qwertyydude
Flashlight Enthusiast
- Joined
- Aug 10, 2008
- Messages
- 1,115
Just a theoretical, if the power supply wasn't affected by temperature, is there a minimum operating temperature for led's? Say will it work while submerged in liquid nitrogen?
Led operation in extreme cold.
- Thread starter qwertyydude
- Start date
I would think theoretically they should work at any temp but super cooling stuff does mess with the speed of electrical stuff oddly.
qwertyydude
Flashlight Enthusiast
- Joined
- Aug 10, 2008
- Messages
- 1,115
Exactly my point. I know they work just fine in a really cold freezer. I left mine in there overnight with no battery. Put one in and it lit up fine. I wanted to know if there was a lower limit for just the led alone.
Are you going to put the LED in liquid nitrogen with the battery not in it? :D
qwertyydude
Flashlight Enthusiast
- Joined
- Aug 10, 2008
- Messages
- 1,115
No unfortunately I don't have access to liquid nitrogen. But I'll tell you what I plan on doing. I have an xp-g led on a star with a tiny nick on the dome. Not even visible except with a 10x loupe but it bothers me so I won't use it. So I'm going to buy some dry ice and press the star on top and let it get really cold. I wonder if it's true that they get brighter as you go colder so if you go super cold you get super bright.
It may get a little brighter but I seriously doubt it would be more than perhaps just noticeable if it does.
While its not quite cryogenic :nana: I just ran an unknown bin SSC-p7 on a star in a bath of tetrafluoroethane (canned duster :whistle:) which wikipedia sez should be -26.3C or -15.3 F
I ran the LED at 10.2A ( :caution: ) for 3 minutes, and the voltage at the star (esitmating out the losses in the 2x 6" 28 awg leads) was 4.0v at first, and crept up to 4.3v after a minute or so.
Seoul says 3C/W thermal reistance through the slug on the base, which at 43.86w (4.3v @ 10.2A) is 131.6 C + (-26.3C) means the die was at a comfy, still in spec 105.3C (max die temp is listed as 140) If you can get the heat out to the LN2, (-196C) you could probably hit the LED with 100W :sssh: EDIT: nope apparently not. Around 16A in, the Vf got too high and the ESD diode started to conduct, and made a mess of things. The ESD diode went poof, and it appears that there was signs of overheating bond wires to the die, though they were still intact, (The silicone around the bondwire appeard to have been separated and making gas as the wire got hot. ) Not sure what else failed but no more light out from this LED, even though it failed open with the bondwires appearing intact. [/edit]
But a LED in a bath of boiling liquid doesnt do much good, so on a cold plate youre not going to be quite cold as if the LED was immersed. Either way, with a refrigeration setup (something that runs colder then R-134) or DICE/acetone or LN2 you could push a LED quite hard.
I wonder if the cold temps might affect the effiency of the phosphors on top the die. Pushing them hard like this, obviously they wont stay cold for too long, but running at a normal drive current and LN2 or below temps I'd wonder if the phosphor's output would increase or decrease.
I ran the LED at 10.2A ( :caution: ) for 3 minutes, and the voltage at the star (esitmating out the losses in the 2x 6" 28 awg leads) was 4.0v at first, and crept up to 4.3v after a minute or so.
Seoul says 3C/W thermal reistance through the slug on the base, which at 43.86w (4.3v @ 10.2A) is 131.6 C + (-26.3C) means the die was at a comfy, still in spec 105.3C (max die temp is listed as 140) If you can get the heat out to the LN2, (-196C) you could probably hit the LED with 100W :sssh: EDIT: nope apparently not. Around 16A in, the Vf got too high and the ESD diode started to conduct, and made a mess of things. The ESD diode went poof, and it appears that there was signs of overheating bond wires to the die, though they were still intact, (The silicone around the bondwire appeard to have been separated and making gas as the wire got hot. ) Not sure what else failed but no more light out from this LED, even though it failed open with the bondwires appearing intact. [/edit]
But a LED in a bath of boiling liquid doesnt do much good, so on a cold plate youre not going to be quite cold as if the LED was immersed. Either way, with a refrigeration setup (something that runs colder then R-134) or DICE/acetone or LN2 you could push a LED quite hard.
I wonder if the cold temps might affect the effiency of the phosphors on top the die. Pushing them hard like this, obviously they wont stay cold for too long, but running at a normal drive current and LN2 or below temps I'd wonder if the phosphor's output would increase or decrease.
Last edited:
kaichu dento
Flashaholic
I don't know about liquid nitrogen, but this winter I'll try putting some lights in the creek after it gets really cold!
In the led emitter specs there is a temp limit usally somewhere between - 20 to 80 deg C. While the upper temp is logical, the lower limit isn't? In Sibiria it have being -80 deg C. If you needed light then did you have to use incandescent?
Optical Inferno
Enlightened
At very cold temperatures, the bandgap will change causing the LED to become more efficient if I'm not mistaken.
At extreme cold temperatures like liquid nitrogen temperatures the bandgap should become so insignificant won't the LED just short?? Or won't the forward voltage get so low that it just won't work?? I can't remember how it works...now I'm going to have to whip out the old quantum physics and solid state textbooks...
At extreme cold temperatures like liquid nitrogen temperatures the bandgap should become so insignificant won't the LED just short?? Or won't the forward voltage get so low that it just won't work?? I can't remember how it works...now I'm going to have to whip out the old quantum physics and solid state textbooks...
UberLumens
Newly Enlightened
- Joined
- Feb 24, 2010
- Messages
- 143
a semi conductor can become a superconductor at very cold temperatures.
so the LED would be ok, but maybe end up with an exponential need for amps
so the LED would be ok, but maybe end up with an exponential need for amps
AnAppleSnail
Flashlight Enthusiast
Tell you what, next time I go by Harris Teeter I'll blow a few bucks on dry ice, then drop a prototype bike light (remote power) in and see what happens. I won't have any way to verify that the LED reaches dry ice temperatures, but if I drop it in and apply power 10 minutes later it's a good bet.
mattheww50
Flashlight Enthusiast
ALL metals and semi-conductors become superconducting if you make them cold enough, but we are talking about temps below 1 Kevlin for most materials, with a handful that go up to about 10 Kelvin, and some special metalic compounds that are superconducting up to about 90 Kelvin.
The issue with temperature limits is usually a materials issue. If the LED and substrate material or mounting have significantly different coefficients of explansion, extreme cold or extreme heat could case the either the LED or the substrate (most likely the LED) to fail mechanically.
I would be surprised if a LED didn't work at liquid nitrogen temperatures, as long it didn't fracture due to mechanical stress caused the differences in coefficient of expansion.
While 'weird' stuff does indeed happen at very low temperature, very low means just that, something on the order 4 Kelvin or less. Liquid Nitrogen is 77K IIRC.
uk_caver
Flashlight Enthusiast
At low enough temperatures, would there be a chance that differential thermal contraction could damage parts of an LED, or its mounting arrangement?
Yes
You need to lower the temp slowly rather than quickly.
Using LN2 should be no issue but has to be controlled else yes the thermal stress will kill the chip.
Also there will be limits still on passable current that can pass through the micro filament connecting wires to the die.
HarryN
Flashlight Enthusiast
Just a few points of interst. The LED Vf does keep rising for quite a while as you get cold.
Some practical limits include that the silicone in the dome is only rated down to (-40C). Finding ones that are officially rated below this probably involves custom formulations.
Cree and Lumileds both offers temp specs in this range, but most LED makers do not.
As far as Siberia, yes it does get really cold there. My understanding is that mechanical / electrical items that are used in those types of regions either are never shut off in the winter, or include substantial auxiliary heating to warm them up prior to use. They even electrically heat the exhaust pipe of trucks there to keep the water vapor in the exhaust from condensing and freezing the exhaust pipe shut.
Some practical limits include that the silicone in the dome is only rated down to (-40C). Finding ones that are officially rated below this probably involves custom formulations.
Cree and Lumileds both offers temp specs in this range, but most LED makers do not.
As far as Siberia, yes it does get really cold there. My understanding is that mechanical / electrical items that are used in those types of regions either are never shut off in the winter, or include substantial auxiliary heating to warm them up prior to use. They even electrically heat the exhaust pipe of trucks there to keep the water vapor in the exhaust from condensing and freezing the exhaust pipe shut.
Most of Cree's low temp thermal and thermal shock testing bottoms at -40C. The basic rules are, power supply outside, silicon gland to the leds
The vf will increase and the led will produce more light. There is very little in lumen maintenance at these temps
The vf will increase and the led will produce more light. There is very little in lumen maintenance at these temps
You may have problems due to the one-sided nature of the cooling and the pressure and vibration inherent in pressing warm items onto dry ice. You can make a cryo-fluid using dry ice and alcohol. Isopropyl works well but I recommend using a high percentage. 99% is available in some areas and is usually cheaper and has less residuals [no I don't know why]. The reason you don't want low percentages is the same as why large chunks of dry ice slowly stop bubbling in plain water...the water freezes forming a heat shield.
And remember: DON'T TOUCH IT!!! Yes, you can touch dry ice if you're quick and careful but you WILL hurt yourself if you touch a liquid at the same temp. [and I know there are people saying duh to this but people like me are too curious for our own good sometimes 8o)
HarryN
Flashlight Enthusiast
This is deceptively true. We are so used to liquids being frozen before they get really cold that when you encounter a liquid that is around 0 F or less it is really dangerous. I was goofing around one time and put a finger in a silicone oil that was very cold - it numbs you so fast you don't realize what is going on, and frost bite can occur very fast.
