LEDs were once little more than weedy little indicators with low light output. Today, they’re absolute powerhouses, efficiently turning a flow of electrons into a searing beam of light. Despite their efficiency, they can still put out a fair whack of heat. Thus, if you’re building a powerful flashlight like [CrazyScience], you might wanna throw some active cooling on there just to keep things happy. Check out the video below.
The build will not be unfamiliar to any casual observer of the modern DIY flashlight scene. It uses a flatpack LED module of great brightness and a wad of 18650 lithium-ion cells to provide the juice to run it. The LED itself is mounted in a 3D-printed frame, which leaves its rear exposed, and a small PC fan is mounted for air cooling. It’s not the most optimized design, as airflow out of the fan is somewhat restricted by the 3D-printed housing, but it’s a lot better than simple passive cooling. It allows the torch to be more compact without requiring a huge heatsink to keep the LED at an acceptable temperature.
The final torch doesn’t have the most ergonomic form factor, but it does work. However, as a learning project for a new maker, it’s a start, and the learning value of building something functional can’t be understated. If your desire for flashlights swerves to the more powerful, we’ve covered those, too. Just be careful out there.
I’ve dealt a lot with these types of led’s and this is a horrible idea for nothing more than clickbait channel. These led’s on air alone last at most a month and it is far more common for them to fail much sooner than that with only air. at a minimum there needs to be a heatsink on them. Again horrible execution, horrible idea to tell people to do just horrible all the way around.
“The build will not be unfamiliar to any casual observer of the modern DIY flashlight scene”
Why do I feel that the author here themselves is not a part of that scene? Cause this is a horrible example of how to do it.
I thought hackaday was beyond the hot glue youtube diy clickbait stuff (you guys know the type if you’ve spent some time over there).
To do it right you actually use a heatsink. There’s not enough surface area on the back of one of those chips to air cool it directly. You need a hewtsink of some type. I’ve personally done this with a 100W COB module and an old gpu heatsink; works great. With an led like that at full power, your dumping almost half of its supply power into heat. It’s gotta go somewhere.
Yeah, and look at how ugly that thing is. It’s literally useless as a flashlight (as in you would never carry it around for portable lighting), will fail almost immediately, and doesn’t even solve some niche problem.
So sick of YouTube clickbait trash. There are projects / hacks that are being done earnestly rather than just “content for content’s sake”.
I sure would appreciate the editors here taking the time to filter out things like this – it might result in less articles for a while, but surely the authors here are capable of finding more interesting things if given the motivation to do so
It’s stuff like this that makes me want to write a browser plug-in to filter out specificly low quality youtube repost hackaday posts… I suppose that would be easy to do with the rss feed.
Honestly what really happens is I just read the site less.
Hello Nathan! Do that please, and send it to me! I need it!
Yes, please!
I’ve seen those 5W LEDs work just fine in open air without heatsinking
My 20W flashlight with a 20W LED, now that required a heatsink
I’ve never had one of these no name COB LEDs last more than a few hundred hours even with a proper heatsink. The current is never balanced between the LED chips, so one string burns out quickly, then the rest get overdriven and burn out too.
I have a theory that any cob LED being sold in less than 1K MoQ is from a factory reject batch, because that is the only way I can rationalise the literal volts worth of difference in forward voltage between units, that is to say nothing about many LED strings not working entirely until you poke the yellow gel thing with your finger.
Wild!
The LED will die within no time.
There is no current limitation happening so it will overheat quickly despite the poor active cooling.
The LED is right at the center hub of the fan where no fins are and therefore the airflow will be rather limited.
Poor execution and an even sadder to see it posted here in the Hackaday feed, I must say.
A “five minute craft” type of video wrapped as an instructables. The LED will burn out very soon.
when the comments are 1000x better than the post…
A nonsense!
Why didn’t you use cardboard and hot glue for the entire “project”?
You made a useless print where neither the LED nor the fan are properly fixed, and you used hot glue to secure them.
What you did is not useful at all.
It doesn’t have a closed and functional casing for proper use.
The LED doesn’t have a current driver, and there is no BMS for the batteries.