1.1 Definitions

In the present paper we'll call a “full LEDs headlamp” any lamp using only LEDs as light sources in order to achieve its main function, i.e. creating a passing and a driving beam and any other desired beams defined by the Rxx regulation (bending light, motorway, adverse weather or town light).

By “discrete optical system” we mean a complete optical system, including its light sources, physically separated from the other optical systems of the lamp.

1.2 Status

Several functional full LEDs headlamps (by opposition to other ones, mainly seen on concept cars or in earlier years, which did not provide really usable beams) have been already disclosed. Since a single LED does not yet provide enough light to create any of the beams mentioned in the previous chapter¹, the designers of these headlamps had to use several (and in truth many) LEDs (for example, we use 16 LEDs in the headlamp described in the present paper to create the passing beam alone).

In order to be able to more easily integrate many LEDs and the corresponding optical system inside existing headlamps housings and also to be able to show something never seen before when only bulbs were available, many existing full LEDs headlamps use discrete optical systems for nearly each LED involved.

This design strategy has advantages:

---

1 It is possible today to buy LEDs delivering around 80lm at 25°C junction temperature and

engineering samples have been measured by us above 100lm: at the real operating

temperature inside a headlamp (more than 90°C!), one can expect 80 to 100lm.

1. the single LED modules are small and easy to integrate: one can put useful LEDs in the smallest nooks and corners of the headlamp (this design flexibility also relates to a great styling freedom),

2. the same module can be used several times and so be manufactured as a standard element, at a lower cost.

It also has drawbacks:

1. For beams with a cutoff line (all beams under discussion, except the driving beam), each discrete optical system requires a mechanical adjustment device in order to properly superpose all cutoff lines on a screen 25m in front of the car²: with as many as 16 modules, the mechanical complexity of the system, the time required to make the final adjustment and the difficulty of designing an automated process to do it in mass production make this solution completely unrealistic for the automotive industry (moreover, for all modules with a 15° (or more as allowed by the AFS regulation) cutoff line, a two axes aiming system is necessary).

2. All considered beams require a leveling system. The leveling in a full LEDs headlamp may be achieved in two radically different ways: one can fasten every optical system to a common mounting plate or one can add an actuator to each module, several smaller mounting plates holding a reduced number of optical systems being a middle term approach. The first principle creates great problems: since the luminance of current LEDs is low³, the front surface used must be high to get the proper maximum intensity: the optical setup tends to use as much as possible of the outer lens surface and such a design combined with the leveling generated movements leads to high clearances and forbids using many out of the way corners of the headlamp which nevertheless could be large enough to accommodate a single LED optical system. The second strategy is both potentially very expensive and

---

² For a good beam, it seems necessary to keep all cutoff lines inside a less than 0.2% wide band, which corresponds to a ±0.1% angular tolerance (±0.06°). Such a tight tolerance will probably be unreachable without a final adjustment.

³ Around 14 Cd/mm² for the very best components when an halogen bulb (like an H7) reaches 20 Cd/mm² and a D2 xenon bulb 65 Cd/mm².

 

[1] [2] [3] [4] [5] [6] [7] [8] [9] 下一页