Homemade 50 LED light, ideas and help wanted.

[Jayls5]

For fun I purchased 50 LED lights from deal extreme. I figured I'd throw them all together to make a decently bright light.

I took a picture just to show the size/amount of LED's i'm dealing with:

Obviously, it's squished together in that picture, and i'd probably need to expand it out 1.5-2x the surface area to wire/aim it appropriately.

When I actually went to test one of the lights, here's what I came up with (camera flash on):

It appears to need ~.18 amps at ~4.5v to get maximum output. Anything higher than 4.5 doesn't really improve the brightness.

Here's another pic without camera flash:

So it seems I'm dealing with 4.5v * .18a * 50 = 40.5 watts as an estimate. Is this correct? Does anyone have an idea of what I should power it with? If possible, for simplicity, I'd like something that would allow me to wire it all in parallel.

 

[Norm]

This thread will probably interest you.
5mm & flux LED fade tests with light meter
I wouldn't be running a 5mm LED at 180mA you probably need to keep the current down around 25mA.
Norm

 

[moviles]

180 ma and 0.8w ,maybe too much for this leds without heatsink ...

 

[Jayls5]

This thread will probably interest you.
5mm & flux LED fade tests with light meter
I wouldn't be running a 5mm LED at 180mA you probably need to keep the current down around 25mA.
Norm

These lights would be too dim to make an effective light if I drove it like that. Let's assume I'm properly heat-sinking it. If heat-sinked, would I still have to worry about them quickly going down the tube?

Since I don't really know what brand my lights are, I'm not sure what to make of the link you gave me.

I have these: http://www.dealextreme.com/details.dx/sku.1107


I want to keep it simple, and I don't care about losing a 9 cent LED, so I think the best mode of action would be to do as best of a job on a voltage regulator as I can. If some break, I simply remove them and replace (or not, the others will still run). Since i'm dealing with so many of them (which probably vary), I don't see how making a good circuit limiting circuit for them is going to be feasible... or practical for that matter. Correct me if I'm wrong here.

 

[Locoboy5150]

Just in case you need some inspiration...

http://www.youtube.com/watch?v=NkkU0UO3sek

 

[John_Galt]

You can't really heatsink these 5mm LED's. They are made of optical plastic which doesn't transfer heat well. They are, in fact, heatsinked through the legs.

You would be able to make a much more effective light with regular LED's, such as the Cree line-up, Phillips-LumiLED's, Luxeon, SSC, etc. Just sayin.

 

[Norm]

Since I don't really know what brand my lights are, I'm not sure what to make of the link you gave me.

Just trying to show you how 5mm LEDs deteriorate when even slightly overdriven.

 

[Jayls5]

You can't really heatsink these 5mm LED's. They are made of optical plastic which doesn't transfer heat well. They are, in fact, heatsinked through the legs.

You would be able to make a much more effective light with regular LED's, such as the Cree line-up, Phillips-LumiLED's, Luxeon, SSC, etc. Just sayin.

And how much would that cost?

As of right now, it's not looking too expensive for what I'm doing.

I'm broke at the moment.

Just in case you need some inspiration...

http://www.youtube.com/watch?v=NkkU0UO3sek

That is hilarious. Seems he's driving them with a lot less current though. Perhaps I should attach a peltier unit to the leads of these LEDS to keep them cool?

 

[spencer]

I would say don't waste your money on this. Especially not with a peltier. Go get a high power LED. I would say the lifetime on a 5mm LED driven at 180 mA would be around half an hour to less than 50% output. Not worth it.

 

[Jayls5]

Just trying to show you how 5mm LEDs deteriorate when even slightly overdriven.

Well that's depressing. The whole reason I was even interested in this particular project was the fact that just a single LED got extremely bright with what I was pushing through it.

 

[Jayls5]

I would say don't waste your money on this. Especially not with a peltier. Go get a high power LED. I would say the lifetime on a 5mm LED driven at 180 mA would be around half an hour to less than 50% output. Not worth it.

I'll tell ya what. I'll run it for half an hour right now on my unit and I'll tell you how it does

I'll take a before and after pic.

 

[John_Galt]

Take a look at Cutter Electronics. they will have driver's, LED's, Optics/Reflectors, heatsinks, thermal paste (to ensure good heat transfer between the LED and the heatsink), etc. All at a reasonable cost.

You will also be able to purchase certain bins and tints of the various LED's. Take a look in our LED sub-forum to comprehend it all. This way your light will have a nice tint/color, and not just be blue at any distance over 3inches.

 

[John_Galt]

I'll tell ya what. I'll run it for half an hour right now on my unit and I'll tell you how it does

I'll take a before and after pic.

Unless you can manually adjust your camera, and keep the settings exactly the same, your camera will likely readjust it's sensitivity, rendering before/after photos pointless.

 

[Jayls5]

Unless you can manually adjust your camera, and keep the settings exactly the same, your camera will likely readjust it's sensitivity, rendering before/after photos pointless.

I put it in manual mode and made specific settings on things like shutter time. I'm keeping the distance the same as well. The sun is down and the lighting is the same in my room. Gotta keep it quasi-scientific. :thumbsup:

 

[Jayls5]

Unless you can manually adjust your camera, and keep the settings exactly the same, your camera will likely readjust it's sensitivity, rendering before/after photos pointless.

One hour between pics of overdriving, camera angle wasn't absolutely perfect but you get the idea:

 

[John_Galt]

A beam shot at a wall (let us know the distance) would be better. LED light sources have very high surface brightnesses (brightnessi

) so even if they are dim they appear bright.

 

[Jayls5]

A beam shot at a wall (let us know the distance) would be better. LED light sources have very high surface brightnesses (brightnessi

) so even if they are dim they appear bright.

Yeah, that's why I picked a -2 exposure to dim the appearance a bit.

I'm still running it right now and it appears exactly the same as when I started it 2.5 hours ago.

Should I grab a new LED for the wall test you suggested, or just keep using the same one?

 

[John_Galt]

A new LED, to keep things "quasi-scientific". Run one for an hour, taking a beamshot at the beginning, and one at the one-hour mark.

 

[Jayls5]

A new LED, to keep things "quasi-scientific". Run one for an hour, taking a beamshot at the beginning, and one at the one-hour mark.

Will do... I'll post when done.

 

[Jayls5]

A new LED, to keep things "quasi-scientific". Run one for an hour, taking a beamshot at the beginning, and one at the one-hour mark.

Ok, turns out the one I was using previously was already dimmed from messing with it the night before.

This one definitely dimmed down after the hour or so of use.

It's not really impressive anymore if I have to drive these at a low current. Oh well, thanks for saving me the time/effort.