Xan Presents: Everything You Need to Know About Automotive Lighting

[bshanahan14rulz]

Maybe something to look at that could shed some light on the situation is the lux of a halogen filament at 55W compared to the lux of common LEDs used in these "cutting edge" LED drop-ins.

LEDs are making leaps and bounds in efficiency, but since the majority of the LED lighting industry is focused on lm/w, most of the common single-chip emitters with high luminous flux are the ones with larger dies. Larger dies means more power needed to reach a target lux, and more waste light that is not going to be at the focal cylinder of the incandesent reflector. Lumens are nice, but for lumens to be usable, they must originate from where the reflector expects them to originate from. So while your LED emitter chip is 2.5mm^2, you can only accurately use a small strip of that. Most of your light ends up "wasted" as glare, hotspots, or otherwise unintended reflection.

The bare fundamental question to ask regarding replacing halogen filaments with LEDs is, is it feasible to drive an LED so hard that its surface brightness matches the surface brightness of a halogen filament? In a hot engine bay, while mounted to a tiny twig of a heatpipe? I'm a bit behind, but I don't think LEDs have it in them, even today, to generate high enough lux.

You can put as many of these LEDs in whatever arrangement is marketable, but if the surface brightness isn't there, they will not produce the intended beam at the expected brightness level.

That said, if someone's dead-set on upgrading their lighting responsibly, a retrofit might be worth looking into. If you pull OEM components, you know they will perform as expected.

p.s.:
My 2000 Corolla had HIDs. OEM projector optics from an infiniti G35, new ballasts and bulbs from Philips/Hella and Philips, mounted to meet aiming specifications and rotational level. HIDs ain't as cool as LEDs, apparently, but they're easier to retrofit!
p.p.s: Trying to get my post count up, I should post shorter posts!

 

[SlowDunny]

All I get from this thread is that you need the right housing for different bulb types... It is informative but it doesn't build on any new concepts.

Maybe one of you can go out and build/test some lenses and light combos for the FiST. If you know how many lumens and the projection pattern, figuring out the right combination would just take some testing. It seems like there is enough knowledgable people and crazy people here to build things.

 

[danbfree]

Maybe something to look at that could shed some light on the situation is the lux of a halogen filament at 55W compared to the lux of common LEDs used in these "cutting edge" LED drop-ins.

LEDs are making leaps and bounds in efficiency, but since the majority of the LED lighting industry is focused on lm/w, most of the common single-chip emitters with high luminous flux are the ones with larger dies. Larger dies means more power needed to reach a target lux, and more waste light that is not going to be at the focal cylinder of the incandesent reflector. Lumens are nice, but for lumens to be usable, they must originate from where the reflector expects them to originate from. So while your LED emitter chip is 2.5mm^2, you can only accurately use a small strip of that. Most of your light ends up "wasted" as glare, hotspots, or otherwise unintended reflection.

The bare fundamental question to ask regarding replacing halogen filaments with LEDs is, is it feasible to drive an LED so hard that its surface brightness matches the surface brightness of a halogen filament? In a hot engine bay, while mounted to a tiny twig of a heatpipe? I'm a bit behind, but I don't think LEDs have it in them, even today, to generate high enough lux.

You can put as many of these LEDs in whatever arrangement is marketable, but if the surface brightness isn't there, they will not produce the intended beam at the expected brightness level.

That said, if someone's dead-set on upgrading their lighting responsibly, a retrofit might be worth looking into. If you pull OEM components, you know they will perform as expected.

p.s.:
My 2000 Corolla had HIDs. OEM projector optics from an infiniti G35, new ballasts and bulbs from Philips/Hella and Philips, mounted to meet aiming specifications and rotational level. HIDs ain't as cool as LEDs, apparently, but they're easier to retrofit!
p.p.s: Trying to get my post count up, I should post shorter posts!

The point I've been trying to make is that some LED's really do mimic the exact positioning quite nicely and look absolutely perfect... these shots are from a model unfortunately only available through Asian online markets but is absolutely incredible and word has gotten around... In fact if anything, there is no lack of output, it's that many overpower the reflectors to the point of glare and raised cutoff. With this example, you don't see that or the spotiness often caused by poor LED's or the anti-glare caps in our housings.

 

[SlowDunny]

The point I've been trying to make is that some LED's really do mimic the exact positioning quite nicely and look absolutely perfect... these shots are from a model unfortunately only available through Asian online markets but is absolutely incredible and word has gotten around... In fact if anything, there is no lack of output, it's that many overpower the reflectors to the point of glare and raised cutoff. With this example, you don't see that or the spotiness often caused by poor LED's or the anti-glare caps in our housings.

You can likely fit lenses to the housings with the right tools. Sourcing the right size lenses with the right diffusion pattern will be the challenge.

A guy in the mazda speed world did this and found it was just too expensive to source the right parts to get a good projection pattern/cut off.
Prices may have changed a bit but this is still the same conclusion. Different light emitters need different light diffusers. /thread

I don't think anyone here has the ability of a OEM to do the R/D to create a cheap/reliable LED projector. If there is, prove me wrong and I will be happy to buy the product.

 

[danbfree]

You can likely fit lenses to the housings with the right tools. Sourcing the right size lenses with the right diffusion pattern will be the challenge.

A guy in the mazda speed world did this and found it was just too expensive to source the right parts to get a good projection pattern/cut off.
Prices may have changed a bit but this is still the same conclusion. Different light emitters need different light diffusers. /thread

I don't think anyone here has the ability of a OEM to do the R/D to create a cheap/reliable LED projector. If there is, prove me wrong and I will be happy to buy the product.

I thought that's why people got the projector housing Euro headlights? LED's are far more forgiving in projector lenses than reflectors it seems anyway, I know some 4 sided mini LED's would work great in them.

 

[XanRules]

Sure but, the problem, as I've stated, is that our eyes are unfortunately very poor judges of how well a light source is working in ways that matter for safety and that without testing these kinds of things I don't think it's worth the risk. We can see "ah, this looks brighter" and that's about it - sometimes you think you can see better until it's too late, and I'd rather go with what I know works until someone can give me empirical proof that something else works better.

Get me documented results, like the ones from CalCoast Labs...

... and some pictures of Spider-man!

I'm gettin' the euro projectors anyway but I'm not stuffin' LEDs in 'em. They look nice and they work better.

 

[danbfree]

Sure but, the problem, as I've stated, is that our eyes are unfortunately very poor judges of how well a light source is working in ways that matter for safety and that without testing these kinds of things I don't think it's worth the risk. We can see "ah, this looks brighter" and that's about it - sometimes you think you can see better until it's too late, and I'd rather go with what I know works until someone can give me empirical proof that something else works better.

Get me documented results, like the ones from CalCoast Labs...

... and some pictures of Spider-man!

I'm gettin' the euro projectors anyway but I'm not stuffin' LEDs in 'em. They look nice and they work better.

Fair enough man... I strongly feel certain LED's have gotten there, especially the very thin tip four sided in mimicking halogen light output in the right place but for ultimate legality you are 100% right.

Sent from my LG-H932 using Tapatalk

 

[XanRules]

Yeah, I'm not even so much worried about legality - technically projector retrofits aren't legal but if someone wanted to throw dedicated LED/HID projectors into their headlights I wouldn't hate. I don't expect everyone to do the type of bugfuck crazy stuff I plan to do* and I don't really care about legality in a lot of ways, I just want everyone to be safe!

And I want pictures of Spider-Man!

* = For example, the fog lights from an SRT8 Cherokee, Challenger, high-trim-model Pacifica, or other higher end FCA cars are very very good, DOT approved for a fog pattern, and the same size as the mostly-decorative Fiesta OEM fogs, I intend to retrofit them if I can find a set. I do not expect other people to do this kind of dumb shit.

 

[danbfree]

Yeah, I'm not even so much worried about legality - technically projector retrofits aren't legal but if someone wanted to throw dedicated LED/HID projectors into their headlights I wouldn't hate. I don't expect everyone to do the type of bugfuck crazy stuff I plan to do* and I don't really care about legality in a lot of ways, I just want everyone to be safe!

And I want pictures of Spider-Man!

* = For example, the fog lights from an SRT8 Cherokee, Challenger, high-trim-model Pacifica, or other higher end FCA cars are very very good, DOT approved for a fog pattern, and the same size as the mostly-decorative Fiesta OEM fogs, I intend to retrofit them if I can find a set. I do not expect other people to do this kind of dumb shit.

Nice, we can agree to that and seriously, I'm not going to settle for crap just to get a nice white color either... I've seen a lot of different fog options, I consider them more like "extra pedestrian warning lights" so the shitty factory ones do OK for now with some random old LED's I had laying around, heh...

 

[KKaWing]

(rant) why can't Canadian cars use the fogs as intended... With the low beams off! We just have glorified accessory / skunk spotting lights. (/rant)

Back on topic. Ft86speedfactory has a projector that attaches to the ft86 fog bracket... Guess who's cars have the same mounting pattern.

 

[danbfree]

(rant) why can't Canadian cars use the fogs as intended... With the low beams off! We just have glorified accessory / skunk spotting lights. (/rant)

Back on topic. Ft86speedfactory has a projector that attaches to the ft86 fog bracket... Guess who's cars have the same mounting pattern.

Same down here man, switch works the same way, have to have low beams on too, might be the law, not sure... but good find on the fog assemblies, at least we know it uses a common standard so we have plenty of options!

 

[XanRules]

Will be updating this thread shortly with some clearer language, particularly regarding the way "signal image" works as I realize this is particularly confusing and not an adequate explanation. In short, our eyes are very sensitive to blue light, which does not cause your pupil to constrict in the same fashion as yellow light. As such, your eyes perceive that there is more light coming from a cooler blue light than a more yellow one, even if the light output is higher from the "warmer" and more yellow light. I will work this into the original post along with some more updates when I have time.

I would also like to announce that I am not a completely intractable grump, as Philips has released an LED array for halogen fog light housings that is approved by all regulatory authorities. I believe the technical data is available by request. I have reached out to them to see what about the optics of fog lights differs in such a way that these bulb assemblies are incompatible with headlight optics, just for the sake of my knowledge - they do specify repeatedly that they are for fog lights only but I am curious as to how that works. My thought is that the bar is much lower for fog lights since most fog lights on modern cars are just for decoration, but it'll be nice to have pros and experts confirm.

FURTHER EDIT: These are technically not approved by regulatory authorities, which gives Philips some plausible deniability in the event of any trouble. However, the consensus from a handful of people who have run tests is that they are quite good in some fog lamp housings, but as is to be expected they do not provide enough light, nor the right type in the right directions, to be suitable in headlights. I have reached out to Philips for the tech sheet and specs regardless and will try to find the threads/etc in which they were tested. Philips' own internal testing DOES show that they are passable, again, in SOME fog housings.

 

[MKVIIST]

With all the up votes in this thread, I’m making this thread a sticky.

 

[XanRules]

Thank you!

 

[XanRules]

Two good recent-ish posts from Virgil at Candlepower Forums, who I believe also has a professional background in automotive lighting, regarding the current state of LEDs - the links, of course, work in the forum but are broken in the quotes:

The basic answer to your question is that legitimate LED retrofits for halogen headlight bulbs are gradually becoming technically feasible, but we're not quite there yet. We're still in the early stages where a few of the LED retrofits work passably-to-reasonably well in certain headlamps, but we're still not to the point where there's an LED retrofit bulb of any type (say, H4 or 9006 or whatever other kind) that dependably works acceptably well in every/any headlamp designed to take that kind of bulb.

(Also, when acceptable retrofits do come into existence, they will not be sold under nonbrands through fly-by-night operators on Alibaba for $25/pair...these are clues...)

The SAE Lighting Committee has a task force working to define and describe the design and function parameters necessary to make a technically acceptable LED retrofit bulb, and the major-name brands are fielding early-version products, unfortunately following the same path of questionable ethics and legality that they've been following on LED retrofits for signal bulbs: not really legal, but there are loopholes in the way the regs are written that they sneak through. First line of products totally unsafe in any application (first Sylvania Zevo LED retrofits, with the plastic dome). Make money off of that, "for interior and off-road use only" [nudge-nudge, wink-wink], then put out a second line of products not really safe in a large number of applications, but they just pretend that problem doesn't exist (see here for example).

There has never been a shortage of bogus claims for illegitimate products. First the maker or vendor makes the claims, then they're picked up and regurgitated -- and elaborated, like a fishing story where the fish gets bigger every time the story is told -- by eager enthusiasts who (1) want reality to be different than it is, and/or (2) want to feel validated in their purchase decisions, and/or (3) want to feel important, like dispensers of wisdom. This is modern "viral" marketing, and it's very effective. Truth and reality have nothing to do with it.

As far as the "retrofits" being discussed in the thread you saw: total nonstarter. There's a guy pushing them hard, claiming to be involved in aerospace engineering so he can't really discuss the details and all kinds of other cloak-and-dagger baloney; Dan Stern forwarded me a conversation he had with that guy, and the guy is just completely full of beans: really does not know even the basics of how headlamps work, but he's sure his guesses and assumptions are good enough. Claims to be an aerospace engineer but rants for paragraph after paragraph and page after page about how he has no use for people who talk about measurements and numbers, because they don't know what they're talking about in the real world (ummm...every real engineer I know spends most of their professional time dealing with measurements and numbers so as to create the things we have in the real world), etc.

He sent Stern samples of the retrofit you saw discussed in that thread. Stern tried them in a bunch of headlamps...with lousy results every time. As he related it to me: "At least the "aerospace engineer" was telling the truth, just not the way he thought: you don't need measurements to tell these are a complete nonstarter." He sent the samples on to me for a second set of eyes on 'em, and...yup, he's right; they're a total joke. There is no evidence of knowledgeable or informed engineering in these at all; there are very basic mistakes in their design and construction, and they don't even physically fit in a fair number of the headlamps they're supposedly designed for (kind of an important gate to pass before we even get to the pickier questions of whether they work). Those headlamps they fit in, they fail in.

Pictures: No. A picture will give a clue as to how bad a beam pattern might be, but won't inform as to how good it might be. Understandable details are described here.

http://www.candlepowerforums.com/vb...etro-fits-getting-there&p=5251057#post5251057

An excerpt from a much larger post:

Maybe in some cases, but that's only two of the many different light source parameters that have to be correct for the headlamp to comply with the law and work in an effective, safe, and legal manner when installed in a headlamp. And even if the emitters on an "LED bulb" are the same size and in the same location as a filament would be, they're still not the same shape or orientation because there are no cylinder-shaped LEDs, and the backer boards and heat sinks create significant shadows where there's supposed to be light.

So that's one big reason why "LED bulbs" fail to comply with Federal Motor Vehicle Safety Standard No. 108.

Also, separately, there is the issue of selling vehicle equipment which, when installed, takes the vehicle out of compliance with any applicable Federal Motor Vehicle Safety Standard. A good example of that is bulbs for lamps other than headlamps: they aren't directly regulated by FMVSS 108 (or any other), but if I sell bulbs that physically fit in the turn signals, brake lights, etc of a vehicle certified as compliant with all applicable FMVSS requirements, and my bulb causes those lights to no longer perform in a compliant manner, I can wind up wearing significant legal liability for it. The same applies to "LED bulbs" for headlamps: even if I magically manage to come up with an LED bulb that complies with all the requirements for one of the halogen bulb types in Part 564, if it changes the performance of a headlamp in ways that make the beam pattern noncompliant, my bulb is still illegal.

http://www.candlepowerforums.com/vb...-LED-Upgrade&p=5251529&viewfull=1#post5251529

 

[XanRules]

Couple of people have asked what I mean when I reference "third-party testing."

Many "testers," such as bulbfacts, measure peak brightness of a bulb or array. There are, in fact, many other factors that need to be taken into account - peak brightness, sure, but also variance of light at different points, light spread, light color, bulb temperature, that kind of thing. All of these are governed by the aforementioned regulations, which, unlike many regulations, are written under consultation of professional automotive lighting engineers. The specific regulations can be found here: https://www.regulations.gov/docketB...=0&dct=N%2BFR%2BPR%2BO%2BPS&D=NHTSA-1998-3397

A third-party test looks something like this one, which was done on an HID kit: http://dastern.torque.net/techdocs/HID/HB1_HID_Retro.pdf

This was done by CalCoast Labs, a photometric testing laboratory that tests lights for compliance with federal safety regulations. It's owned and staffed by physicists, engineers, and other experts in the field. They're the real deal. They used professional photometric testing in a controlled environment to test an HID "kit" for light output, light variance, color, cutoff, etc. In this case, the kit failed because the light was too blue and way too bright.

As LED kits continue to be refined, we will start to see more tests like this being done on those kits. Someday, they may even start to pass. As of now, I have never seen such a test done on an LED kit, and as a result, between the lack of tests and the opinion of automotive lighting engineers, I'm not changing my story regarding their safety or efficacy. I just wanted to give folks an example of what constitutes a satisfactory, professional, third-party test. I intend to keep updating this thread for the benefit of folks who are reading it!

 

[danbfree]

Couple of people have asked what I mean when I reference "third-party testing."

Many "testers," such as bulbfacts, measure peak brightness of a bulb or array. There are, in fact, many other factors that need to be taken into account - peak brightness, sure, but also variance of light at different points, light spread, light color, bulb temperature, that kind of thing. All of these are governed by the aforementioned regulations, which, unlike many regulations, are written under consultation of professional automotive lighting engineers. The specific regulations can be found here: https://www.regulations.gov/docketB...=0&dct=N%2BFR%2BPR%2BO%2BPS&D=NHTSA-1998-3397

A third-party test looks something like this one, which was done on an HID kit: http://dastern.torque.net/techdocs/HID/HB1_HID_Retro.pdf

This was done by CalCoast Labs, a photometric testing laboratory that tests lights for compliance with federal safety regulations. It's owned and staffed by physicists, engineers, and other experts in the field. They're the real deal. They used professional photometric testing in a controlled environment to test an HID "kit" for light output, light variance, color, cutoff, etc. In this case, the kit failed because the light was too blue and way too bright.

As LED kits continue to be refined, we will start to see more tests like this being done on those kits. Someday, they may even start to pass. As of now, I have never seen such a test done on an LED kit, and as a result, between the lack of tests and the opinion of automotive lighting engineers, I'm not changing my story regarding their safety or efficacy. I just wanted to give folks an example of what constitutes a satisfactory, professional, third-party test. I intend to keep updating this thread for the benefit of folks who are reading it!

Thanks for the updates, I'm with you on hoping for real 3rd party tests on LED's soon... BulbFacts does test brightness, beam pattern/cutoff and color temp but missing the critical projection distance, among more technical items you mention. I think *some* sets I'm sure could pass on *some* reflector housings, but it has to be the right combo... but like other gov't agencies like the FDA, it's likely that DOT doesn't care to spend time or effort or would charge FAR too much for one-off individual testing like your example docs... do you know if there are newer versions out there published? I wonder how much that testing costs, and again, because it's a per housing model too, it would be way too much to get full certification probably. Also to consider, are possibly the tests TOO stringent to test? Sometimes they go heavy handed out of an abundance of caution, but then see some very new cars with *factory DOT approved* LED's and HID's that seem to have a LOT of glare, and have even seen halogen's with more glare than their LED counterparts in the same housing, although a bit of glare does help with lighting up signs up ahead too, so what standards are truly best? Hmmm...

Anyway, we'll have to wait and see and from where I stand if the basics of proper pattern, cutoff, brightness and projection distance all in basic testing match up, then to me I feel safer with those kinds of LED's than the multiple halogens I've tried. I know we're not going to agree, just food for thought... I'm sure one day, hopefully soon, we'll find our common ground point.

 

[XanRules]

Thanks for the updates, I'm with you on hoping for real 3rd party tests on LED's soon... BulbFacts does test brightness, beam pattern/cutoff and color temp but missing the critical projection distance, among more technical items you mention. I think *some* sets I'm sure could pass on *some* reflector housings, but it has to be the right combo... but like other gov't agencies like the FDA, it's likely that DOT doesn't care to spend time or effort or would charge FAR too much for one-off individual testing like your example docs... do you know if there are newer versions out there published? I wonder how much that testing costs, and again, because it's a per housing model too, it would be way too much to get full certification probably. Also to consider, are possibly the tests TOO stringent to test? Sometimes they go heavy handed out of an abundance of caution, but then see some very new cars with *factory DOT approved* LED's and HID's that seem to have a LOT of glare, and have even seen halogen's with more glare than their LED counterparts in the same housing, although a bit of glare does help with lighting up signs up ahead too, so what standards are truly best? Hmmm...

Anyway, we'll have to wait and see and from where I stand if the basics of proper pattern, cutoff, brightness and projection distance all in basic testing match up, then to me I feel safer with those kinds of LED's than the multiple halogens I've tried. I know we're not going to agree, just food for thought... I'm sure one day, hopefully soon, we'll find our common ground point.

You do touch on an interesting issue, which is that there are some LEDs that perform acceptably in some housings - thus showing the importance of optics and how preposterously complex this stuff really is. I think the big hangup is that nobody has yet been able to engineer one that will reliably pass tests in all housings. Even the Philips Fog Light LEDs had that issue, which was "well, these are great in some fog housings, but not enough that we could really sell these. I agree that a lot of current-gen headlights seem to be laser beams pointed into the eyes of oncoming traffic, but i attribute this more to user error (improper aiming or lifting a big-ass truck so that the lights point right at eye level). I am sure there are newer tests than that one, and have likely even been some done on LEDs, I will dig around and see if I can find some if I have time this weekend.

 

[danbfree]

Just wanted to say that I have Depo Euro replicas arriving tomorrow! I'm the one who started the LED thread to encourage people to at least use the very best and cleanest output available but I wasn't satisfied personally... Luckily, I accidentally found a sponsor, the owner of BulbFacts.com, one of the only independent testers of all types of bulbs in both reflector and projector housings. Anyway, he helped with a significant portion for me to order Depo replicas for $350 shipped. They will have taken only 9 days total to arrive from Latvia, which isn't bad, but did take 2 weeks to ship after they ended up needing $17 more for shipping after paying, but I got decent service from Cars245.com. So, they take an H7 bulb so I got an HID kit for them. Surprisingly they say the projectors are for LED, which is strange but HID rules for sure. I went with a very solid budget kit, the DDM Tuning Premium for only $60 but tested out with excellent results. I will likely upgrade the projectors later, so we will see how my ~$410 total investment works out soon! The Depo, compared to the TYC or OEM Euro options do have the LED DRL already wired for Low to be on at all times, but I will also likely do the harness to make true bright for day and dim for at night marker lights, the amber marker in front goes from a combo blinker/marker light to just blinker.

 

[danbfree]

Just a quick update, so those can see what you get when you upgrade to the Euro projectors. total investment is $400 including the pair of actually excellent quality Depo aftermarket Euro projectors shipped and after TONS of research, Katana LED's in the Euro H7 format that are only $47. VERY happy with the perfect clean cutoff and the light spread is fantastic for LED's. The Katana's have proven to work amazing in projectors actually originally intended for halogen, but with projectors so much more forgiving to LED's, ones that are designed well are indeed perfect! Even more amazing, the Katana's are clearly a ripoff of the SuperNova V.4 LED's that are $169 but the Katana's are actually brighter in a proper projector! I tried HID's originally and the glare was insane, you have to at least use bulbs attempting to mimic at least the proper filament length of a halogen bulb AND proven to have clean light spread. This was with the bulbs brand new, they are very clearly more white vs. blue now, thankfully, I need to get a new pic but it could be the cell camera even in basic no enhancement mode seems to add blue.. Only downside is with a 100% clean sharp cutoff I did have to watch my vertical aim carefully to get proper distance but not too high as to glare oncoming traffic as they enter the projection area. I need to bring both very slightly horizontally to the right as well for the ECE cutoff to only be on the the side of the road and not back right of cars in front of me.