Audi LEd Headlamps
Victor Roberts
that stated the 2007 Audi R8 would use Luxeon LEDs for both
high and low beam headlamp functions. I cannot confirm this
on the US Audi site. The information there says only that
the R8 has full LED tail lamps.
There is no reason why this can't be done if Audi is willing
to spend the money for enough LEDs. I think this is the
first production car that uses LED headlamps, but I could be
wrong.
--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.
This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.
LAI
A historic development for the LED and automotive industries was the
announcement in early 2004 of the first use of LEDs in the front end of
a car. The 2004 Audi A8 6.0L (a $120,000 luxury vehicle) used six 1-W
white Luxeon LEDs (manufactured by Lumileds Lighting of San Jose, CA)
in its daytime-running-lamp assembly. White LEDs use blue gallium
nitride (GaN)-based LED chips in combination with a yellow phosphor in
the same package to produce a white-appearing spectrum. Additional use
of white LEDs in daytime running lamps and fog lamps on other car
models is expected over the next several years.
http://lfw.pennnet.com/articles/article_display.cfm?article_id=241158
TKM
"Victor Roberts" <x...@lighting-research.com> wrote in message
news:bc5eo25vqcu3offp8...@4ax.com...
>I received an e-mail announcement from Lumileds on Friday
> that stated the 2007 Audi R8 would use Luxeon LEDs for both
> high and low beam headlamp functions. I cannot confirm this
> on the US Audi site. The information there says only that
> the R8 has full LED tail lamps.
>
> There is no reason why this can't be done if Audi is willing
> to spend the money for enough LEDs. I think this is the
> first production car that uses LED headlamps, but I could be
> wrong.
>
>
> --
> Vic Roberts
Hmmm. Didn't think there would be enough intensity for the headlamp
application just yet. The right beam pattern shouldn't be impossible
though. Maybe DS has heard something.
Terry McGowan
Ian Stirling
> I received an e-mail announcement from Lumileds on Friday
> that stated the 2007 Audi R8 would use Luxeon LEDs for both
> high and low beam headlamp functions. I cannot confirm this
> on the US Audi site. The information there says only that
> the R8 has full LED tail lamps.
>
> There is no reason why this can't be done if Audi is willing
> to spend the money for enough LEDs. I think this is the
> first production car that uses LED headlamps, but I could be
> wrong.
LEDs could be much more flexibly packaged than one filliment.
Instead of your typical large glass sealed unit, imagine a flat, or
curved panel, with a couple of dozen moulded reflectors each with an
LED in.
Also, aim each reflector differently and you can fade them in and out
to dynamically alter the light pool.
This may simplify the structure, remove completely any need for
mechanical headlight trim motors, reduce weight, give more flexibility
than simply dip/main-beam, amongst other things.
Boxman
Schefenacker will be detailing how they developed their LED headlamp at
the 2007 Strategies in Light Conference
(http://sil07.events.pennnet.com/fl/index.cfm) and Hella had an all LED
headlamp prototype on the VW Golf in ~ 2005 that I think met all SAE
requirements, but was short on the ECE high beam requirement.
The led headlamp offers some styling advantage and will be useful for
the implementation of AFS lighting systems where the headlamp provides
multiple beams for different conditions when those regulations are
finally approved. But right now, they are going to be too costly for
use on most mass produced vehicles and their power consumption might
exceed that needed for even conventional incandescent lamps in some
cases. The other problem is they are going to need massive heat sinks
(I think Reiter and Schefenacker mentions that they are using active
cooling to keep the LEDs cool) initially to control the heat
generation, and until heat sinking technology improves, or the light
output improves enough to lower the number of LEDs needed, the
headlamps are going to be heavier than conventional designs as well.
THe luminance of currently available LEDs is also going to require the
LED headlamps to be pretty large in comparison to what can be used for
incandescent and arc designs.
The optical design is extremely challenging and buidling an LED in
headlamp in mass production is going to be difficult. While the arrays
of LEDs might look cool, imagine trying to correctly line up an array
of optics over an array of LEDs and get the alignment correct so the
optical performance is ok. The production tolerances will need to be
very tight, which isn't necessarily conducive to mass production. In
the traditional 1 bulb system the tolerances aren't nearly as hard to
control. Some companies appear to be going to using an individual
optic separate to each LED rather than part of an array to provide
better control of the alignment and maintaining the beam pattern.
The arc discharge headlamps have just as much if not more light output,
long life times and consume less power, probably cost less than the
initial led headlamps and with only the single arc to position in the
optical system make it much easier to control in production.
But as usual the LEDs have that cool factor that will probably make
them very popular despite their drawbacks.
Ian Stirling
> I'm not sure who developed the Audi Headlamp, but Reiter and
> Schefenacker will be detailing how they developed their LED headlamp at
> the 2007 Strategies in Light Conference
> (http://sil07.events.pennnet.com/fl/index.cfm) and Hella had an all LED
> headlamp prototype on the VW Golf in ~ 2005 that I think met all SAE
> requirements, but was short on the ECE high beam requirement.
>
> The led headlamp offers some styling advantage and will be useful for
> the implementation of AFS lighting systems where the headlamp provides
> multiple beams for different conditions when those regulations are
> finally approved. But right now, they are going to be too costly for
> use on most mass produced vehicles and their power consumption might
> exceed that needed for even conventional incandescent lamps in some
> cases. The other problem is they are going to need massive heat sinks
Taking for example, Cree current production LEDs, in my very boring car.
It has 55W halogen headlights.
These produce around a thousand lumens.
Assuming LEDs with a total output of 2000 lumens (per light) are wanted,
then this is around 16 LEDs, for a total cost of around 180 pounds.
Assuming that the LEDs are half the total cost of the total assembly.
This produces about 20W of heat (half the 'beam' being on at once).
20W, over an area of 30cm*20cm or so isn't a very large amount at all.
> The optical design is extremely challenging and buidling an LED in
> headlamp in mass production is going to be difficult. While the arrays
> of LEDs might look cool, imagine trying to correctly line up an array
> of optics over an array of LEDs and get the alignment correct so the
> optical performance is ok. The production tolerances will need to be
> very tight, which isn't necessarily conducive to mass production. In
I'd be looking at a glass outer lens, an aluminised reflector, with
a single reflector per LED, with behind it an aluminium pressing, with a
flexible circuit carrying the LEDs glued on.
SimonLW
must spend extra for color matched parts. I bought several Luxeon
flashlights on closeout for $9 each and each one has a different shade of
white. I opened the head to confirm it was a real Luxeon and not the Chinese
knockoff.
I see that you can buy a 3 watt Luxeon Maglight at Home Depot for $16 now,
so the price is getting quite reasonable. I was wondering when these LEDs
might start appearing on the front end of cars, but already were!
-S
Boxman
> It has 55W halogen headlights.
> These produce around a thousand lumens.
> Assuming LEDs with a total output of 2000 lumens (per light) are wanted,
> then this is around 16 LEDs, for a total cost of around 180 pounds.
> Assuming that the LEDs are half the total cost of the total assembly.
>
> This produces about 20W of heat (half the 'beam' being on at once).
>
> 20W, over an area of 30cm*20cm or so isn't a very large amount at all.
Too optimistic of an analysis. The LED luminance is much lower than
your 55 watt headlamp filaments and thus require more lumens to get the
same intensity levels. The high brightness LEDs actually present a
larger source area to design with than most conventional halogen lamp
filaments. Even with extremely efficient optic designs you won't get
more than 85% of the light collected from the LED and then you have at
a minimum another 10% loss on the outer lens. Practically speaking
because the low beam pattern is such an assymetric pattern, most
headlamp optical systems won't even approach a 50% efficiency.
The high beam requires nearly twice the lumens as the low beam, so if
you need 16 for a low beam, you might need another 16+ the low beam to
get high beam. Now you are running 32 LEDs at one time.
The amount of wattage for each LED is going to be near 3.5 - 4 watts if
you want to get the 125 lumens you are assuming in your calculation.
So even if they were able to get away with around 16 LEDs on the low
beam you are still looking at around 50 watts of power. And that has
to be pulled away from the LEDs to keep their junction temperatures
from getting too high and decreasing the light output along with
jeopardizing the life rating for the LED. There can also be a
noticeable color temp shift with the increasing junction temperature.
I think you would be surprised what it would take to keep the LED
junction temperature to even a modest level. Check out Lumileds
application brief on the K2 LED heat sink requirements.
> I'd be looking at a glass outer lens, an aluminised reflector, with
> a single reflector per LED, with behind it an aluminium pressing, with a
> flexible circuit carrying the LEDs glued on.
Again an aluminum reflector is only 85% efficient to start with because
of the reflectivity of the mass produced aluminum vapor coating.
Reflectors are not good systems to use with the high brightness LED for
efficient optics because the LED emits entirely in the forward
direction into basically a 60 degree half angle. A typical reflector
can only collect a little less than half the light from the LED and you
would have to block the upward direct light portion not captured by the
reflector to keep it from going above the horizon and causing glare.
Putting the LED glued to a flexible circuit then pressed to aluminum is
creating a high thermal resistance path between the LED and the ambient
air which isn't going to work too well. The thermal resistance of some
of the best LED packages themselves is around 9 deg c/w mounted to a
metal core PCB so if you have 4 watts flowing through the LED you are
creating at least a 35 degree change in junction temperature even if
you are efficiently sinking the heat away from the base of the LED
package. That means the LED junction is going to get to 60 degrees
under even really good design conditions assuming an ambient
temperature of 25 degrees C. Which is probably not a good assumption
for ambient temperature in an engine compartment. Allowing the LEDs to
approach their maximum junction temperature will destroy their lifetime
and color performance.
Richard
of 9011 and 9012 bulbs in a adaptave fixtures. Self aiming, aiming into a
turn and all. Close to HID intensity and easy for the consumer to replace
the bulbs. I have used these bulbs in a Toyota RAV4 and a PT Cruiser and the
improvement over non-IR bulbs is significant and the cost is quite
reasonable. [But what do I know?]
Richard.
Victor Roberts
This is not about saving money, it is about spending money
to show off.
Daniel J. Stern
Victor Roberts wrote:
> This is not about saving money, it is about spending money
> to show off.
At this stage of the game, you're absolutely right. Automotive headlamp
applications using LEDs are just beginning to be introduced in limited
volume production, but have been undergoing very active development for
about 5 years. The present state of the art units have performance and
power consumption roughly equal to existing halogen headlamps, with
extremely high system cost. Currently, LED headlamps require large
packaging and a large number of the most powerful LED emitters
available. As LED technology continues to evolve, the performance of
LED headlamps is predicted to improve to approach, meet, and perhaps
one day surpass that of HID headlamps. LED headlamps, foglamps and
other forward illumination devices have so far generally been featured
only on manufacturers' concept cars, but the first series-production
LED headlamps will be factory-installed on the 2007 Lexus LS 600h / LS
600h L. They will also appear on the version of the 2007 Audi R8 sports
car sold outside North America.
LEDs actually produce a significant amount of heat per unit of light
output. Rather than being emitted together with the light as is the
case with conventional light sources, an LED's heat is produced at the
rear of the emitter. The cumulative heat of numerous high-output LED
emitters operating for prolonged periods poses thermal-management
challenges for plastic headlamp housings. In addition, this heat
buildup can materially reduce the light output of the emitters
themselves. LEDs are quite temperature sensitive, with many types
producing at 30° C (85° F) only 60% of the rated light output they
produce at an emitter junction temperature 16° C (60° F). These
needs, to exhaust heat from the headlamps and to keep LED emitter
junction temperatures under control, require expensive powered
ventilation systems.
Additional facets of the thermal issues with LED headlamps reveal
themselves in cold ambient temperatures. Many types of LEDs produce at
-12° C (10° F) up to 160% of their 16° C (60° F) rated output. The
temperature-dependency of LEDs' light output creates challenges for the
engineering and regulation of automotive lighting devices, which are in
some cases required to produce intensities within a range smaller than
the variation in LED output with temperatures normally experienced in
automotive service. Of course, the temperature-related output variation
can be managed with effective thermal management and
thermal-compensating driver circuitry...but this further elevates the
system cost.
Cold weather also brings another thermal-management conundrum: Not only
must heat be removed from the rear of the headlamp so that the housing
does not deform or melt and the emitters' output does not drop
excessively, but heat must in addition be effectively applied to the
front lenses of the lamps-which are not heated by the cold light beam
produced by LEDs-to provide rapid and complete thawing of snow and
ice accumulation.
(If you think you've read parts of this text above, you might've. I
wrote most of it for another forum.)
I was at the SIA-V.I.S.I.O.N. automotive lighting technology congress
in France last Autumn. As was the case at PAL in Germany the previous
Autumn, there was a great deal of activity related to LED forward
illumination. There were some good papers presented, some new
developments shown, and several night-drive test cars equipped with
various makers' LED headlamps. It was interesting to note the
performance improvement of the headlamps on Osram's fully-LED-lit BMW
at V.I.S.I.O.N. relative to the earlier version shown at PAL in 2005.
It was also interesting to compare the realisations of different
optical techniques and philosophies regarding light distribution within
the beam pattern, CCT, packaging and thermal management.
This is, however, the first time the development of motor vehicle
forward illumination has been driven by stylists and marketers rather
than by engineers. In every previous technological jump or shift, the
engineering has come first and the stylists have worked to integrate
the physical attributes of the new technology into their designs. In
this case, stylists and marketers have demanded LEDs and the attendant
"wow" appearance, and engineers have had to scramble to provide it in
an efficacious and cost-effective manner.
Stylists and marketers have also demanded an appearance significantly
bluer than is provided by current-production HID headlamps, which has
touched off a debate between those researchers whose work suggests
bluer light gives better seeing performance, and those whose work
suggests bluer light creates more glare without any attendant seeing
performance improvement.
The regulatory debates centre around the colour issue and whether & how
to standardise LED light sources or light source modules -- the latter
idea being to see to it that a vehicle owner facing a dimmed or dead
LED emitter isn't faced with the choice of driving with degraded
headlighting (possibly failing a roadworthiness inspection) or spending
a large sum of money for a unitised replacement headlamp. This is a
thornier issue than it has been with past headlamp light source
innovations, because of the extremely steep evolution curve of LEDs and
the much more critical focusing issues mentioned by Larry Boxler
earlier in this thread.
DS
Daniel J. Stern
Boxman wrote:
<big snip>
> Reflectors are not good systems to use with the high brightness LED for
> efficient optics
Larry-
Certainly conventional reflectors as we've known them aren't well
suited to LEDs, but there were some highly innovative reflector optic
systems shown and described for LED headlamps at SIA-V.I.S.I.O.N. If
you'd like me to send you the relevant PDFs, send me an e-mail, dastern
at torque dot net .
DS
mp...@oscintl.com
Boxman wrote:
> > Taking for example, Cree current production LEDs, in my very boring car.
> > It has 55W halogen headlights.
>
> > These produce around a thousand lumens.
> > Assuming LEDs with a total output of 2000 lumens (per light) are wanted,
> > then this is around 16 LEDs, for a total cost of around 180 pounds.
> > Assuming that the LEDs are half the total cost of the total assembly.
> Too optimistic of an analysis. The LED luminance is much lower than
Great discussion on HB LED's
>From your description of the 1 watt HB LED's and needing about 16 of
them it makes me think that in the auto field that they are not using
pulse width modulation (PWM) to drive them harder (more current 750mA)
in order to get a higher luminance with fewer HB LED's. If they were
it might be possible to use 8 or 10 to get the same total time averaged
output. Does the SAE spec allow PWM in headlights?
What is the correct SAE spec number to cover headlights and tail
lights?
This PWM is being used to drive the HB LED's in digital projectors to
get a small but sufficient number luminance out for viewing in a dark
room.
Another thought that comes to mind for cooling HB LED's in autos is to
use peltier or thermo-electric coolers with some energy from the
alternator or some of the antifreeze from the radiator or the heat
pipes that are being used with laptop computers and HB LED's in digital
projectors.
By the way it is fairly easy to measure if the head or tail lights are
being PWM by using a fast si diode detector right in front of the head
or tail light.
Michael
www.oscintl.com
Ian Stirling
>
> Boxman wrote:
>> > Taking for example, Cree current production LEDs, in my very boring car.
>> > It has 55W halogen headlights.
>>
>> > These produce around a thousand lumens.
>> > Assuming LEDs with a total output of 2000 lumens (per light) are wanted,
>> > then this is around 16 LEDs, for a total cost of around 180 pounds.
>> > Assuming that the LEDs are half the total cost of the total assembly.
>
>> Too optimistic of an analysis. The LED luminance is much lower than
>> your 55 watt headlamp filaments and thus require more lumens to get the
> Great discussion on HB LED's
>
>>From your description of the 1 watt HB LED's and needing about 16 of
> them it makes me think that in the auto field that they are not using
> pulse width modulation (PWM) to drive them harder (more current 750mA)
> in order to get a higher luminance with fewer HB LED's. If they were
> it might be possible to use 8 or 10 to get the same total time averaged
> output. Does the SAE spec allow PWM in headlights?
The LEDs in question - and I'm pretty sure that I'm aware of all the
major players at the moment - all have significantly lower efficiency at
higher drive currents. (once you get over a few tens of milliamps).
From memory, typical is 40% over nominal efficiency at 20mA, and 40%
under at 700mA - for a part that is rated at 350mA nominal, 700mA max.
And unless you make the light flicker visibly, there is no benefit in
signalling your presense to others, as it will be dimmer, due to the
lower efficiency.
Anyone designing intentionally flickering headlights should be
forced to only work under strobe lighting for a few years.
Daniel J. Stern
> From your description of the 1 watt HB LED's and needing about 16 of
> them it makes me think that in the auto field that they are not using
> pulse width modulation (PWM) to drive them harder (more current 750mA)
> in order to get a higher luminance with fewer HB LED's. If they were
> it might be possible to use 8 or 10 to get the same total time averaged
> output. Does the SAE spec allow PWM in headlights?
I haven't looked, and don't have time at the moment, but I doubt it's
prohibited.
> What is the correct SAE spec number to cover headlights and tail
> lights?
No such animal, per se. Each and every function (tail lamps, stop
lamps, directional indicators, back-up lights, fog lamps, headlamps in
general, HID headlamp systems, LED headlamp systems, etc.) is covered
by one or more SAE standards and/or recommended practices. There are
also overspanning SAE standards for such things as colorimetry, plastic
materials used in vehicle lighting devices, etc.
Here's a non-exhaustive list of relevant SAE standards:
J2650 (LED forward illumination device systems):
https://shop.sae.org/technical/standards/J2650_200509
J1889 (LED vehicle signals):
https://shop.sae.org/technical/standards/J1889_200507
J578 (Color):
https://shop.sae.org/technical/standards/J578_200612
J576 (Plastic materials):
https://shop.sae.org/technical/standards/J576_199107
J2009 (Discharge forward illumination device systems):
https://shop.sae.org/technical/standards/J2009_200510
J2591 (Adaptive Forward Lighting systems):
https://shop.sae.org/technical/standards/J2591_200209
J586 (Brake lamps for non-large vehicles):
https://shop.sae.org/technical/standards/J586_200003
J1957 (Center high mount brake lamps for non-large vehicles):
https://shop.sae.org/technical/standards/J1957_200003
J588 (Turn signals for non-large vehicles):
https://shop.sae.org/technical/standards/J588_200003
J319 (Rear fog lamps):
https://shop.sae.org/technical/standards/J1319_200505
J592 (Sidemarker lamps for non-large vehicles):
https://shop.sae.org/technical/standards/J592_200508
J222 (Parking lamps):
https://shop.sae.org/technical/standards/J222_200612
Keep in mind, The SAE standards themselves are legally authoritative
only to the partial degree they are referenced and/or codified in US
FMVSS 108 and Canadian CMVSS 108. Frequently, SAE standards relevant to
any particular device or system are not adopted in their entirety.
Certain technical provisions may be omitted, modified, or provisions
from older superseded versions of SAE
standards may remain in force in MVSS 108.
In the ECE regulations used outside North America, SAE standards are
generally not referenced or codified in whole or in part. Technical
provisions are sometimes identical, sometimes diifferent but
compatible, sometimes wholly different. You can freely download all the
ECE regulations here:
http://www.unece.org/trans/main/wp29/wp29regs.html
> By the way it is fairly easy to measure if the head or tail lights are
> being PWM by using a fast si diode detector right in front of the head
> or tail light.
In the case of parking and tail lamps, it is frequently even easier
than that: Just scan your gaze rapidly across the lamp, and in many
cases you will get a very conspicuous bead effect!
DS
Boxman
> >From your description of the 1 watt HB LED's and needing about 16 of
> them it makes me think that in the auto field that they are not using
> pulse width modulation (PWM) to drive them harder (more current 750mA)
> in order to get a higher luminance with fewer HB LED's. If they were
> it might be possible to use 8 or 10 to get the same total time averaged
> output. Does the SAE spec allow PWM in headlights?
>
I'm just not sure how PWM would help increase the output. For
stop/tails as Daniel pointed out, It's used to decrease the output from
a steady burning stop lamp to a tail lamps to get the required ratio of
intensity between a stop and tail instead of cutting the current down.
I recall that this is mostly due to the HB LEDs not necessarily
operating reliably with the lower currents.
Because of the high output requirements for headlamps, the HB LEDs used
in these applications are probably driven near or close to their max
current ratings already. How would PWM increase the amount of light
available? Even if they are at something like a 33% duty cycle, you
would need to run 3X the maximum steady state rated current to get the
same brightness. Seems to me this might cause some color shifting or
life deprecation for little if any benefit. Maybe I'm missing
something.
Victor Roberts
wrote:
There is an urban myth that states you can fool the eye by
pulsing a light source at rates higher than the flicker
fusion frequency. It doesn't work.
Boxman
>
> There is an urban myth that states you can fool the eye by
> pulsing a light source at rates higher than the flicker
> fusion frequency. It doesn't work.
>
Thanks Vic - I'm not as familiar with the electronics side so I didn't
want to say one way or another. I suspected this was the case.
mrob...@worldnet.att.net
> As LED technology continues to evolve, the performance of LED headlamps
> is predicted to improve to approach, meet, and perhaps one day surpass
> that of HID headlamps.
Do you think that the technology will eventually be good enough to
create a self-contained LED headlamp module that would be a drop-in
replacement for a plain ol' sealed beam? If the technology does get
there, will there be enough cars with sealed beams still on the road
for any of the manufacturers to care?
As far as I know, there aren't yet any (legal) HID modules that can be
installed in place of a sealed beam. I think the main problem is that
the reflector still needs most of the volume of the lamp, which leaves
little room for the power supply. But it seems like the working part of
an LED headlamp could end up being a relatively thin LED - reflector -
lens "sandwich", leaving plenty of room for electronics in the rest of
available space.
> Cold weather also brings another thermal-management conundrum: [...]
> heat must in addition be effectively applied to the front lenses of
> the lamps-which are not heated by the cold light beam produced by
> LEDs-to provide rapid and complete thawing of snow and ice
> accumulation.
How are the current designs doing this - blowing the LED-cooling air at
the lens? Using resistance heat like a rear window defogger? Or...?
> The regulatory debates centre around the colour issue and whether &
> how to standardise LED light sources or light source modules -- the
> latter idea being to see to it that a vehicle owner facing a dimmed
> or dead LED emitter isn't faced with the choice of driving with
> degraded headlighting (possibly failing a roadworthiness inspection)
> or spending a large sum of money for a unitised replacement headlamp.
Given the number of cars I've seen on the road this winter with no
apparent accident damage but only one working headlamp, apparently
US$10 qualifies as "a large sum of money".
Matt Roberds
LEDNeonFlex
in China and Nichia has LED's that 150Lumens/Watt CFL is at most
80Lumens/Watt... You will see this year home and commercial lighting
drop in replacements with many many cool mini light power house fixtures.
Derek Breneol
LED Flex Lighting Co.
www.ledneonflex.com
www.ledflexlight.com
Victor Roberts
<de...@ledneonflex.com> wrote:
>This market is moving fast, Edison Opto is manufacturing 100watt led's
> in China and Nichia has LED's that 150Lumens/Watt CFL is at most
>80Lumens/Watt... You will see this year home and commercial lighting
>drop in replacements with many many cool mini light power house fixtures.
>
>Derek Breneol
>
>LED Flex Lighting Co.
>www.ledneonflex.com
>www.ledflexlight.com
I'm not sure what you mean by "cool mini" but if you mean
decorative, very low lumen fixtures then I have no issue
with your prediction. However I do not expect that we will
see fixtures producing a useful amount of white light for
general lighting this year, except for a few very expensive
concept fixtures designed to replace incandescent-based
fixtures.
That wonderful LED efficacy we read about is still confined
to low power devices that are tested at a junction
temperature of 25C and often don't have any light absorbing
packaging. Real world devices do not perform as well. The
DOE just concluded a round of tests of LED-based fixtures
and they were surprised to find that the fixture
performance was far below the performance that they expected
by reading the LED data sheets. They should not have been
surprised. They seem to have ignored at least the junction
temperature issue, plus perhaps driver efficiency and the
"typical" vs. "minimum guaranteed" issue.
Daniel J. Stern
> Do you think that the technology will eventually be good enough to
> create a self-contained LED headlamp module that would be a drop-in
> replacement for a plain ol' sealed beam?
Indubitably. The question is whether they could be sold at a price
anyone would pay, and whether there's sufficient demand for form
factors that have been almost entirely abandoned in new-vehicle
production.
Truck-Lite are keen to develop LED headlamps, and it is logical to
guess that they might try for a standard-format unit.
> As far as I know, there aren't yet any (legal) HID modules that can be
> installed in place of a sealed beam.
Correct-there aren't any HID modules (illegal or otherwise) that can
presently be installed in place of a sealed beam. However,
ValeoSylvania now have a single-emitter round LED lamp unit sized to
replace a PAR36 in Harley-Davidson's so-called "passing lamp" units
found on some models outboard of the central main headlamp. This
PAR36-sized (4-1/2" diameter) beam unit meets SAE fog lamp
requirements, which isn't especially difficult since the minimum
allowable performance for SAE Fog is quite minimal indeed, but that's
sort of beside the point, isn't it!
> > Cold weather also brings another thermal-management conundrum: [...]
> > heat must in addition be effectively applied to the front lenses of
> > the lamps-which are not heated by the cold light beam produced by
> > LEDs-to provide rapid and complete thawing of snow and ice
> > accumulation.
>
> How are the current designs doing this - blowing the LED-cooling air at
> the lens? Using resistance heat like a rear window defogger?
I'm pretty sure both approaches are being played with. Remember, the
current designs are only just beginning to exist!
DS
Ian Stirling
> On Sat, 06 Jan 2007 03:24:42 GMT, LEDNeonFlex
> <de...@ledneonflex.com> wrote:
>
>>This market is moving fast, Edison Opto is manufacturing 100watt led's
>> in China and Nichia has LED's that 150Lumens/Watt CFL is at most
>>80Lumens/Watt... You will see this year home and commercial lighting
>>drop in replacements with many many cool mini light power house fixtures.
> That wonderful LED efficacy we read about is still confined
> to low power devices that are tested at a junction
> temperature of 25C and often don't have any light absorbing
> packaging. Real world devices do not perform as well. The
For applications in which the beam is not critical, and the LEDs can be
kept reasonably cool (say 60C), exceeding your common or garden CFLs
efficiency is not hard.
Not by much, but by a bit, and it's lots more expensive in most cases.
(the exception is where you need 1/20th or so of the light of a 100W
bulb, in a small package, and have a convenient power supply already)
It's going to be a while before we see the '135lm/W' banner headlines
translating into lights that can beat the efficiency of linear
fluorescents.
LED lights at the moment, as a replacement for CFL are barely
economically viable, and for them to be so, you have to pretty much
assume that it's a constantly on light.
As a replacement for incandescant, they are easily viable, but...
mp...@oscintl.com
> Here's a non-exhaustive list of relevant SAE standards:
>
>
> In the ECE regulations used outside North America, SAE standards are
> generally not referenced or codified in whole or in part. Technical
> provisions are sometimes identical, sometimes diifferent but
> compatible, sometimes wholly different. You can freely download all the
> ECE regulations here:http://www.unece.org/trans/main/wp29/wp29regs.html
>
Wow thanks very much for this list of SAE and ECE spec & regulations.
> > By the way it is fairly easy to measure if the head or tail lights are
> > being PWM by using a fast si diode detector right in front of the head
> > or tail light.In the case of parking and tail lamps, it is frequently even easier
> than that: Just scan your gaze rapidly across the lamp, and in many
> cases you will get a very conspicuous bead effect!
>
> DS
I saw this bead effect last year 2006 on what I think was a Cadillack
STS-V at dusk. I knew it was PWM LED's because of this bead effect. I
found a photo of it at www.ledsmagazine.com in the January 2005 issue.
There is also a photo of the AudiR8 head lamp in the December 2006
issue of the same magazine.
On the flicker frequency dont forget about the white light LED's use a
phosphor to up convert from uv/blue to white and the phosphors have
longer decay time than the LED die. Sorry I dont have any quantative
data.
Michael
www.oscintl.com