I realize that its not legal (&reckless) to use DIY LEDS on a car. That being said, I am interested in exploring the design of an LED system for an automotive application that could theoretically meet DOT requirements for safety and efficacy. Kids, do not try this at home.
I measured the light pattern shining from my Infiniti G37 HID projectors last night. The lens is approximately 2 feet from the pavement. The light cast at any point in front of the car is cut off at about 2 feet from the ground, so if you are approaching a hitchhiking cheerleader, all that is really illuminated well is her legs. Anything above that line is kinda dark. The pavement that is in front of the car is illuminated from some distance far away up to 9.3 feet in front of the bumper. So, a light that projects no higher than itself (2 feet) and no lower than 9.3 feet ahead is about 12 degrees. Car lights also need to cast light to the sides of the road. So, if you didnt have access to the fancy optics that the car manufacturers are using, you could approximate the same beam pattern with multiple LED units. If we could find a combination of LED and lens that completely and evenly fill a 10-14 degree cone (without significant spill outside the cone) we would be set. It is important that the light does not exceed the spread limit because it will cause glare for other drivers. Each LED unit would need to be angled to the left and right to cover the field of view. 72 degrees of light seems to be what my car produces (well kinda...my lights swivel as I turn so the field is actually much wider, so side firing LEDs could be switched on to replicate this feature). Assuming only one unit is needed per section, a single bank of 6 units at 12 degrees each on a bicycle/motorcycle would approximate what my G37 headlights illuminate when pointing forward. Since automotive applications involves 2 banks of lights, additional emitter units would need to face inward to avoid a 6 foot wide dead zone directly in front of the vehicle. for 12 degree units, the math works nicely if each side has a unit with partially redundant coverage pointing dead ahead and has an additional unit inboard covering the middle. When looking at the front of the vehicle, the angles would be:
30,18,6,0,-12________________-12,0,6,18,30
One part of the math that I'm not able to wrap my head around is how the units relate to each other on the Y-axis. If the unit pointing dead straight is dipped down 6 degrees from the horizon, the neighbor units mounted to the same plate would rise higher than the horizon the farther they are swiveled to the side. In fact, if a side light was pointing perpendicular in relation to the forward light, it would be shining 6 degrees above the horizon.
If anyone knows the equation to calculate the Y axis angles I would appreciate it if you shared it with me.
It would be nice to incorporate high-beam capability...It would be possible to use wider angle units but incorporate a cutoff. In another thread I saw a guy fab a "hood" with a small lip hanging down in front of the light. I tested the idea with a dinner plate and a stock mag and it worked very nicely. My cars high-beams work by retracting a similar hood with a solenoid. Although it works, moving parts will eventually break...
An alternative to the solenoid and hood would be to place a switchable LCD (liquid crystal) plate where the cutoff should be. These things are sometimes called electrochromatic,electrochromic,SPD,smartfilm,smartglass,and thermoglass. Essentially it is frost white until current is applied, then it is clear. While it is translucent, light does in fact get through, but it would probably not produce much glare & look kinda neat. However, I have no idea if smartfilm wastes electricity and it is also very hard to find. There are youtube vids of license plate covers and window tint...all controlled by switches. A couple problems...Although the filter goes from frost to clear instantly, it takes a moment to go from clear to frost. Another problem is cold weather...I have no idea how all of this deals with freezing temps.
Has this been tried before? Any examples of 12 degree LED units?
Thanks for any input, Steve
I measured the light pattern shining from my Infiniti G37 HID projectors last night. The lens is approximately 2 feet from the pavement. The light cast at any point in front of the car is cut off at about 2 feet from the ground, so if you are approaching a hitchhiking cheerleader, all that is really illuminated well is her legs. Anything above that line is kinda dark. The pavement that is in front of the car is illuminated from some distance far away up to 9.3 feet in front of the bumper. So, a light that projects no higher than itself (2 feet) and no lower than 9.3 feet ahead is about 12 degrees. Car lights also need to cast light to the sides of the road. So, if you didnt have access to the fancy optics that the car manufacturers are using, you could approximate the same beam pattern with multiple LED units. If we could find a combination of LED and lens that completely and evenly fill a 10-14 degree cone (without significant spill outside the cone) we would be set. It is important that the light does not exceed the spread limit because it will cause glare for other drivers. Each LED unit would need to be angled to the left and right to cover the field of view. 72 degrees of light seems to be what my car produces (well kinda...my lights swivel as I turn so the field is actually much wider, so side firing LEDs could be switched on to replicate this feature). Assuming only one unit is needed per section, a single bank of 6 units at 12 degrees each on a bicycle/motorcycle would approximate what my G37 headlights illuminate when pointing forward. Since automotive applications involves 2 banks of lights, additional emitter units would need to face inward to avoid a 6 foot wide dead zone directly in front of the vehicle. for 12 degree units, the math works nicely if each side has a unit with partially redundant coverage pointing dead ahead and has an additional unit inboard covering the middle. When looking at the front of the vehicle, the angles would be:
30,18,6,0,-12________________-12,0,6,18,30
One part of the math that I'm not able to wrap my head around is how the units relate to each other on the Y-axis. If the unit pointing dead straight is dipped down 6 degrees from the horizon, the neighbor units mounted to the same plate would rise higher than the horizon the farther they are swiveled to the side. In fact, if a side light was pointing perpendicular in relation to the forward light, it would be shining 6 degrees above the horizon.
If anyone knows the equation to calculate the Y axis angles I would appreciate it if you shared it with me.
It would be nice to incorporate high-beam capability...It would be possible to use wider angle units but incorporate a cutoff. In another thread I saw a guy fab a "hood" with a small lip hanging down in front of the light. I tested the idea with a dinner plate and a stock mag and it worked very nicely. My cars high-beams work by retracting a similar hood with a solenoid. Although it works, moving parts will eventually break...
An alternative to the solenoid and hood would be to place a switchable LCD (liquid crystal) plate where the cutoff should be. These things are sometimes called electrochromatic,electrochromic,SPD,smartfilm,smartglass,and thermoglass. Essentially it is frost white until current is applied, then it is clear. While it is translucent, light does in fact get through, but it would probably not produce much glare & look kinda neat. However, I have no idea if smartfilm wastes electricity and it is also very hard to find. There are youtube vids of license plate covers and window tint...all controlled by switches. A couple problems...Although the filter goes from frost to clear instantly, it takes a moment to go from clear to frost. Another problem is cold weather...I have no idea how all of this deals with freezing temps.
Has this been tried before? Any examples of 12 degree LED units?
Thanks for any input, Steve
Last edited:
