Any interest in a 42 LED MagLite?

[Go Go Gadget Flashlight]

I have a design for a 3D cell MagLite in the works. Just wanted to know if anyone on the board would consider building (or buying) one, or am I wasting my time?

Only thing I have designed so far is the PC board to make sure 42 LEDs will fit. They do.

And the current draw should be close to an amp (42 * .025 = 1050 ma).

What do you all think? Brock, KenB, are you guys drooling or shaking your heads in dismay?

 

[Brock]

Have you tried running 42 LED's with 3 D cell's to get that current draw? Alkaline really can't handle a draw like that, well it can but the voltage will drop quite a bit. I would guess that many would drop it so much it wouldn't pull that much power?

Brock

 

[Go Go Gadget Flashlight]

Good point Brock.

I don't HAVE 42 white LEDs. Anybody out there wanna test it for me?

 

[Go Go Gadget Flashlight]

Then again, maybe a 4D without any resistor?

 

[Chris M.]

I have a design for 37 of them in a 3D MAG- mounted on a thick disc of aluminium in the head, using thermally conducting glue. This is necessary im my design because in order to really blow the new 19 LED Trek light away (forget its proper name now!) it`ll push them as hard as is safe to do, and the heatsinking properties of all the metalwork means it may be possible to go up to 50mA per LED safely with no long term overheating problems.

Trouble is, I only have 7 of them little white shiny things and no money (well I do but it`s set aside for the impending E2!). One day soon though I`ll do it. Guaranteed I`ll post the full details here too!


How do you get 42 in there? My pattern is sort of circular- I guess you`re using a tighter packing like honeycomb arrangement?

 

[Alan]

You might want to make it on
4C/4D cell light with NiMH batteries. Those batteries could sustain high current draw without signicant voltage drop.

Alan

 

[Badbeams3]

Well...let me add one thought. There is something called wave cancellation...I`m not sure how much of a factor it would play in your design. But it has to do with multiple points of light cancelling each other out to some degree. This is one reasons the soon to be out (maybe...if it`s true) led`s from Lumileds www.lumileds.com will be brighter than multiple led`s having the same amp draw. It has long been known that...all things being equal...two led`s is not truly twice as bright as one. I have heard this called "diminishing return" or something like that. However, I would never want to styfull your creative mind (well...the eye was a bit of a strech ). So while I think it is a cool project, I would wait and see how these new led`s pan out before spending to much on multiple led lights. How about building a led book light? That`s something no one else is doing and would be quite usefull. You could sell them on E-bay to finance some of your projects.

 

[Go Go Gadget Flashlight]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Chris Millinship:
I have a design for 37 of them in a 3D MAG- .... <snip>
.... How do you get 42 in there? My pattern is sort of circular- I guess you`re using a tighter packing like honeycomb arrangement?<HR></BLOCKQUOTE>

I used a Cad program to draw up a perf-board (from Radio Shaft) and took precise measurements of my Nichia 5mm's.
I constrained the board's size to the inner diameter of the Mag's head.
See here:
CAD view of 42 LEDs
(Edited Link above - should work now)
Obviously, I could leave out alternating LEDs to reduce the number some...

I'm beginning to think that it would require a 4 cell to produce enough voltage. NiMh's would be my first choice anyway, I don't like alkinlines much.

Regardless, I'd like to stay away from using any resistors.

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR> By Ken B:
However, I would never want to styfull your creative mind (well...the eye was a bit of a strech ). <HR></BLOCKQUOTE>

But seriously... No, that's why I posted this, to see what everyone thought and to bring up points like Ken's (which I did not know about). Thanks.

As for the book light, I have seen one somewhere (can't recall). If I find it again, I'll let ya know.

This is not intended to be a marketable project. But for the cash outlay to build it, it damn well better be bright!

 

[Go Go Gadget Flashlight]

DOH!!! I just realized, I can't even count!!!

There are 44 LEDs in the design. Sheesh! I need to take a vacation!

 

[Badbeams3]

Yep, I believe a 4 cell mag light would be the way to go and agree with your opinion that resistors may not be nesassary. It should be bright!

 

[**DONOTDELETE**]

I don't know if it's due to 'wave-cancellation' or what, but the 12 white LED Mag conversion I made was originally planned to hold 24 LED's. As I was building the LED assembly, I added the LED's in pairs and checked brightness and throw after each addition. Because of what became apparent to me, I stopped with 12 LED's.

I've come to the conclusion that in multiple LED flashlights, there's a point of diminishing returns ... that point when the cost of the LED's outweigh the percentage improvement in light output. This is all just conjecture on my part ... I haven't backed any of it up with measurements. The only instrument I've used is the good old Mark 20/20 eyball. But the improvement from 8 to 10 to 12 LED's, wasn't nearly as significant as stepping from 4 to 6 to 8.

I'd love for someone to put the matter to a test and measure the effects of adding LED's and then chart the resulting total light output (keeping all other factors constant). From my simpleton's observation, I'm thinking the chart will have parabolic characteristics and not show a straight line improvement.

Oh, my flashlight IS bright. It has a great color temperature and a beautifully diffuse beam pattern. But, to KenB's point, I too believe an improvement in a single LED's output will be more significant than adding more LED's of lower mcd.

FWIW and as something of an illustration of this point: the head I made with 7 blue-green 20,800 mcd LED's is MUCH brighter than the head with 12 whites. Yes, I understand the color differences and perception issues ... but the bottom line is the Blue-green head is BLINDING to look at and the white's merely hurt. (Nice, scientific comparision, huh?).

My two cents worth.

Luffokc

 

[Chris M.]

Sorry but it seems your CAD picture won`t show- something in your web host`s software prevents outside access.

However they`re arranged, one thing I would suggest is- beware of overheating. Cramming that many LEDs in such a confined space, and just mounted on a plain fibreglass PCB, you`re going to have problems if you push them over their rated 25mA maximum. That is why in my design they are mounted on a thick aluminuum disc- to conduct the generated heat away into the head and body of the light which means I can drive them harder (and so brighter) and not worry about frying any.

I would suggest to anyone trying the 44 LED design- test the constructed light pcb mounted into the head of the light. Limit the curent with a resistor arrangement, or if you`re lucky enough to have a variable current/voltage power supply, use one of them- and just ramp up the current bit by bit over a long time until they start to get too warm for comfort. Then back it off a bit and leave it cooking for a while- hopefully they will cool off a little and all is well- then you`ll know it`s safe to run them anywhere up to that current. As for power source? Try all batteries that you can but watch your current draw- and use the one that you`re happiest with in the final light.

After all (depending on where you get them), you don`t want to go frying 44x $3 white LEDs!!! That`s a lotta cash for a lotta melted gloppy blackened stinky stuff (ever melted an LED?)! Ouch!

 

[Badbeams3]

Good point Chris. I begin to have some question as to the outcome. In order to have maximum brightness each Led should be driven around 50 millamps, based on Brad`s research. This might be tricky getting the heat thing figured out. I suppose moving to the north pole is out of the question Gaget?

 

[Badbeams3]

Well, I just pulled the metal board out of my Expedition 7 led light and hooked it up to a 6 volt batt. Within SECONDS it became hot...very hot. I don`t believe your going to get by without some resistors Gaget...no way...not with 6 volts any way. Also the clips I used were the same one`s I used with my W-D 40 experment...and I discoverd they had welded shut all those months ago...not kidding...wow. Says a lot about the durabilty of the LED though...still works! I don`t think your going to be able to run 44 at anywhere near 50 millamps each with out a heavy duty nitrogen cooling system built in.

 

[Go Go Gadget Flashlight]

Well, 50 ma each is never gonna happen with 44 LEDs. That's 2.2 amps!

D alkalines would only give it 8 hours run time.

The whole idea here is too figure out which is better, 2 leds at 25ma or 1 at 50?
Same power in, but which gives off more light?

If I can package 44 at 25ma, heat shouldn't be a problem.

 

[Badbeams3]

Yep, your dead on Gaget. Hmmm, I wonder...does anyone know?

 

[Go Go Gadget Flashlight]

I went looking for that light meter. If I can't find it, I'll never know.

I may have to rig up some test setup like Craig (and others) use. Shine both onto white paper for comparison. But then I have to build another regulated power supply.

 

[Go Go Gadget Flashlight]

Somebody try this link for me please.
http://photos.yahoo.com/bc/bajagadget?d&.flabel=fld1&.src=ph

I don't know if it lets me in because I have the right cookie or what...

 

[Badbeams3]

It`s a pic of a dune buggy, a 44 led light setup, and some insane guy, probably a flashaholic.

 

[Go Go Gadget Flashlight]

Yeah, the more I hear you guys talk about it, the more I'm sure this isn't going to be easy.

I've got an old photo light meter (I think). I'll have to do some experimenting to see what's what before I do too much.

Luff,
When you were doing the 6,8,10,12 to 24 led buildup, did you compensate the current for each step to make it the same per LED?
Say 120ma for 6, 200ma for 10, 240ma for 12...

If not, then I bet this is why you lost "linear brightness gain". If you did compensate for current, then the Law of diminishing returns wins.

I will do my tests by keeping the current through each led constant, and see what the light meter tells me.

If you ate pasta and antipasti, would you still be hungry? (George Carlin)
Or would there be a huge matter explosion? (Gadget)