[Holzleim’s] flashlight build is quite impressive. And the rise and fall of the hardware is quite a story. He designed it as a handheld light, relying on batteries to power a multitude of high-power LEDs inside.From the collection of four 5350 mAh lithium polymer cells he was able to achieve a peak power output that surpasses 500 Watts!
Projects that make use of these powerful LED modules have heat dissipation as a major design element. With this in mind, [Holzleim] used large bars of copper as a mounting surface for each group of modules. The size and thickness of these bars made it difficult to use traditional soldering techniques to affix the bodies of the modules so he ended up using a clothes iron to ramp up the temperature high enough to reflow the solder. Once mounted, most of the LEDs were paired with optics, including that Fresnel lens at the center, to help focus the light.
To keep the batteries topped off, [Holzleim] designed his own external charger which can run off of mains or from a car’s electrical system. Unfortunately he must not have included a way to monitor the battery temperature because one of the cells failed, causing a fire that burnt out the components inside the flashlight. Luckily his wife was home at the time and got the failing device outside where it didn’t cause more damage. We’ve linked to his posts regarding the charging system, the properly functioning light burning paper with its beam intensity, and his sad announcement about the fire after the break.
Man, that’s a lot of money to go setting on fire.
Wow, such a nice build, and it was totally torched. That sucks!
it was literally painful looking at that pic of it all crispy
LiPo is an inherently combustible technology.
The Original Hack is awe inspiring for the level of effort and final resultant beams. I will def consider using a clothes iron for reflows. That alone justifies the post, Even if it’s not a new idea to everyone- it was to me.
Oh, I know “why” we build with LiPo and often omit containment structures robust enough for zero damage upon failures. It’s in this case- a proof-of-concept trade off to make it buildable?
Making 100% zero damages survivability of a worst-case battery “event” possible is often not practical. But making no provision for failures is arguably near suicidal or even negligently HOMICIDAL. If it burns down an inhabited structure or causes a death- criminal charges are a very thinkable result. Though law is a poor consolation for lives being lost or damaged if we have an avoidable incident.
Be Careful, but not Too Careful- is a pretty phrase for many situations. And often it’s “Good Enough” as a guideline.
I at one time charged some cell phones etc in a crock pot type bowl. usually in laundry sink etc so any fire might be minimally destructive to anything beyond the dead device itself. I don’t feel so overcautious every time I read of an incident.
Sadly- all of us- myself included have our perceptions hacked by simple lazy- my phone’s charging on the table next to me. It’s going to shortly be living in one of the mini crock pots again.
And _I’m_ unashamedly trying to “Socially Hack” the mindfulness of EVERYONE HERE about not killing anyone with our weird/wonderful creations. Bystanders or our selves.
Darwin’s also our toughest QC inspector..
You (and the HaD staffer who wrote the article) might want to read the thread, especially page five before going off on a lecture.
The chap did build in numerous safety devices, including temperature cut offs.
LiPos are great, but I think I’ll stick to NiMH for my projects.
Lithium phosphates are better, the energy density is about 70% of Li-Ion and they can’t explode, melt down or otherwise ruin your entire day.
Plus when you overdischarge these you can recover them by careful heating of the pack to 65 degrees C then charging, to get it over the “energy hill” which normally prevents recovery above 2.1V
See 4HV.org for more details.
Anyone have an estimated cost?
I love his improvised stand for the iron.
I agree with one of the responders to his thread. Is it possible to have the battery disconnected from the device while charging? (safely away from the house in some fire resistant container?)
You have to wonder whether this was one of those not-so-reputable cell manufacturers which use inferior materials in the plates.
How did I know that CPF would be involved with this. It’s not the only impressive build I seen from there.
Glad it made it,im on my way to building foglights out of 4 or 8 20W leds… =D
@buzzles:
My comments were generic to all high energy density systems. And- all the safety devices one can design still won’t save you from a failed cell.
Yes- good safety practices of the charge and discharge level control parameters can lower the failure risks. But- that failed cell is just as lethal to be around regardless of “how” the failure was caused. That point of a cell failing is where electronic design becomes superseded by mechanical design.
So- containment becomes the most robust overall “Last Defensive Strategy” element. If a cell fails- are you willing to bet your life and the life of others on your design skills??
Call it designing for Failure Risk Reduction?
My Hack under current experimentation is corrugated Stainless gas tubing for cell containment. It’s potentially fire resistant enough to make a difference. And- the pressure management of that corrugated expanding to dissipate pressure energy may be a worthy design Hack too.
Complete the design exercise by finding a thermal insulation barrier that’s not a fire load risk or prohibited by any other logic..
nice project! at the very least, this will surely inspire numerous “reflow iron” projects. i know i’m gonna make one as soon as i find a six pack of karlskrone. ;)
the fire is a real bummer, but i can’t help but think the charger may have been at fault.(page 4) the light was such a clean and professional build, but the charger looked like a rats nest just waiting to short out and cause havoc.
While it may be inconvenient, consideration should be given to keeping high power battery power supplies as this separate from the device they power, or at least removable. I think we can never know the number of builders who have interdependently came up with using a clothes iron in this or similar fashion. This is the first high power LED light builder that I read that understands that a flood light option in addition to the spot light mode. I have a 31 LED 3 cell Maglite wannabe. While it works well as far what in can do goes, it pales in versatility when compared to my 2 cell maglite with an adjustable beam pattern.
No battery management circuit hardwired to the cells is what let this happen. The cells were just paralleled up. If one cell in the pack get over or undercharged a fire can result. A battery management circuit would have shut down the pack and disallowed charging or discharge.
Perhaps the concept of failure modes needs some clarification. LiPo can fail from violation of charge or discharge profiles. Those become theoretically- at least-reducible risk factors by cell level management circuits. And if you’re an OEM able to do exhaustive real world destructive proving out tests, those risks become closer to actuary math models. Which can be insured against by a megacorp, but can we mere Hackers risk becoming the legal test case for failing due diligence?
If you’re a Hobby Hacker- *YOU* are the certifying authority over YOUR design+build+QC overall. It’s the reason I am evaluating methods like the corrugated Stainless for risk reduction. Perhaps using a small engine spark plug as feed thru pressure bulkhead? The potentially over looked risk trading is going from a flamethrower to a pipe bomb in one THUMP. Be wary that a mitigation attempt does not create a bigger risk.
I’m VERY respectful that to make magic of technology entails risks. “It Is the Business of the Future to Be Dangerous” so to speak, but it’s up to us as guardians of HOW dangerous.
Thus- go ahead and make magic! Dance with no safety net while doing so. Just be damned mindful of what’s at stake for you and bystanders.
http://www.candlepowerforums.com/vb/showthread.php?221261-Handheld-500W-LED-Power-….-finally-finished-(Beamshots-added)&p=2946501&viewfull=1#post2946501
Sorry.
No sympathy for this guy.
He spends all that time building a moderately well built flashlight, then totally cheaps out on the charger and produces that *thing*?
Just look at it! Does that look safe to anyone?
No wonder the lipo cells failed in the charger.
-SC
I hate to also be negative, but I have to agree with SingChao. I wouldn’t let volatile batteries anywhere near that board. I’m not a PCB snob and have hand cut copper-clad many times, but that charger looked like a train wreck even before the fire. The batteries flamed during charging. Looking at that project, the fault is far likely to be the charger than the cells.
ha, actually built somethin similar with a phenomx4 heatsink, an’ 4 xml-u3 emitters, driven @ 3500ma off a 12v, 20ah SLA agm. used reflecters an’ hand magnifiiers put on with hot glue t’ focus the light. wouldn’ quite burn paper, but it’d make a bright hotsopt s’ uncomfertable t’ shine on yer skin fer more’n a second er two.