Page 1 of 2 12 LastLast
Results 1 to 10 of 20

Thread: Single piece headlight assembly with LED projection

  1. #1
    Join Date
    Oct 2004
    Location
    Beaverton, OR
    Posts
    2,842
    Feedback Score
    1 (100%)

    Default Single piece headlight assembly with LED projection

    So this will be the thread that I will use to design and fabricate a single-piece headlight assembly for the 1987/1988 Thunderbirds.

    Purpose of this build:

    1. Single piece headlight assembly with no seams between
    2. Use a completely clear lens. Ditch the sloppy light control and losses that fluted lenses create.
    3. Assembly must bolt up to the stock header panel holes and NOT require cutting the header panel
    4. Use one or more projectors for the headlight
    5. Utilize LEDs only, including the headlights
    6. Keep high beam, likely using a HID projector with solenoid
    5. Provide adequate heatsinking for LEDs that are installed.

    What this build will NOT incorporate:

    1. A way to "replace" bulbs, since I will be using LEDs, they should last forever. Any replacement would require dis-assembly and removal of dead LED, then installation of new LED. After all, by the time an LED fails, there should be much better ones out there to use at that time.
    2. The initial build will not be made from a mold that can be cast with resin in the future. If a housing and lens assembly can be found to work, I MAY make a mold for other people to buy castings of, then design their own internal headlight designs. I will not make multiple of these headlights to sell due to the complexity that the first attempt to have.

    Foreseen issues to resolve:

    1. How to aim the projector assembly? This includes rotation, vertical, and horizontal.
    2. How to form the lens over such a wide headlamp assembly and get the sides to wrap around enough to provide a good seal all around. This may just work, but it will take some practice with scrap sheets first.
    3. What can be done to make the entire headlight assembly look filled out? We have one large headlamp, and two fairly big marker lights. The huge amount of surface to do things with so that the headlights look good will be difficult to design. I'm not that creative.
    4. What kind of heat will I experience if all heatsink fins are contained within the headlight assembly? This provides thawing of the lenses, but it is also an enclosed environment that can heat up when pushing 40-50 Watts of heat into the housing, and the output diminishes as the LEDs heat up. This will just need some test results with a thermometer.

    Any and all ideas and aide will be appreciated. On that topic, anyone have any busted TC inner marker lights? I have a spare set of outer markers, headlight aiming brackets, and housings. I also have a headerpanel to mock this all up on. I also have plenty of mediocre quality (3M) fiberglass from using one of those $50 off $100 home depot coupons. Either way, I think most of the housing will be made out of aluminum rather than any sort of resin.

  2. #2
    Join Date
    Oct 2004
    Location
    Beaverton, OR
    Posts
    2,842
    Feedback Score
    1 (100%)

    Default

    So start, I put 4 emitters next to one another, each about an inch away with the two edge ones aimed inward, the front half of a projector housing, and took a picture of what the light distribution looks like on the ceiling. The left and right sides of the beam were cutoff by the desk lamps, one of which you can see in the lower right.

    Moving the LEDs closer together will provide much more even distribution of light. This proves to me that LEDs WILL work. The front half of the projector was held 1 inch above the LEDs. There is a lot of light loss as I have no reflector currently to keep the LED's light going through the projector's lens. I would be surprised if even 10% of the light is making it through the small projector opening above the cutoff and in between everything else. Now the question is how do I capture this light without keeping the full depth of a projector. The emitters can't be moved forward since the lens is focused on the point where the cutoff shield resides. I will have to play with some things, and aluminum foil, to see what I can do.

    If I could get more distance, I'd expect the cutoff to shift from yellow to blue.


  3. #3
    Join Date
    Feb 2012
    Location
    Grand Rapids, MI
    Posts
    1,808
    Feedback Score
    7 (100%)

    Default

    I've been tinkering with this a little, and I've determined that my LED headlamps will NOT be projector based. The reason is that my array has to be so close to the shield that I can't get a wide enough pattern from them. I'm using a Bridgelux BXRA-C2002-00E00 with an old Core2Duo aluminum/copper HS. I've decided to investigate reflector optics, but I've yet determine how to produce an upper cutoff. I've also thought about a inverted chrome cone to spread the output out into a wider reflector. I'm pretty sure Audi's LED headlights are reflector based. I'd really like a good close look at one...

    I know Seek likes the multiple LED idea, but that C2002 makes 3000 lumens at 17.5v. Just sayin.
    1988 Thunderbird 5.0
    GT40P, TFS-1, KB domes, 30#, MAF conversion, Explorer GT40 Upper/Lower, Smog delete, Wide Ratio AOD Mod, Unlocked Speedo, 3G Charging System Upgrade, Hi-Torque Mini Starter, 3.73 LSD, BBK 1 5/8 unequal headers, H-Pipe.

    88 Thunderbird/Cougar EVTM
    (Thanks to Trinom for hosting)

  4. #4
    Join Date
    Oct 2004
    Location
    Beaverton, OR
    Posts
    2,842
    Feedback Score
    1 (100%)

    Default

    Quote Originally Posted by TheFoeYouKnow View Post
    I've been tinkering with this a little, and I've determined that my LED headlamps will NOT be projector based. The reason is that my array has to be so close to the shield that I can't get a wide enough pattern from them. I'm using a Bridgelux BXRA-C2002-00E00 with an old Core2Duo aluminum/copper HS. I've decided to investigate reflector optics, but I've yet determine how to produce an upper cutoff. I've also thought about a inverted chrome cone to spread the output out into a wider reflector. I'm pretty sure Audi's LED headlights are reflector based. I'd really like a good close look at one...

    I know Seek likes the multiple LED idea, but that C2002 makes 3000 lumens at 17.5v. Just sayin.
    I was thinking of trying a large emitter, only due to the evenness of light and simpler point source. I will have to play with some things a bit more though.

    As for the simplest way to get a cutoff - it doesn't look pretty, but you put emitters side by side, create a "visor" made of a reflective material, and you'll get your cutoff. Basically you have your LEDs visible, and it doesn't look pretty like a projector, but it works. You can even make a very small change in the surface to get a step like projectors have. You still get good width due to how wide some emitters throw light, or you can make use of the side markers to get some extra light on the side, like many manufacturers currently do when implementing LEDs. I started doing this with some aluminum stock for a bicycle light, but the alumiweld cut a hole directly through my metal......I'm currently picking out a hobby-level TIG welder. I think a Dynasty 165 will do well for the time being, although everyone seems to love the Syncowave 200/250 units MUCH more and the ability to do thicker aluminum would be nice at some point I'm sure.

  5. #5
    Join Date
    Feb 2012
    Location
    Grand Rapids, MI
    Posts
    1,808
    Feedback Score
    7 (100%)

    Default

    What about a collimator? A 60mm collimator inside the projector bowl might get me a better dispersal and maybe work out to a better use of the projector. POL makes these for bridgelux arrays. I'm still on the lookout for a good driver, too. Have you worked with these?
    1988 Thunderbird 5.0
    GT40P, TFS-1, KB domes, 30#, MAF conversion, Explorer GT40 Upper/Lower, Smog delete, Wide Ratio AOD Mod, Unlocked Speedo, 3G Charging System Upgrade, Hi-Torque Mini Starter, 3.73 LSD, BBK 1 5/8 unequal headers, H-Pipe.

    88 Thunderbird/Cougar EVTM
    (Thanks to Trinom for hosting)

  6. #6
    Join Date
    Oct 2004
    Location
    Beaverton, OR
    Posts
    2,842
    Feedback Score
    1 (100%)

    Default

    Quote Originally Posted by TheFoeYouKnow View Post
    What about a collimator? A 60mm collimator inside the projector bowl might get me a better dispersal and maybe work out to a better use of the projector. POL makes these for bridgelux arrays. I'm still on the lookout for a good driver, too. Have you worked with these?
    I have two of the H6FLEX drivers, but not the B6.

    The problem is I want to find a way to distribute the light WITHOUT using the reflector portion of a projector. We don't have much depth to work with, if I want to keep the headlight assembly entirely in front of the header panel. While I'm fine with chopping them up to fit more in there, I have a lot more options and less worries IF I can get the light source to fit in shallow enough of an area. I'm not sure this is possible though, a there are no projectors with an extremely short focal length for the lens. If there was, the lens would probably be so tall that it would make up for any other space savings to be found with the shorter focal length...I think an array of LEDs to create a wall of light just behind the cutoff shield is the only way to effectively get a good beam without needing another 2-3" behind the lens mount.

    I should add that my comment earlier about using a "visor" to create a cutoff would also create a ton of foreground light if no other optics are used, not the light that you need up higher to fill in further distances. The useful light is up higher, which you can see with any car headlight where most of the light is near the "cutoff" point.

    I should have some time tomorrow to play with some optics. I also have a TSX projector coming to play with since it has about the shortest focal length available for projectors. Of course it has no high beam but I will work with what I can. I will also review what people have done lately with LEDs and their beam quality as it may help give me some ideas. I would love to get something put together for some beam shots against a wall though, then play from there. Holding a projector up makes it difficult to get pictures.

    I also had to put my tail light project on hold last night as I need some more screws and insulating washers before the LEDs can be mounted and everything put together. All the machining is done though and ready for bolting everything together.

  7. #7
    Join Date
    Feb 2012
    Location
    Grand Rapids, MI
    Posts
    1,808
    Feedback Score
    7 (100%)

    Default

    If the collimator does it's job, I shouldn't need to screw with focal depth. With that big array, I need a boost driver. The H6Flex is a buck driver. Are they any good, the ones you're using?
    1988 Thunderbird 5.0
    GT40P, TFS-1, KB domes, 30#, MAF conversion, Explorer GT40 Upper/Lower, Smog delete, Wide Ratio AOD Mod, Unlocked Speedo, 3G Charging System Upgrade, Hi-Torque Mini Starter, 3.73 LSD, BBK 1 5/8 unequal headers, H-Pipe.

    88 Thunderbird/Cougar EVTM
    (Thanks to Trinom for hosting)

  8. #8
    Join Date
    Oct 2004
    Location
    Beaverton, OR
    Posts
    2,842
    Feedback Score
    1 (100%)

    Default

    Quote Originally Posted by TheFoeYouKnow View Post
    If the collimator does it's job, I shouldn't need to screw with focal depth. With that big array, I need a boost driver. The H6Flex is a buck driver. Are they any good, the ones you're using?
    But the collimator has quite a bit of its own depth (1.5"). I'm not sure there would even be a benefit of using one on a larger LED die that provides even light, as it would likely destroy the beam width that you could gain a bit from by using a reflector instead. Again, it needs testing through, and a reflector to help get the light through the 1/2" opening in the projector.

    I don't have any long-term testing, but the general consensus is that the h6 drivers haven't failed in any of the products I've seen them used in. Heatsinking the things are very easy, so high loads don't concern the driver in any way. I'll have to look again, but if I remember right, their boost driver was more efficient than their buck driver?

    Their driver offerings have been expanding lately. I believe the boost driver is relatively new?

  9. #9
    Join Date
    Oct 2004
    Location
    Beaverton, OR
    Posts
    2,842
    Feedback Score
    1 (100%)

    Default

    I'm not sure the C2002 is very usable as a car headlight The 5500-7000K color temperature (the higher luminous rated LED will likely be of higher color temperature) would really work against you, as would the 65 CRI rating. The N1203 is much better suited but only good to 2300 lumens emitted at max rated current. I'd expect greater losses in just about anything you can come up with to make the LED work within a projector, than if you were using a D2S bulb. If you use multiple though, they could work out quite well, and not cause as much glare to others on the road due to the light being spread out, rather than a ton of light coming from a single point.

    There are also Cree emitters on 10mm metal core pcbs which would allow four to fit into the area of one bridgelux array, and provide 3600 lumens of neutral white light. The crees cost a bit more though...I think that you can get three bridgelux for the price of four Crees. The Brudgelux only offers 60 lumens per watt in neutral white, so they don't look very appealing compared to Cree's 85lm/W at max.

    Anyway, only testing will tell the story.

  10. #10
    Join Date
    Oct 2004
    Location
    Beaverton, OR
    Posts
    2,842
    Feedback Score
    1 (100%)

    Default

    I ordered a BXRA-40E2200-B-00 to play with. 3500 lumens at 40W and neutral white. 23mm star so it will need some optics magic to get a wide beam.

Posting Permissions

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