RMSK, Inc.        
Results 1 to 12 of 12

Thread: 50w LED headlight experiment

  1. #1

    Shrug 50w LED headlight experiment

    Hey guys. I'm new here and quite new to LED lighting. I've been searching around the forums and Google for a lot of the answers to my questions and some of which I found the answers to. But I'm kind of looking for help on a specific setup. I'm kind of fearing that I may have bought the wrong parts that are now already on the way here. Figures... I'm an impulsive shopper.

    OK, so I'm building a 50w LED headlight system. At the moment, it will be going into a C5 Corvette for testing. C5 Corvettes have flip-up headlights, so space can be both good and bad considering there's a bit of room set aside for lighting alone.

    My original plan was to use a single 50w LED for each low beam. (I'm not worrying about the high beams at the moment.) Those LEDs would be powered by a 100w DC/DC converter which I now realize is likely not enough at all. The LEDs have a working voltage range of 18-22V. The DC/DC converter is programmable (via USB) to any voltage at the 0.01v precision. My plan was to power them at 20v.

    I'm also trying to figure out how I'm going to focus these lights. My basic idea is to create a reflector in front of the LED that uses the OEM housing to put out light. This would be similar to a reflecting telescope. My other thought was to use a custom arrangement of mirrors. And a final though to use a lens. (But the lens would have to be DOT approved.) Either way, the light has to be cut correctly.

    Finally, I've been trying to figure out cooling for the lights. I was thinking I'd be able to get away with a simple P4-style heatsink and a 12v blower fan. But I'm getting the feeling that won't be enough.

    So here's my thoughts at this point:
    Sell the 50w LEDs I've already purchased and buy 50w LEDs meant for 24-26v operating range and purchase something like this: http://www.current-logic.com/shop/in...&products_id=9

    I'd assume with a setup like that, it should have enough power and be far easier to implement.

    As for cooling, I was thinking about purchasing 2 30x30 Peltier coolers with a Qmax 50w rating. Although, I think Pelteirs defeat the purpose of LEDs (which is to be more efficient). Is this correct? Then behind the pelteirs I'd put standard p4-style heat sinks and a small 12v push fan.

    I'd appreciate any input you guys may have.


  2. #2
    Retired Administrator Norm's Avatar
    Join Date
    Jun 2006

    Default Re: 50w LED headlight experiment

    Moved to Transportation Lighting - Norm

  3. #3
    Join Date
    May 2010

    Default Re: 50w LED headlight experiment

    It won't be street legal and would be for offroad use only.

    Problem with the LEDs is that the voltage range means little. You need to watch current. This means, regardless of the voltage, you need to keep an eye on current and math out the actual wattage.

  4. #4

    Default Re: 50w LED headlight experiment

    I think a single LED brute force approach isn't the best way to go about this. Most of these LEDs are rated to 4000lm, which is most likely optimistic. Large LEDs like this have few choices in usable optics. A better approach would be to use a few smaller, more efficient LEDs that have better lens options. At least that way you can use a combination of narrow conical lenses and ellipticals to tweak the beam pattern. Using current generation quality LEDs like Cree, Luxeon, or Luminus that have a lower forward voltage (more suitable for automotive applications) will be more efficient than that big lump of a Chinese LED. With LEDs like the Cree XM-L, you can get 4000lm or more for less wattage than that 50W LED.

  5. #5

    Default Re: 50w LED headlight experiment

    Quote Originally Posted by deadrx7conv View Post
    Problem with the LEDs is that the voltage range means little. You need to watch current. This means, regardless of the voltage, you need to keep an eye on current and math out the actual wattage.
    YES.The actual wattage is very important, you should keep in mind.
     The more we do, the more we can do; the more busy we are, the more leisure we have.

  6. #6
    Flashaholic* LukeA's Avatar
    Join Date
    Jun 2007
    near Pittsburgh

    Default Re: 50w LED headlight experiment

    50W per side? With modern LEDs? That's waaay too much light. Think about a quarter to a third of that amount of power passing through each headlamp, of you're building a pair.

    A few years ago I went on a similar quest to the one you're on now. Here's my thread. I posted in a thread in the bike lighting section with some basic forays into an on-road-type beam pattern. That's actually much harder than making enough light.

    My light, even with LEDs from 3+ years ago and very inefficient drivers, drew ~50W per headlamp on full beam, and was brighter than a pair of OE 65W halogen high beams.

    An idea that my design incorporated that you may find interesting to ponder was forced-air cooling. I have a large fin heatsink that I can mount to the back of the aluminum the LEDs are mounted to, and can easily duct a sufficient mass flow rate of air through those fins to cool even inefficient LEDs in very hot conditions. Imagine a racecar's brake cooling ducts, which use forced air from the moving car. Same thing.This is to address your concerns about active cooling. I and most others, will agree with you that using power to cool something that is extremely practical to cool passively is pretty silly.
    Last edited by LukeA; 07-27-2011 at 07:37 PM.
    A little madness never hurt anybody.

  7. #7

    Default Re: 50w LED headlight experiment

    Powering this beast is only one problem. Negating the heat is another. A 50 watt Chinese LED requires approx 500square inches of passive convection area to cool. Peltiers wont help because you need to cool the peltier.

    Not to mention these Chinese LED are horribly inefficient and you easily get the same amount of lumens per half the wattage of Cree's, or 2/3 the wattage of Bridgelux.

    Last, you'd need a huge diffuser for the 50watt LED because it will blind oncoming drivers worse than any HID - provided you can actually figure out how to power and cool it, which you wont.

    To put practicality in perspective it would only take about 500-600 lumens of 4100k LEDs at a 80ish CRI to start to surpass a high performance 55watt halogen given the fixture inefficiency and scotopic limitations of the halogen. IMHO, a dead neutral quality 4100k LED source with proper diffusion that doesn't blind oncoming drivers would likely draw more attention to the vehicle and say 'class' than an obnoxious light source.
    Last edited by blasterman; 08-12-2011 at 09:42 PM.

  8. #8

    Default Re: 50w LED headlight experiment

    Well, assuming this project will never leave your garage, I'd be fairly concerned with cooling. Are you intending for the entire light assembly to pivot as with the originals? You're going to need active cooling, either a big heat sink with a fan, or a water cooling set-up, and it's going to have to flex...

    Also, it's been mentioned, but you should really use a current regulated driver vs voltage regulated, since LEDs resistance changes as the temperature changes. As they heat up, they'll pull more current. You need something that will lower the voltage accordingly.

    50W LEDs are probably what you should be using, keeping in mind that the lumen figures are based on the emitter alone, and possibly over-stated. I wouldn't run them at rated power either, just below, as it will be more efficient. The losses are going to be 1/3 to 1/2 of your total light, so you'll probably end up with as much light out the front as a decent HID setup.

    Don't try this on the road. Unless you own the road, and there's nobody else on it.

  9. #9

    Default Re: 50w LED headlight experiment

    this is what it takes audi to build led headlights
    Last edited by alpg88; 04-05-2012 at 10:02 PM.

  10. #10

    Default Re: 50w LED headlight experiment

    I've toyed with the idea of putting a Bridgelux LED behind a projector, just replacing the HID bulb and spherical reflector with a flat LED should yield favorable results, but I wouldn't use it on the road. No matter how good it looks, it's not worth getting in trouble over.

  11. #11

    Default Re: 50w LED headlight experiment

    The lighting modifications you're asking about (home-made headlamps) are illegal. Rule 11 of this board prohibits advocating illegal activity. Please stop now.

  12. #12
    Flashaholic* AnAppleSnail's Avatar
    Join Date
    Aug 2009
    South Hill, VA

    Default Re: 50w LED headlight experiment

    If they're show-only lights, they'd be okay to plan here. But a 50W LED array is too big for headlight use. That would blind people, not look like headlights.

    PS: Replacing the stock headlights with non-road-legal lights makes the whole car non-road-legal. I can't tell you if it would only be illegal at night; I'd guess unsafe lighting is unsafe 24/7.
    My biggest light-hog is my camera.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts