| ||||||||||||||||
| ||||||||||||||||
|
"LED bulb" retrofits are not safely usable *not* just because they lack an E-mark, but because they are fundamentally incompatible with optics designed for use with a filament bulb. It's not a question of "optimisation", as you stated. It is a question of fundamental optic design, and the output characteristics of LEDs differ so radically from those of filaments that a drop-in "LED bulb" replacement is just not physically possible right now. About 10 days ago, I had dinner with Osram Optoelectronics' LED R&D chief, and we discussed in detail the question of an LED device to directly and effectively/safely/legally replace a standard filament bulb. He's thought about this question at great length, and done a great many experiments and studies. At the **CONCEPTUAL** level, it could theoretically be done using a Lambertian-emitting LED powerhead (this is the side-and-rear emitting type which has the same surface luminance regardless of the angle of view -- in contrast to the conventional "straight ahead" emitters which produce a spot beam facing straight away from the emitter). The powerhead would have to be engineered such that its surface luminance and far-field illuminance are within the 90th percentile tolerance of the luminance and illuminance of the specific filament bulb one wanted to replace, *from all polar angles*. For a P21W or P21/5W replacement, this would take an estimated 10,000 man-hours of R&D because no existing product even remotely approaches the 90th-percentile tolerance range relative to P21W or P21/5W. The R&D chief estimated this would begin to become an economical product to develop at high OEM volumes of around 10 million, which defeats the point of developing such a device, because at that point it becomes more economical to develop a new LED solution independent of what went before, i.e., we discard the notion of backwards compatibility with filament bulbs. There is an existing first-generation Lambertian-emitting LED powerhead, the Osram Joule, http://ledsmagazine.com/articles/news/2/6/26 . It is used in the current Mercury Mountaineer SUV in the North American market. An optical system (lens and reflector) can be designed that will produce compliant performance from either a Joule LED or a conventional bulb, **If contemplated from the lamp design stage that there will be bulb and LED versions of the lamp**. He has spent many hours testing Joule LEDs in pre-existing bulb-type optics-he has specially-modified Joule powerheads that can be moved axially to correct the focal length for whatever lamp reflector is being tested. He mounts them in an axial micromanipulator and finds the optimal placement of the powerhead. He has found a compliance rate of a little under 50%. That is, with the state-of-the-art LED emitter, he can make about half the bulb-type lamps he tests produce minimally compliant photometry. There's no predictability to it; some lamps that are conceptually identical but stylistically different will give different pass/fail results. And, the compliance success rate is helped by the relatively low intensity requirements for ECE-spec vehicle exterior lighting devices relative to North American-spec items. US amber rear turn signal intensity requirements are 130-750cd, ECE requirements are much lower at 45-180cd. Likewise, a US brake lamp is 70-300cd, while an ECE brake lamp is much lower at 54-167cd. So, the minimum permissible intensity of ECE brake and signal lamps is between 35% and 77% of the minimum permissible US intensity. Where overall intensity is a limiting factor in LED retrofit success, these numbers are significant. But all of this is theoretical right now. The "LED bulbs" presently available consist of a cluster of axial-emitting LEDs grafted onto a bulb-type base, and they are not only without E-mark, they are physically without hope of producing even barely minimally adequate safety performance. The result of using such an "LED bulb", under the best of conditions, is a 25mm spot of insufficiently intense light. Distribution and intensity of light through the required vertical and horizontal angles, and dispersion across the required minimum projected effective surface area, is not even remotely approached. And yes, this applies to the fancy/expensive "deluxe LED bulbs" with side-facing as well as rear-facing emitters, too. So, no, your guess that the legality of these devices "may change" has no basis in fact. Daniel Stern |
| ||||||||
|
| ||||||||
This site has no affiliation with |