3rd Gen HID vs LED vs Halogen H11 projector headlights

Power rating 18~22.5W is because it starts at around 18W and stable at around 22.5W (1.3A @13.8V start up = 17.94W) that's the reason for wide power rating range in spec.

One important message from me is,

No, in Koito H11 projector being used in Tacoma, really not compatible. And do not recommend for projector use. it's just not beneficial, not taking advantage of its design.

I gave explanation and Crash stated above. In projector optical design principals, hotspots are generally created at rear end of near bulb insert area. 12 o'clock zone is responsible for beam center intensity ( includes foreground center zone)
Because chip size and position is so tightly configured, there will be no unintentional beam spread that can mask off nature of bi-directional LED


Due to design principal difference between projector and reflector, there really isn't way to make projector user happy. Bi-directional LED configuration never can gain proper efficiency in projector. Some less refined larger chip bulb will show much smoother and higher volume output in projector, but this is because there is lot of unintentional accidental beam spread conveniently cover missing illuminance area.
4.0 has least amount of unintentional beam spread = as a result, in projector, it become too lean not be able to mask missing illuminance zone.

Since projector optical principal is not compatible with bi-directional LED config to begin with, my focus was spent for as proper as possible in reflector housing.
When paired with reflectors, 4.0 will show strongly tight hotspot with least amount of glare. In some low beam reflector shows near compliance level of glare control, and not adding excessive foreground while substantially increase hotspot intensity

Could I make LED bulb that works well in both projector and reflector? No, not to my standard. To give volume and smoothness in projector, it require intentional over sized chip, outside of designed optical beam spread. But doing so result in huge glare issue and weaker hotspot in reflector housing.
reflector has much more positive compatibility with 4.0. Then my decision was to focus on optics that can take advantage of tech and design principal.


Therefore,, It must be odd message from designer, but since I am with most information about 4.0 bulb, I am advising all here, 4.0 is not recommended in Projector usage. It's just not beneficial, type of optics isn't taking advantage of design.
But do enjoy in high beam reflector. I have been using it in high beam nearly a year, very enjoyable high beam performance.



I am checking reply time to time, feel free to ask me any questions, concerns

 

Yes they will do much better in the high beam reflectors than the low beam projectors. The 3.0s may provide initial higher output for short durations in the high beam, whereas the 4.0s ramp up as they warm up, but the 4.0s will have better optical focus than the 3.0s. The 4.0s do resonate in the 3rd Gen high beam reflectors as well whereas the 3.0s do not, I tested the 4.0s in the high beam location in trying to determine the resonating issue.

 

I appreciate the insights in going over the unexpected results.

 

I was going to pose a few questions and ask for speculation from Crash, but Yoshi I's response basically covered them all. Thank you both!

Crash do plan to do a highbeam comparison between the H9 and the 3.0 and 4.0 bulbs?

 

Once again, thank you @crashnburn80 for the testing. Surely saved a bunch of us from buying them!

 

H9 produces more light that either 3.0 or 4.0 bulbs. I would think H9 would be better than 3.0 and 4.0 in high beam application.

 

I updated the 4.0 review with a halogen pattern wall shot comparison which I had forgot to provide in showing the difference vs pnp LEDs.

I do not. PnP LED distance testing is a lot of work. The Philips H9 at runtime voltage should put out more lumens, have better optical focus and superior omnidirectional light output for use in a halogen reflector. If wanting 6000k LED color temp, the 3.0s should offer higher instant-on output while the 4.0s will provide better focus for download projection and higher output intensity once they warm up.

 

There is a new player in the game from Diode. Their new SL2 pro are supposed to have improved focus and clockable mounting collar so allegedly one could address the 12/6 oclock vs 3/6 design.

https://www.diodedynamics.com/h11-sl2-pro-led-bulbs-pair.html

 

Wow, thanks crashnburn for testing and Yoshi for the informative post. I'm not a Tacoma owner at all. I own an Elantra but my halogen projector headlights suck (rated poor by IIHS). When I was looking at plug and play LED options and pros and cons, I came across this forum, and have nerded out on crashnburn's posts since then. Your posts prevented me from going the pnp LED route and I ended up upgrading to HIR halogen bulbs instead.

Yoshi's post above brings an interesting question. So, thinner filament LEDs (like 4.0s) are better for reflectors and that makes sense. However, for projectors, is my interpretation correct that thicker filament junk highly rated pnps on Amazon (Ex: Hikari, Fahren, Sealight etc) tend to do better? Since projectors control glare by design, I'm assuming that they won't cause glare to oncoming drivers, given that the projectors are aimed properly. From Yoshi's post, it looks like the unintentional beam spread leads to a better beam pattern / hotpsot in projectors?

@crashnburn80 @Yoshi I

 

No, it won't create better beam pattern. But it coincidentally may mask of unintentional beam distribution unintentionally Very hard to explain..
@crashnburn80 If this type of optical study subject have better thread, feel free to redirect this.

I don't have exact Koito H11 projector model, but I got Stanley H11, principal is similar, so let me use it to explain.

In long story short, larger chip LED, and larger interval between emission surfaces bulb in projector create "loose" less focused beam split in right half and left half. Since it is loose and scatter some, it accidentally overlap in the middle. Result in higher center intensity by some of outside of intentional beam to be overlapped.


From top, H11 halogen - 2st4.0 and SV4 as example. SV4 has bigger chip + also emission interval is much wider than 2st4.0
What's happening here is
Green box width shows filament diameter (1.4mm) projection region.
Yellow arrow shows where each side of LED emission surface image starts.

2st4.0 start beam pattern nearly just very slightly outside of 1.4mm filament edge zone, and triangle shade is "missing illuminance" 12 o'clock zone.
Then look at SV4 ( bottom images), each side of beam pattern is offset ( this is due to LED interval is too wide) not creating intentional beam, starting much further out compare to H11 halogen or 2st4.0
But, this conveniently going to cover "missing" triangle accidentally. Larger chip size also aid beam to be loose=larger, help incidental overlap amount

So for the goal of having less missing triangle ( typically seen void at beam center bottom in many LED bulb used in projector), it is "convenient" for bulb to have NOT too tight intervals in between emission surface. This help make it look like beam is filled better.
But never been intentionally and correctly filling.


Here is right half and left half contrasting color assigned view.




Therefore, more accidental overlap is conveniently covering each side of beam pattern's missing zone.
And often, overlap also aid beam to have higher chance of gaining higher intensity stack at beam center, or larger/wider hotspot in projector.

However!!!! The problem is chip size any bigger, or interval any larger than filament 100% result on glare in reflector application, and unintentional foreground intensity increase in projector housing.
Make incidentally obtained higher peak useless by also adding foreground value.

In here, this was my past study 2st3 vs SV4 vs Halogen Look at bottom image, that is actual photo record of Stanley H11 behavior

peak intensity point also move way far down, adds foreground
SV4 uses pretty tight sized LED, is 4.6mm x 1.49mm Even that size, it still shift hotspot this much, add foreground this much due to interval of emission also much larger than 2st



Hope this explained why tighter tolerance LED configuration result in much lean, not filling beam in projector. It simply is nature of optical principal.

To me, not much motive to create LED bulb do well in projector, because it still can't gain meaningful intensity boost ( the reason I said meaningful is because it also will add foreground)
reflector has much more compatibility and design principal that can take advantage of tight tolerance emission.

for LED bulb to avoid glare in reflector low beam, then it is absolutely critical to have as accurate and tight as possible emission config.
Reflector uses far outer segment to create hotspot, direct above segments for foreground.

LED bulb naturally has lambertian peak strength assigned to far segment = stronger hot spot ( it can be stronger than halogen, because unlike filament, LED has directional power distribution curve difference)
Basically, selectively focusing available lumen into hotspot creating surface. Higher the lumen density chip has, exponentially higher hotspot intensity in general
then, 12 and 6 oclock is not available in LED projector, which is responsible for foreground, tend to shoot less volume to foreground and this can be convenient for some application.
However, once again, any emission size bigger than filament directly result in severe glare issue in reflector low beam.
This is the exact reason LED bulb had been told, never use PnP in reflector, because it never had emission size as tight as filament, simple physics.

 

Thanks for the detailed explanation @Yoshi I . I clearly understand why thicker filament bulbs do not work well in reflectors (mainly uncontrolled glare) and why thinner LEDs do better.

For Projectors, I'm summarizing my understanding below. Let me know if that doesn't make sense.
- Thicker and brighter filament LEDs lead to a more 'filled' beam pattern in projectors due to unintentional beam distribution. This usually creates a bigger and brighter hot spot and the unintentional spread also covers the dark spots generated by the 2 sided led bulbs (unlike 360 degree halogen)
- However, while the beam pattern looks more filled, it also creates a lot of foreground illumination which results in less usable light and makes vision worse.

I was wondering why companies do not create separate thicker filament projector specific LED bulbs and thinner filament (like 4.0s) reflector specific LED bulbs. While thinner filament LED pnps for reflectors might still be a possibility, your explanation has helped me understand why thicker LEDs are not necessarily a good fit for projectors. Thank you

 

Your understanding is generally correct
Some missing aspect is brighter LED definition. It is tricky term, but can’t forget about total lumen vs lumen density. When it comes to intentional hotspot creation, both projector or reflector, higher lumen density light source create much stronger hotspot

the reason manufacture don’t make projector specific LED bulb or reflector specific LED bulb, I think vast majority isn’t even comprehending optical principals difference. They think more lumen is better. And sadly, there are so many unaware customer buying those bulbs all over, then why they would invest more money to target narrower split applications market. And from my point, I concluded there is no realistic benefit to make LED bulb that can be compatible with projector designed to use filament ( point light source ) lack of 12 and 6 o’clock illuminance is destiny of weak hotspot Poor upper illumination. Projector really needs full 360 degree emission ( frontal view) and also back shoot towards center of reflector bowl. Projector by nature much smaller than reflector, light source position discrepancy magnify

using larger powerful chip, some more interval in between still won’t make LED bulb useful in projector due it excessive foreground consequences

and then that bulb for sure not be useful in reflector due to severe glare issue

make it tight tolerance bulb that maybe able to take advantage in reflector is logical conclusion
Basically I surrender to design LED bulb that I good for projector. Because it will be at the best, mediocre, or worse by excessive foreground weak upper illumination, unbalanced beam no matter what
Making effort and spend moneys on system that is fundamentally not compatible with really won’t make sense
Especially, halogen bulb is well established. Cheap and work as intended !

In reflector, I must compose tons more contents to explain potential but if someone interested in such academical subject, reach me directly or have someone create appropriate study thread. I share info as requested as much as I can

 

Awesome work @crashnburn80 !!!! great stuff and your work on this thread (and others) is extremely appreciated!!!! thank you @Yoshi I as well!!

great info for us all to be smart and safe out there!!!

 

This is hands down one of my favorite threads to reference and send friends who swear by LED bulbs in reflectors or projector housings. I still can't believe my $16 H9s outperform a crazy expensive (comparatively) set of LED bulbs at least in the taco housing. I know the halogens will probably not last as long but I have a spare box in the truck if they fail. Thanks for all the knowledge over the years @crashnburn80.

 

Yoshi, please continue.

 

Here is reflector optics and LED bulb study topic.

This is generic low beam reflector sample I am using here, does not always validate all aspects identical way, but you can grab good amount of generalized behavior character.
Hope this help many to make wise judgement what to try!

First of all, halogen filament is very compact 1.4mm diameter, length about 4.5mm~5.0mm ish cylindrical point light source. Once of the reason LED never can replicate filament emission behavior is because LED is directional surface light source.
To make LED work as close as possible to replicate filament, a lot of effort had made to make "identical" size chip, place at " identical position.
However, this still does NOT mimic halogen emission character. Closer, but some pro and cons in different balance.



In this study, I am using H11 halogen reflector low beam sample to simulate how each LED bulb character affects beam in what way.


1. How chip size affect beam quality

LED chip size is CRITICAL parameter how well LED bulb regulate glare level. In short, larger the chips, exponentially higher probability of causing severe glare.

Base LED bulb design for this study : 1.0mm PCB
LED thickness 0.8mm each
Total interval 1.0+0.8 x 2 = 2.6mm
This is pretty common emission surface intervals for standard range of LED bulbs nowadays. Some are tighter, some are thicker, but this help to focus on chip size based behavior difference.

3 samples are prepped
1.4mm x 4.5mm ( identical to H11 halogen filament's side profile size)
2.0mm x 5.0mm ( larger chip, some higher power LED bulb uses this range of emission size chip)
3.4mm x 5.3mm ( out of question type, something like 6-die chip, 12,000lm!!!! type of chip)

All 3 are se tot have 700lm per emission surface (total 1400lm) output
base reference, H11 bulb also simulated at 1400lm



Results were probably easily guessed.
Larger the chip, more glare, less focused hot spot. Blurry cutoff, is our educated logical guess.

Below are typical photo being posted online of " before after of LED bulb upgrade!!"

Did LED bulb retain cutoff ? replicated beam reasonably ?




This is huge beam quality discrepancy. Seeing visible brighter and dark edge does not define " cutoff" Some larger efficient reflector may shows rather distinct illuminance border, but do not mistake this as "compatible with LED bulb"
This is image of "out of question" type super large chip LED model in comparison. 3.4mm x 5.3mm ( out of question type, something like 6-die chip, 12,000lm!!!! type of chip)

Some may say, you can aim down, it still has cutoff. No that is not acceptable.
By doing so, it will cause massive foreground value increase, making low beam output absolutely useless.


Here is how chip size affected cutoff and beam proportion
https://imgur.com/CQYgXZN




Large the chip, more glare, and also less output focus. Loosing intensity while adding glare is most stupid configuration. Of course, they can add more lumen to advertise flashy numbers but that only makes glare far worse.
This is explained by lumen density below.
1.4x4.5x2=12.6mm^2 1400lm (111lm/mm^2)
2.0x5.0x2=20.0mm^2 1400lm (70 lm/mm^2)
3.4x5.3x2=36.0mm^2 1400lm (39 lm/mm^2

This case is showing, even moderately smaller size chip at 2.0mm x 5.0mm is already showing glare.
At filament side profile chip size, while it still within glare limit allowance, value is elevating. On the positive aspect, without hitting glare allowance limit, it increased hotpot intensity considerably higher than halogen.

This is one of reflector optics and LED bulb's unique relationship. This indicate, when LED bulb is designed with strongly tight tolerance, even at identical total available flux.





2. How intervals of emission surfaces affects?

Now, we learned (or just re-confirmed) how critical is to have emission size as close as possible to filament. Next we observe using "identical to filament's side profile view size" chip
1.4mm x 4.5mm model, but at different intervals




First, we will look at How well it replicate if chip size is same as filament profile and, interval at filament thickness





Pretty close? As I explained above, LED bulb in reflector tend to show less foreground with higher hot spot intensity

https://imgur.com/AjDlqnl


In this particular reflector, from identical total flux of 1400lm, LED bulb that has interval of 1.4mm demonstrated about 150% of peak intensity at hotspot, in exchange of reduction of foreground flood volume.



https://imgur.com/9scoB5X

This difference maybe difficult to observe in low res gif image, please go download link and see in full screen suggested.
All of sample did NOT exceed glare allowance limit on two test points while gaining higher peak hotspot intensity.
Main difference is focus efficiency, and some affect on glare region value. It is not glare issue at 1400lm config, but this data also is showing if bulb has more than 1772lm total flux, even at 1.4mm interval, it will start showing glare.



3. Emission surface position accuracy influence

So far, we learned it is critical to have chip size as close as filament, interval as close as filament diameter is the key to take advantage of reflector optics unique pro and cons relationship with LED bulb.
However, one another aspect we can't forget is

There always will be manufacturing tolerance, installation tolerance. especially many LED bulb now has "clock-able" rotating bulb base, this can give undesired emission position misalignment.
Soldering emitters to PCB also will have reflow solder position tolerance.
Such tolerance stack can be easily added up to +/-0.2mm

And lower section is defined as "glare zone"



Emission must be contained within tolerance box, high intensity zone must be well contained within B This is very important because any emission expand into glare zone(right below tolerance box lower line)

In this case study, filament equivalent sized chip, but intentionally misalign lower by 0.2mm to see how badly it affects to glare
Interval is set at commonly found 2.6mm in this case to highlight misalignment influence.
https://imgur.com/iyhCKEd


https://imgur.com/brWwA3g



If emission surface is NOT accurately contained within tolerance box, and especially misaligned to lower than box, 0.2mm Y axis discrepancy was enough to start glare in this optics.


Conclusion of study

* Chip interval mainly affect focus efficiency, large the interval, lose hotspot. While it is desired to have minimum intervals, since reflector optics are generally larger, it is more forgiving to emission surface intervals.
• Chip position accuracy is critical. 0.2mm to lower position caused glare even with chip size equivalent to filament side profile size of 1.4x4.5mm. 0.2mm tolerance can be very much possible with standard soldering position tolerance, bulb base to bulb shaft coordination accuracy, bulb base centering accuracy
• Chip size had even larger influence to glare handing. Chip size at 2.0mm x 5.0mm already showing above maximum allowance of glare limit at 0.5U 1.5L-L and 1.0U 1.5L-L ( those two test points are evaluation point for oncoming traffic driver eye input)
• Considering tolerance stack of soldering accuracy, bulb base centering accuracy/limitations, even at chip size of 1.4mm x4.5mm( equivalent to filament side view profile size) chip still carry certain amount of glare risk.



To take advantage of reflector optics without raising risk of glare, LED bulb ideally shall have chip size further compact within tolerance box B
This red 3 boxes are highest luminosity image of filament emission. This shall be target chip size and better glare control tolerance box.



Even at this level of effort, there will be some optics that aren't compatible with LED bulb with bi-directional emitter layout.
Application based positive list data collection is still essential process to evaluate compatibility, but now at least we know what kind of precision, configuration to look for.

 

Another excellent post, Yoshi! Thank you for the detailed discussion and example photos. Despite their low resolution, you still clearly illustrate how seemingly insignificant physical parameters can have a huge performance impact, especially regarding glare!

All too often it seems people want their low beam UpgraydeZ to result in >picrelated for oncoming drivers, especially when their is no emphasis on checking and adjusting headlight aim after any such modification.

 

About what you said. If the rear end of the projector is responsible for the hotspot, does the bulb cap of the h11 improve the hotspot over the h9 which doesn't have a cap? I mean, does that cap reflect light to the back of the projector to help the hotspot or just simply blocks the light going forward?

 

H11 bulb top shade is purely there as absorber. It's to prevent forward direction leak. Reflector low beam or projector, directly escaping ray will not get any control by reflector = scatter as glare.
High beam reflector does not need bulb top shading, because scatter is acceptable per application intention.

Therefore, use H9 bulb in H11 intended projector can cause glare or increase risk of glare certainly.
In H11 low beam projector with glare shield cap, it may handle glare better than projector depend on design at least for direct leaking ray based glare.

And it is not possible to recycle undesired forward ray turn into backshoot.

 

Thank you @crashnburn80 for the testing and write up. The H9 Philips bulbs were an easy swap by pulling the green inner piece on my headlight plugs.