those budget 3AAA shower head lights. The difference is that the head of these lights comes after the battery tube. Since there is no battery tube core to function inside the light head as an internal reflector, these lights become all flood. Plus these 3AAA lights have led's that have no depth of placement which also creates excess flood.

How about this battery? http://www.batteryspace.com/limnnirechargeable26650cell37v4000mah10arated148wh.aspx

LEDninja said:

Problem #1 - balance the current to each LED...

Problem #2 - To get each LED 1/8th the power each, you need 8 driver circuits...You can use 1 circuit to drive all 8 LEDs but see problem #1.

Click to expand...

Boost converter plus series connected LEDs save it.

LEDninja said:

Problem #3 - Following vaska's link to get 200 lumens from 1 LED you need 500 mA. To get 200 lumens from 8 LEDs each LED needs 25 lumens or 60 mA, total 480 mA. Your improvement in runtime is 4%!!!

Click to expand...

You a looking at the wrong column in the table for XP-G, the owner of the least regressive lumen per watt curve. Critical value is Pin. Improvement in power consumption is about 25% even with the best of the existing LEDs

vaska said:

Boost converter plus series connected LEDs save it.

Click to expand...

We're talking about 8 leds from one D cell? So, you are suggesting that the leds be wired in series for a total of 27 V and you want to boost a 1.5V battery to 27V?

Am I understanding you correctly?

gcbryan said:

We're talking about 8 leds from one D cell? So, you are suggesting that the leds be wired in series for a total of 27 V and you want to boost a 1.5V battery to 27V?

Am I understanding you correctly?

Click to expand...

Just have a look at what I wrote earlier:

vaska said:

I don't think, that one 1.5V cell is a good choice, because voltage drop of the switching elements of the power cirquitry is too much relative to low voltage supply. I prefer higher battery tension.

Click to expand...

One Li-Ion cell quite fits such combination.

3.7 volts boosted to 27? If I tell that to my electrical engineering crew they will be giving me dirty looks again. 8 tiny tiny boost drivers might be better for 8 LEDs. In fact there is a driver that I know of that can give about 20 lumen from an XP-G and not much bigger than the LED. Problem is 8 LEDs and 8 drivers is going to be close to 80 dollars...

KuKu427 said:

3.7 volts boosted to 27?

Click to expand...

I can report that it works at approx 85-88% efficiency. I'm a headlamp manufacturer, and my lamps are complemented with LT3477 boost converters. Once I had a chance to make a couple of custom lights with D-type x 4 cellbox, four MC-E LEDs with 2P8S connection and 20 W consumption at TURBO mode. Quite a realistic task.

maxrep12 said:

LEDninja,
As for #4 or beam pattern; this design would not be a shower head configuration. The led's would be recessed back as much as an inch on the battery barrel. The width gap between the battery barrel and the head would be small. Just enough room to accommodate the emitter. All interior surfaces would be polished, not just the back surface. Throw would be a function of gap width as well as rearward placement distance.

Click to expand...

There is a company who owns the patent regarding parallel shiny surfaces to project light forward from recessed LEDs and they sue to protect that patent. (Peak was sued over their original Matterhorn 3 LED lights - the surfaces of the holes of the current model is no longer parallel)

-

My 1st LED EDC is the Dorcy AAA.
http://www.flashlightreviews.com/reviews/dorcy_aaa.htm
The head is only 1/8 inch thicker than the body but it is the part that pushes against my thigh and very uncomfortable. I finally found a replacement for the head that is the same width as the tail (see my avatar).
My current EDC is the EZAA. It is as wide as the Dorcy's head, but because the head and body is the same width, is more comfortable.
-
My main concern with this concept is the large increase in width to save a little length may end up with a light that is too fat and bulky.
The Mini123 is 2.3"L*0.8"Ø or 58.5 mm L*20.4 mm Ø.
The rebel is 3.17*4.61 mm so that adds 2*3.5 or 7 mm. The metal for the head adds another 3 mm for a total 10 mm. Your light width will be 20 + 10 mm = 30 mm.
Your light length will be 3 mm back + 6 mm spring + 34 mm battery + 3 mm front of battery tube + 3 mm circuit + 3 mm front of light = 52 mm.
I do not think replacing 58.5 mm L * 20.4 mm Ø by 52 mm L * 30 mm Ø is worth it. The 30 mm Ø is too fat for the tight pockets of skinny jeans.

-

As the improvement in runtime is not large, you can go down to 4 or even 2 LEDs. A proof of concept light can be built with 2 LEDs in series direct driven with 2 CR123A for low mode or 2 RCR123A for high mode. 86 mm L is still pocketable to some people.

That is currently my favorite battery. It's a beast in a C+ cell size. 4000mah

guiri said:

How about this battery? http://www.batteryspace.com/limnnirechargeable26650cell37v4000mah10arated148wh.aspx

Click to expand...