LED projector headlight experiment

[tjts1]

I've had a pair of BMW E34 (early 90s 5 series) projector headlights from the junkyard for a while now. At first the plan was to find a way to mount them on my E30 3 series but the housings were too big. Finally the other night I took one apart just to play around with. Seeing that I had a bunch of random LEDs sitting around I decided to try some experiments. at first I tried sticking an LED inside the reflector where the normal bulb would sit. Not much success there. Then I separated the lens and cut off shield from the reflector.
Reflector with 9006 bulb left, lens and cut off shield in the middle.

Took my cheap 21 LED flash light...

And shined it through the projector lens and cut off shield without a reflector.

I expected the cut off to be all screwed up because the LEDs are all over the place with in a 1" circle. But the exact opposite hapend. I got a perfect low beam pattern every time. If I move the flash light to the left, the pattern stays fixed but the left side becomes brighter and vice versa. As it turns out in this projector, if you remove the reflector you can position any number of bulbs anywhere so long as they shine light from behind the cut off shield toward the lens.

To take the experiment one step further, I turned my Ipod to a blank white screen, placed the projector lens on top

And again, perfect beam pattern on the ceiling. Sorry the picture is so dark. It took a few tries with my cheap camera.

I think this setup is even better than a normal halogen or HID bulb because you don't get any of the rainbow effect around the cut off. Its just a perfectly even clean while light with a cut off at the top and a gradient toward the bottom.
So the next step would be to build an actual functional headlight with the correct brightness, color and functionality. Again, this still an experimental step forward. Im not ready to install this in a car quite yet.
I think I'm going to use 10 Luxeon Rebel LEDs in neutral white mounted on a large heat sink.
http://www.luxeonstar.com/luxeon-rebel-led-neutral-white-lambertian-145-lm-700ma-p-172.php
I'm not worried about housing, making it water proof or any of the other practical aspects of a headlight.
Comments, ideas?
thanks
Justin

 

[-Virgil-]

Yep, this shows a critical point to understand about projector headlamps: they will generally maintain more or less their nominal cutoff configuration no matter what light source you might put behind the cutoff shield. Halogen bulb, bunch of LEDs, Zippo lighter, handful of excited fireflies, extremely small star, whatever. As you correctly point out, the tricky part will be getting appropriate amounts of light, appropriately distributed, under the cutoff. (This is also the problem with "HID kits" in projectors; they can give the false impression of a hunky-dory result because the cutoff shape stays about the same, but the light distribution under the cutoff is AFU). Note that projector optics are used with LEDs in the production Lexus LS600h and Cadillac Escalade Platinum LED headlamps. They use multiple small projectors in front of multiple emitters.

Keep in mind the thermal side of things, that is a huge management challenge in the creation of LED headlamps that work OK in real-world driving conditions. Not just heat sinking, but heat transport. Remember, you've got to be able to defog or thaw frozen lenses in winter. Maybe not a problem if you live someplace where it's always warm and dry.

(As an aside...why trying to mount E34 projectors on an E30? I can think of two reasons not to do it: The E24/E30/E32/E34 projectors with 9006 bulbs are really inefficient optics, and they made a version of them specifically for the E30...)

 

[F250XLT]

The cutoff will always stay the same?
What controls this?

 

[LukeA]

Keep in mind the thermal side of things, that is a huge management challenge in the creation of LED headlamps that work OK in real-world driving conditions. Not just heat sinking, but heat transport. Remember, you've got to be able to defog or thaw frozen lenses in winter. Maybe not a problem if you live someplace where it's always warm and dry.

There is the same issue with HID projectors.

 

[superjoe83]

The cutoff will always stay the same?
What controls this?

the cutoff shield, all low beam projectors have this in one form or another, here are a couple pics of one:

 

[-Virgil-]

There is the same issue with HID projectors.

Nope. Plenty of heat in the beam from an HID headlamp.

 

[LukeA]

Nope. Plenty of heat in the beam from an HID headlamp.

There's heat in LED beams too.

 

[-Virgil-]

There's heat in LED beams too.

Not anywhere near enough. Most LED heat goes out the back of the junction. Most tungsten-halogen or HID heat goes out the front with the beam. This is a very basic principle of LEDs vs. other headlamp light sources; you'll probably want to study up on it if you intend to try building LED headlamps.

 

[LukeA]

Not anywhere near enough. Most LED heat goes out the back of the junction. Most tungsten-halogen or HID heat goes out the front with the beam. This is a very basic principle of LEDs vs. other headlamp light sources; you'll probably want to study up on it if you intend to try building LED headlamps.

My outdoor LED architectural lights point up and are uncovered by snow, unlike the ground around them.

My LED headlight demonstrator, 10 XR-Es, puts out enough heat to make it uncomfortable to hold one's hand in front of it for very long, about 12W of radiation in all. So in terms of radiated heat, it's like a corner running light.

 

[-Virgil-]

My outdoor LED architectural lights point up and are uncovered by snow, unlike the ground around them.

Your outdoor LED architectural lights are not in service on the front of a vehicle.

 

[LukeA]

Your outdoor LED architectural lights are not in service on the front of a vehicle. Got any more non-automotive non-examples to throw at this known and significant design issue with automotive LED lamps?

My architectural LED lights point nearly straight up and still do not accumulate snow, despite accumulation of several inches around them. That means they radiate sufficient heat out the front to prevent their obfuscation by precipitation in that situation. I'm not pretending that it is a situation identical to that faced by automotive headlights, but I'm also not dumb enough to think that there are no parallels.

 

[blasterman]

Got any more non-automotive non-examples to throw at this known and significant design issue with automotive LED lamps?

I live in Michigan, which thanks to fast moving cold fronts from the plain states and Canada, along with Lake Effect Heisenberg weather generators causes driving conditions beyond description. Friends visting from Alaska recently had me drive them around town because they were that freaked by the weather conditions that seemed like Ming the Merciless was playing with his crystal ball / planet earth.

This entire 'halogen lights are safer than LED because they melt ice better' myth has me chuckling every time I chip ice off my Mazda M3 head lamps after driving home from work.

The reflector housings used in most of today modern headlamp designs are designed to reflect heat and otherwise keep from getting hot because otherwise they'd melt into slags of plastic. Indeed I recall headlamps from yesteryear that were less efficient glass designs getting much hotter.

You defeat your own point - given that Halogen and HID emits most heat in the form of IR then that energy gets projected away and doesn't melt much of anything. Indeed, the front plastic housings of my headlamps are barely warm to the touch after running for hours.

 

[-Virgil-]

This entire 'halogen lights are safer than LED because they melt ice better' myth has me chuckling

Then you're holding a golden ticket to a very generous salary with any of the world's headlamp OEMs, all of whom are investing a great deal of money, time, and effort to lick the LED headlamp thawing problem.

 

[blasterman]

I guess you're right and they're all wrong and stupid and dumb and stuff, eh?

I think you need a 'golden ticket' to a class on economics, supply chain mgmt., and obviously don't have a ISO 9000 cert (like I consequently do have, not that I care to brag about it).

The front lenses on my M3 put out LESS heat than some LED lights I've designed and used in home fixture designs, so that straw horse needs to be euthanized and buried.

Thermal issues with LEDs is the last thing on a long list that automobile makers are concerned with right now. Care to argue that point?

Also, Japanese automakers can't figure out how to replace tail lights with LEDs without causing major ripples in their supply chain. It doesn't do any good to have a front head light assembly that put's HID to shame when auto repair shops won't be able to figure out how to replace them, nor insurance companies willing to fork over the extra cash. Halogen bulbs are cheap, and the plastic reflector assemblies now used in most cars as lamp housings are already on the factory floor.

A LED assembly that produces superior light than a typical 55/65 halogen car lamp would be cheap to make and solve problems created with over-size halogen assemblies designed mostly to dissipate heat. The reduction in front chassis size and complexity alone is staggering when you think about it. However, this isn't going to happen soon because of the re-tooling costs and supply changes will cost too much. .

If you've invested tens of millions on plastic injection technology to make high tech lamp housings you aren't going to invest more millions on converting to LED even though it will save money in the long run. Might as well upscale those car buyers to HID so you can hopefully blind the deer standing in the road so it humanely doesn't know what hit it when you total out your BMW. About the only advantage to HID I can tell.

Also consider most drivers come from the same school as the dude who thinks putting a neodymium coating on a halogen bulb makes it brighter.

Not sure what you're problem is, but you're defending an industry who thinks 26mph is 'high gas mileage' and puts back seats in commuter cars to keep insurance rates down. Not much sympathy there.

 

[-Virgil-]

I'm not talking about the auto industry, but rather about the automotive lighting industry, who generally don't have a lot to say about fuel economy. It is in the interest of the automotive lighting industry for every driver to want (and every car to have) the most advanced lighting technology; if they had a magic wand to wave and give every new car LED headlights tomorrow, they'd certainly wave it. Vast money and effort are being poured onto the technical challenges; if it were as simple a matter as you seem to think, that would not be the case.

The literature is abundant and not very hard to get hold of -- you might try a search of the UMTRI library, or of Driving Vision News -- and it's being produced by scientists and engineers and researchers specially educated and specifically knowledgeable about headlamps, and they provide ample relevant demonstration and proof of their claims. I'm pretty sure they're not keeping anything behind their backs. Here's a sample page from a recent Hella (of Germany) technical presentation on the all-LED headlamp for the Cadillac Escalade. It's a PowerPoint slide, and there have been many such PowerPoint slides from the whole rest of the industry all over Europe, Japan, and small pockets of America. The industry really does seem to agree that lens thawing is a particularly challenging issue with LED headlamps:

If you're interested in more, I will try to dig up some paper titles for you (though it'll take some time; I am very busy with work). There have been presentations on the subject at the two major automotive lighting symposia for the last few years.

 

[DM51]

Thanks for the editing.

 

[mds82]

I am currently making myself LED Fog Lights.

I am using 6 Cree MC-E's and using mirrors as the optics.

I will be posting pics in the next few days, but it is going to look very good. I have it almost done and the cutoff is perfect.

 

[MichaelW]

When I was clearing off the snow/ice off the car last month, I turn on the low & high beams. (HIR retrofits) The plastic in front of the high beam gets plenty warm.
After five minutes, the very light falling snow would vaporize on contact with high beam area.
The low beams don't do this (the snow 'only' melts), my guess is the glare shield (non-projector headlights) reduces forward IR output.
http://www.autoguide.net/apf/images/products/keystone/xs7497886.jpg


A side marker bulb isn't 12 watts, maybe half that.

 

[-Virgil-]

Scheinwerfermann, your posts have become increasingly disrespectful and unpleasant in tone.

You're right; I let my keyboard run away with me. Sorry 'bout that, all. I've gone back and edited for tone and style.

 

[erikiksaz]

To the OP, are the projector lenses on the E34 completely clear? I know that the E30 series had clear lenses, but I haven't heard anything about the E34. Could you please take a picture of the lens sitting flat on a table over some lettering?