I'm becoming a mad scientist

[evan9162]

I\'m becoming a mad scientist

So this all started when I thought I could take a Luxeon/O assembly, and extract the bare emitter from it. *cough* wrong. All I ended up doing was popping the lens and leads assembly off the alum pedistal that the Luxeon die lives on.

At that point, I figured I'd killed a blue-green Luxeon, and blown $15. Just in playing around, I touched a couple of wires directly to the die, and *flash!* blue-green light!

So it's Friday night, and I'm pretty bored. I figure I would get the most fun out of this now-unusable (but still functional) cyan luxeon.

I decided to see how much current I can push through this little beast. I also figure to use the best cooling possible: water!

So I immersed the wounded luxeon in some tap water, and grabbed a couple of needles (to make the connection)

This happened:

That's pushing 2A through it.

The water was boiling directly from the die.
It is insanely bright
Something wierd is happening though, there was some residue left floating in the water, and the luxeon die is slightly discolored. Wonder if the LED substrate is boiling away or something.

Igor! Get my electrodes!

 

[MrAl]

Re: I\'m becoming a mad scientist

Hello,

Looks like a pretty artistic coffee heater
🙂

Actually, it might be nice to see how long
it would live at a more reasonable current,
like 1 ampere. Measure the temperature
of the water and see how long it takes to
rise up to maybe 60 degrees C.

Also, was the water boiling as a result
of a temperature increase or because
conduction in the water between the two
electrodes was occurring? What kind of
water did you use, regular tap water or
distilled water?

Pretty interesting idea.

Good luck with it,
Al

 

[evan9162]

Re: I\'m becoming a mad scientist

I saw the needles form bubbles on them due to electrolysis happening. However, I think it was boiling from the die, as it made a slight fizzing sound as the bubbles formed from the LED die. However, I'm not sure if the surface temperature of the die would be high enough to cause instant boiling. Perhaps it was causing a more accellerated form of electrolysis to happen? Even after running it for a few minutes, then removing it from the water, no part of the luxeon was significantly warmer than before.

Unfortunately, I got greedy and tried to push more current through the luxeon (by running 12V across it), only to burn/crush the die. Oh well - I don't think of it as $15 lost, I just now have 3 $20 white luxeons

anyone else crazy enough to try immersing their luxeons?

-Darin

 

[**DONOTDELETE**]

Re: I\'m becoming a mad scientist

Hmmm...how about a small, waterproof flashlight head, filled up and sealed with water inside to cool the Luxeon, and then run the 2 amps across it? If your pic is any indication, that would make for one BRIGHT light!

 

[Andre]

Re: I\'m becoming a mad scientist

Has anyone tried immersing a Luxeon in liquid nitrogen yet?

I seem to remember a story about someone doing this with a discrete LED and getting a very large increase in brightness.

Or maybe just sitting one on top of a piece of dry ice (frozen CO2) would be easier.

Andre

 

[Lux Luthor]

Re: I\'m becoming a mad scientist

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by MR Bulk™:
Hmmm...how about a small, waterproof flashlight head, filled up and sealed with water inside to cool the Luxeon, and then run the 2 amps across it? If your pic is any indication, that would make for one BRIGHT light!<HR></BLOCKQUOTE>

I ran the idea of liquid cooling by Peter on another thread regarding the 5W ones. If we can read his silence the same way we can read PK's silence, he may be working on it. At least maybe?

 

[PsycoBob[Q2]]

Re: I\'m becoming a mad scientist

You could also use fluids other than water to cool it.... just keep track of the refraction, or you'll wind up with an odd light. Might just be easier to stick several LS's in the head of a D-cell maglight, and load it up with distilled water.

Computer overclocking maniacs had many of the same problems- look at their pages for tips- LS's should be easier, due to only one component. Silver heatsink plate, Arctic Silver 3 mixed with AS epoxy to stick things together.... and a ZLT circuit built with a heatsinked power MOSFET instead of a transistor.

 

[signals]

Re: I\'m becoming a mad scientist

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by PsycoBob[Q2]:
Silver heatsink plate, Arctic Silver 3 mixed with AS epoxy to stick things together....<HR></BLOCKQUOTE>

Slightly off topic:

Why do I keep seeing people mixing the Arctic Silver grease with epoxy? Is there some reason this is better than the Arctic Silver epoxy that I've been using? Or just cheaper/more versatile?

-Kevin

 

[Roy]

Re: I\'m becoming a mad scientist

How are you going to cool the water that you're using to cool the LED? The heat has to go somewhere. As a liquid heats up, it expands and if your system is sealed, then you get pressure! Enough pressure...BOOM!

 

[Lux Luthor]

Re: I\'m becoming a mad scientist

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Roy:
How are you going to cool the water that you're using to cool the LED? The heat has to go somewhere. As a liquid heats up, it expands and if your system is sealed, then you get pressure! Enough pressure...BOOM!

<HR></BLOCKQUOTE>

You don't need to fill up the flashlight head completely, you could just leave a little space in there.

As far as cooling the liquid goes, the total volume of liquid should have a large heat capacity compared to the LED (i.e. the LED is much smaller). This would spread out the thermal energy into the liquid, keeping the LED cooler. It isn't necessary to entirely cool it, just to get that peak temperature at the junction down (i.e. reduce the temperature gradients).

Yes, heat still needs to be dissipated, but that would happen in a metal flashlight by convection and radiation at the outer surface, as long as the liquid is in contact with the body.

 

[EMPOWERTORCH]

Re: I\'m becoming a mad scientist

I read somewhere (maybe it was Don Klipstein's web site) about an underwater bulb that was specifically designed to be run underwater. The bulb was fitted with a very high wattage filament, which, if the bulb was run in air at its full working wattage would cause the bulb to overheat and break or melt. Under water however, the bulb's glass envelope would be kept cool by the water.
I am sure this same technique could be extended to solid state lights in the form of an underwater torch or diving lamp. The lamp could be designed so that by means of a pressure switch, it would operate in air at normal Luxeon currents, but in water say more than a 1/2 m deep, a much larger current may be applied giving a superb light for diving. In the case of a torch, the problem now isn't a matter of heatsinking but how to design a battery power source powerful enough to operate at such high currents for the length of a typical dive without having a torch so big and heavy as to be unmanageable.
One solution is to incorporate a diving lamp assembly into a diving helmet and the battery into the webbing that carries the aqualung.
Providing the lamp assembly with a"watersink" would cool it even more efficiently...