Ideal road beam pattern? Here!- Mtbr.com

1. ## Ideal road beam pattern? Here!

Hello all!

I wondered what would an ideal road beam pattern look like. And I designed one (several actually)! (not the reflector of course!)

The ideal pattern should have constant head-on illuminance (lux) in the center "lane", and gradually decrease to sides, like shown schematically here:

Note the 4 parameters that determine the geometry:
dn1; dn2; df1; df2; (df=fixed=10m; all distances in meters; 1m~3feet)

Few simple algebraic equations (no calculus or integrals!), like from here(*), and here come the results. Proportional schematic diagram (to better understand beam geometry), and corresponding "wall shot". Isolines are all sqrt(2) from adjacent ones.

Very-very tight beam, with fast falloff (tunnel): dn1=0.5; dn2=0.1; df1=0.5; df2=0.1;

Wider, but also fast falloff: dn1=1; dn2=0.1; df1=1; df2=0.1;

Same wide, but smoother falloff (less tunnel): dn1=1; dn2=0.3; df1=1; df2=0.5;

Same falloff, but the area of const illum is wider: dn1=1; dn2=0.3; df1=2; df2=0.5;

Finally, same wide, and extra-slow falloff: dn1=1; dn2=0.3; df1=2; df2=1;

Now, you can "select" your ideal road illumination, and see its corresponding wall pattern. Later, when you see a wall shot on the internet or in real, you can immediately say if it's something close to your dream, or very different from it.

The octave script that does it (anybody can run!):

Code:
```h=1.05; ez=10; % won't be needed, but left as is
b0=atan(h/ez);
[as,bs]=meshgrid((-40:40)*pi/180, (-3:40)*pi/180);
dn1=1; dn2=0.3; df=10; df1=2; df2=1;
b=h*sqrt(1+1./(tan(b0+bs).^2)).*tan(as);
e=h./tan(b0+bs);
de=dn2+(df2-dn2)*e/df;
db=dn1+(df1-dn1)*e/df;
M=2.^(-(abs(b)-db)./de); % the fall-off multipler
ind1=find(abs(b) < db);
M(ind1)=1; % the central lane with const illuminance
lint=7*M.*(1+tan(as).^2).*(1+1./(tan(b0+bs).^2))/(1+(ez/h)^2);
% luminous intensity!
% "7" is simply to match my colorbar

% that's it! the rest is visualization

decade=[1 1.5 2 3 5 7];
cl=[];
for ii=-3:1
end
col_lims_myW=[0.03 30];

pcolor(as*180/pi,bs*180/pi,log(lint));
hold on;
title('Luminous intensity');
xlabel('alpha, deg.')
ylabel('beta, deg.')
axis('ij');

zimax=max(max(lint));
isl=isolines_sqrt2(zimax,cl(1));
contour(as*180/pi,bs*180/pi,log(lint),log(isl),'k');
axis equal
caxis(log(col_lims_myW));
colorbar('YTick',log(cl),'YTickLabel',cl);

grid minor on
ylim([-5,40]);
```
(*) Need to fix few mistakes there, like interpret E (lux is wrong) as I (cd is correct) in (3.116), or exchange e <-> h and -beta -> beta in (3.119), but the final plot, Bild 3.145 is correct.

2. Curious as to why you loose so much intensity at such a short distance. The rest though not exactly the same numbers pretty much do the rapid loss of intensity as it spreads wider. But especially on road you want more light out directly in front. The thermal imaging type pic looks like a optimized flood optic (assuming your aiming for a cut off type beam).
I'm just going off of what I'm making sense of there. But seems almost the opposite of what majority is looking for in a road light. Needs more clarification

Sent from my SM-G950U using Tapatalk

3. Making a bunch of cool pics is great, but how do you, or are you going to, bring the design to reality? If you aren't or won't then it seems to me to be a waste of time.

I'm also with RAKC on wondering why there is so much light right at the wheel and losing so much at a distance. As far as I prefer, the area out to 6-10 feet in front of the tire could be nearly dark. From there, out to distance, is what I want lit up.

4. While I'm trying to understand what others don't understand... Fully aware of the waste, and without any hope... Here's an update for those big rocket scientists who do understand, but keep silence.

The argument to the exponent (2^...) has a kink: zero const, then a tilted line. Not so bad, but better smooth it. Introduce 5th parameter, r, the size at which one segment blends into the other, with a piece of parabola:

y=-(x-(b-r))^2/(4e*r).

If r is taken 0.5 in units of e (rather strong smoothing), the last (5th) pic becomes:

Code:
```...
db=dn1+(df1-dn1)*e/df;

rv=0.5;
M=zeros(size(e));
ind1=find((abs(b) > (db-de*rv)) & (abs(b) < (db+de*rv)));
M(ind1)=-(abs(b(ind1))-(db(ind1)-de(ind1)*rv)).^2./(4*de(ind1).^2*rv);
ind1=find(abs(b) >= (db+de*rv));
M(ind1)=-(abs(b(ind1))-db(ind1))./de(ind1);
M=2.^M;

lint=...```

PS.
And here's Philips Saferide 80 for comparison (pretty good light actually; its beam width roughly half of the ideal above; clicking will open in good quality):

Even Saferide doesn't come even close to the proper concentration of light in the hot line. I wish modern tiny-row-LEDs headlights be closer to that proper concentration, and have properly adjusted near field.

5. ## Graphs

Originally Posted by Vancbiker
Making a bunch of cool pics is great, but how do you, or are you going to, bring the design to reality? If you aren't or won't then it seems to me to be a waste of time.

I'm also with RAKC on wondering why there is so much light right at the wheel and losing so much at a distance. As far as I prefer, the area out to 6-10 feet in front of the tire could be nearly dark. From there, out to distance, is what I want lit up.

I believe the multi-color graphs are showing beam angles as that is what the axis labels are. Not distance in meters for those graphs. Attached here is a SafeRide aerial view color distance graph, wall beam photo of the SafeRide, and abvgdee's multi-color degree angle vertical wall graph.

The SafeRide doesn't have a very wide horizontal spread. It has a far reaching throw out front, with equalized illuminance per distance, as the light beam strikes the ground. If you look at the distance graph you will see the SafeRide has very good lighting (yellow color) out to 40 meters. That is the distance I can easily distinguish between objects such as a small brown paper bag versus a rabbit. And that is on the 1.2 watt low setting riding on grayish/black asphalt. Efficiency of the controlled beam pattern and high performance Task LED driver makes a modified SafeRide an exceptional road light for straighter flat roads. But as abvgdee comments, it is a fairly narrow beam nearfield.
The OL Road version has about twice the usable visual spread in the nearfield which mup and twisty road riders would appreciate more, as well as a much higher lumen output.

6. Don't you wish it were possibly to just be able to dial in "the perfect road beam pattern" with whatever lamp you were using? My ideal beam pattern is pretty much what the OB is doing although I could ask for a bit more throw. You wouldn't want too dim a light close to the bike because then you might not see something you really want to see.

7. Originally Posted by Cat-man-do
Don't you wish it were possibly to just be able to dial in "the perfect road beam pattern" with whatever lamp you were using? My ideal beam pattern is pretty much what the OB is doing although I could ask for a bit more throw. You wouldn't want too dim a light close to the bike because then you might not see something you really want to see.
If someone created a lamp with a perfectly adjustable beam it would certainly be revolutionary.
I have to agree with Vancbiker. I don't want much light in the first 10-15 feet or so since I am always looking way further ahead. But I also have only ridden on a mup once in my life. I usually ride open fairly flat roads at 18-20 mph. That is 26-29 feet per second flying by. If you saw something at 10 feet there wouldn't be enough time to react. Like when a very fast raccoon blasted out of some bushes 10 feet in front of me one night in Michigan and I hit him.

8. I've always thought that the Saferide had an intriguing beam. I've wanted to get a reflector from one to build a light with for years but did not want to spend that much money for a light just to get the reflector. I should keep an eye out on ebay to see if I can get one at a cheap price.

9. Oh, yes, gipsyman, thank you! Right — those are angles. The colored pics are as wall shots, but colored (in log scale) for better evaluation.

The illumination is even (const) in my ideal patterns at all distances, close to wheel, and far away (in the "central lane"). This condition (same illuminance) defines the luminous intensity (cd) at various angles. And this distribution, how much light intensity should be sent at such and such angle — is shown on those colored pics.

I could write what the "frame" diagrams (made of straight green lines) are, but I already explained it on the very first image in the OP. Please see it if you skipped.

Originally Posted by RAKC Ind
Curious as to why you loose so much intensity at such a short distance.
If you mean the very first picture in OP (the straight green lines frame) — yes, that's actually what we want, this fast (called exponential) decay. Because our eyes see logarithmically.

This is why, if some headlight makes brightness linear, like 25->50->75->100% of power, then the 25->50 change will be visibly noticable (but not very much!), but the 75->100 is almost same. Should be like 12->25->50->100, or even better, like 4->11->33->100.

The 3 "exposure value stops" = the 8=2^3 times of intensity difference — is what would our eyes perceive as more or less uniformly-lit area.

The exponential decay on the "straight green lines" sketch continues further to the sides (not indicated), with same "strength". This decay strength is shown as slope on the parabola sketch (post #4; again, right side of the road shown): zero slope is no decay = illuminance is const, const slope of the line = const-decay strength = exponential (the correct for our eyes) decay.

Originally Posted by Cat-man-do
You wouldn't want too dim a light close to the bike because then you might not see something you really want to see.
And the near field will cost you close to nothing in terms of lumens. Expensive (in lumens) will be the beam width far away (the hotspot/line area).

Originally Posted by Cat-man-do
...although I could ask for a bit more throw
The tightly-packed red zone (lines) is actually this cutoff-characteristic throw. Typical circular torch pattern has its very-hot-hotspot (the area enclosed by first isoline) diameter ~5 degrees. Here the corresponding vertical size ~0.5 degrees, much tighter! This is the throw, to pack lumens into such a small angle!

———

I very much welcome those folks who understand both me and others (who don't understand) — please, stop by, explain something. Again, gipsyman, thanks!

10. Originally Posted by Vancbiker
...it seems to me to be a waste of time.
Actually, this I think is great — when people start dreaming:
Originally Posted by Cat-man-do
Don't you wish it were possibly to just be able to dial in "the perfect road beam pattern" with whatever lamp you were using?
Today dreams, tomorrow expectations, then demands, then shipping products!

11. Just remember, it's easy to figure out the ideal" beam pattern.

The real hard part is balancing the losses (reflection, transmission, thermal), the size, the thermal, the mass, the power consumption, etc. That's when it gets real hard.

You end up with guys who want to see 200 yards away, but want 80* wide beam patterns, 2500 lumens, weigh less than 100g, use a single or dual battery, stay cool to the touch, last for 3 hours, oh and cost less than \$100.

It's very hard to overcome reality sometimes.

12. Originally Posted by Outbound
Just remember, it's easy to figure out the ideal" beam pattern.

The real hard part is balancing the losses (reflection, transmission, thermal), the size, the thermal, the mass, the power consumption, etc. That's when it gets real hard.

You end up with guys who want to see 200 yards away, but want 80* wide beam patterns, 2500 lumens, weigh less than 100g, use a single or dual battery, stay cool to the touch, last for 3 hours, oh and cost less than \$100.

It's very hard to overcome reality sometimes.
Ummm....I hope you're not quoting me although I did say that. That was a typing error as I meant to say, "feet". I did go back to correct that btw. As to what you said about the other stuff, I can't help but hear the Rolling Stone song , "You can't always get what you want", playing in the back of my mind. Always going to be more people wanting something that really isn't either possible or worth making due to market demand.

With all this in mind I think you've gotten a lot of feedback here about what it is that people really want. I think the lamp you designed makes excellent headway as far as providing a very usable road lamp system. The only other thing I myself could want is something to provide the occasional "High beam" for when needed. It's my opinion that this might be better accomplished using a separate lamp and have it mounted on the helmet. The perfect setup would be something very light-weight but yet able to throw about 300 feet. I like using something like a torch because it's so lightweight and can throw a very tight spot ( run time not being an issue because only used for moments at a time ). My current favorite is the Convoy M1. The only additional thing I could wish for to make the M1 more useful would be the ability to turn it on/off using a wireless remote ( exactly like the Gemini Duo I have with their two button w-remote )

Now if you had a combo package that provided both your current lamp and another lamp designed for helmet high beam like I described, that would be a killer road setup. ( wireless remote a "must have" for road helmet use ) In the mean time I've been thinking about building a 2-cell AA Li-ion battery just to use with my Duo for road helmet "high beam only" use. Like I said before, don't need or want too much weight on the helmet, run time not a big issue and the Li-ion AA's should help keep things lightweight.

13. Originally Posted by Cat-man-do
Ummm....
In the mean time I've been thinking about building a 2-cell AA Li-ion battery just to use with my Duo for road helmet "high beam only" use. Like I said before, don't need or want too much weight on the helmet, run time not a big issue and the Li-ion AA's should help keep things lightweight.
Gloworm Alpha would work excellent for that too. About the same weight but comes with far better mounts. Xpg emitters only require about 60% of the current to produce similar throw though the beam is narrower. Not sure if they have the conversion kits to make them compatible with the wireless remote yet but when available would be a good option fore someone interested in extended runtime.
Mole

14. Originally Posted by Cat-man-do
Ummm....I hope you're not quoting me although I did say that. That was a typing error as I meant to say, "feet". I did go back to correct that btw. As to what you said about the other stuff, I can't help but hear the Rolling Stone song , "You can't always get what you want", playing in the back of my mind. Always going to be more people wanting something that really isn't either possible or worth making due to market demand.

With all this in mind I think you've gotten a lot of feedback here about what it is that people really want. I think the lamp you designed makes excellent headway as far as providing a very usable road lamp system. The only other thing I myself could want is something to provide the occasional "High beam" for when needed. It's my opinion that this might be better accomplished using a separate lamp and have it mounted on the helmet. The perfect setup would be something very light-weight but yet able to throw about 300 feet. I like using something like a torch because it's so lightweight and can throw a very tight spot ( run time not being an issue because only used for moments at a time ). My current favorite is the Convoy M1. The only additional thing I could wish for to make the M1 more useful would be the ability to turn it on/off using a wireless remote ( exactly like the Gemini Duo I have with their two button w-remote )

Now if you had a combo package that provided both your current lamp and another lamp designed for helmet high beam like I described, that would be a killer road setup. ( wireless remote a "must have" for road helmet use ) In the mean time I've been thinking about building a 2-cell AA Li-ion battery just to use with my Duo for road helmet "high beam only" use. Like I said before, don't need or want too much weight on the helmet, run time not a big issue and the Li-ion AA's should help keep things lightweight.
Haha it's not a direct quote, not trying to bash on any of you guys, more the guy who buys a light from Amazon and believes all the marketing.

I am currently working on designs for some new lamps, including a small internal battery one like the Lumina's or the Bontrager. I might have to put some of you guys on a private email list or something to solicit some feedback. I don't want to be posting renders and beam pattern simulations publicly.

However also working on improvements for the current lights, we'll have to see if they help. Always looking to improve the product so that come this fall or next year can be a serious competitor.

15. Also, there are manufacturing limitations on that sharp cutoff. How do cars do it? Through long focal length projectors that have small shields. What's the problem with doing that for a bike light? Well you end up with a light that is 6" long, and a 40-50% system efficiency, so a light with only 1000 lumen source is going to be putting 500 lumens on the road.

Most headlights with a sharp cutoff/projection style lamp require a minimum source of 2000 lumens because only 900-1000 lumens will actually hit the road. So have to go to either a really novel TIR lens or reflectors to get the system effiency up.

The problem with reflectors and gaining a sharp cutoff is that you need a BIG reflector to do just that, talking 100mm deep type reflectors that'd look perfectly fine on a car, but absolutely massive on a bike handlebar. Combine that with small manufacturing tolerances and the simulated beam pattern may not always be exactly what is expected. Case in point one of our earlier reflectors had the coating pooling too thick, 0.5mm too thick and it was throwing everything off. Ended up finding a new coating supplier, as well as redesigning the reflector so it wasn't so sensitive to manufacturing tolerances.

This is why I found the challenge of bike lighting to be absolutely fascinating. There has been a lot of advancements in optical design to boost system efficiency and create beam patterns that are acceptable for high speed vehicle use. Though for the most part most automakers have no issue with using 3000 lumens of light, burning off half of it to get that sharp cutoff line and pass regulations. New advancements in electric cars and MPG standards means a bigger push for more efficient headlights that require some really advanced software and math to develop. For bikers, that means headlights that can use less lumens, less power, and last much longer, or to reduce the system weight by only needing 2 lithium cells instead of 4. This is what I hope to be a leader in. The guy who is barely using any lumens by properly distributing the light.

16. Originally Posted by Outbound
...small shields. ... 6" long, and a 40-50% system efficiency...
Well... if the area above the cutoff is not contributing to illumination anyway, why not to block it? I personally would prefer the beauty of this (blocked) sharp line and smaller lumens, rather that larger lumens, but that are essentially misleading, bogus.

Back to bike lights. I believe the sharpness of cutoff is determined by mainly 2 things: size of LED and size of reflector. Right?

Saferide LEDs are domed, with dome diameter 2.5mm, makes LED die a square of 1.7mm. The dome is a magnifying glass, so a factor of around 1.3 more. Which makes it twice the size of Altilons. Reflector size is 2.5cm. It could be increased by 1.6 times, to 4cm (we don't want too bulky light). Makes the factor of about 3 (=2*1.6). THREE times sharper cutoff. Isn't it possible right now?

The FWHM (0.5-level of max, 2nd isoline) of the SR is 5 degrees, the 0.7-level (1st isoline) is 3 degrees. Ideals here have absolutely sharp cutoff, so realistically, (vertical) isoline size should be doubled. Which makes it 1.2*2=2.4 and 0.6*1.2=1.2 degrees correspondingly (FWHM=0.5 and 0.7). So, looks like the ideals are within the reach! (and of course, the main beam pattern can also enjoy great improvements compared with SR)

Am I wrong somewhere in my expectations?

17. Originally Posted by abvgdee
Well... if the area above the cutoff is not contributing to illumination anyway, why not to block it? I personally would prefer the beauty of this (blocked) sharp line and smaller lumens, rather that larger lumens, but that are essentially misleading, bogus.

Back to bike lights. I believe the sharpness of cutoff is determined by mainly 2 things: size of LED and size of reflector. Right?

Saferide LEDs are domed, with dome diameter 2.5mm, makes LED die a square of 1.7mm. The dome is a magnifying glass, so a factor of around 1.3 more. Which makes it twice the size of Altilons. Reflector size is 2.5cm. It could be increased by 1.6 times, to 4cm (we don't want too bulky light). Makes the factor of about 3 (=2*1.6). THREE times sharper cutoff. Isn't it possible right now?

The FWHM (0.5-level of max, 2nd isoline) of the SR is 5 degrees, the 0.7-level (1st isoline) is 3 degrees. Ideals here have absolutely sharp cutoff, so realistically, (vertical) isoline size should be doubled. Which makes it 1.2*2=2.4 and 0.6*1.2=1.2 degrees correspondingly (FWHM=0.5 and 0.7). So, looks like the ideals are within the easy reach! (and of course, the main beam pattern can also enjoy great improvements compared with SR)

Am I wrong somewhere in my expectations?
Not totally off, but on the right track. You'll never get the sharpness of a projector cutoff with a reflector, just the nature of it since that sharpness comes from a razor thin edge of the cutoff shield, and the fact it's a light absorber (or reflected back into the housing). Some low beam only projectors are a bit more efficient, but you are then factoring in 3 losses, the reflector loss, the cutoff shield loss, and then the lens loss. Plus you still want some light above the cutoff, not only is it easier on the eyes in actual use, but also needed to spot things like signs.

The sharpness also comes from the actual manufacturing of the reflector. You have draft angles, have coatings, all those things add in minor imperfections that can result in a slightly fuzzy cutoff.

You want a smaller LED in every situation, not a larger one btw. Makes it easier to accurately control light. The closer to a point source the better.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•