Designing good daytime rear commuter lights

BrianMc

Enlightened
Joined
Nov 4, 2009
Messages
940
....and for the back 2 bare red emitters on a piece of folded aluminium, epoxied in for waterproofness...
the backlight runs straight from my hubdynamo without fancy regulators or drivers. MEGABRIGHT with 120° lightning!

Yes, my next project a dyno tail light pair for the errand bike derived the reflector/bubble lights, testing a flare" lens (mostly horizontal), and no lens like the Red Zone 4s. I am impressed with drivers around the RZ4s (see the RedZone thread.) Nice to know that it isn't such a crazy idea.

BrianMc
 
Last edited:

kenwood96

Newly Enlightened
Joined
Sep 24, 2011
Messages
19
I used some optics before too, My conclusion was that without optics, they were far better to spot in all directions and really eyepiercing!
 

BrianMc

Enlightened
Joined
Nov 4, 2009
Messages
940
Yes, my next project a dyno tail light pair for the errand bike derived the reflector/bubble lights, testing a flare" lens (mostly horizontal), and no lens like the Red Zone 4s. BrianMc

A broken cold solder joint on the DX driver (not the best quality) of the MR16 red spot bulb modified to fit MR11 housing, in one unit shown in this post. My excuse to take the reflectors out of both. I test rode the reflector-less lights with these bubble lenses and made videos but need to redo them to post here. Good enough to tell me what I wanted to know. The 1 Watt Luxeons are as bright as the 45 degree angle of the RZ4's, at slightly off axis, brighter than the narrow beam Turbo or HotShot. The paired lights looked like one big light and they showed at about 30 degrees from head-on, for about 300 degrees. Flare lens? They don't got no flare lens! They don't need no steenking flare lens! I need to see how they do at a distance. Considering they are the same output as a single RZ4, they were amazingly bright at 1/8 of a mile.

BrianMc
 

BrianMc

Enlightened
Joined
Nov 4, 2009
Messages
940
The reflector-less lights:

photo0383nx.jpg


On (Single red Zone 4 between and lower:

photo0382w.jpg


Side view:

photo0371n.jpg


Angle view:

photo0379d.jpg


They are about as bright as my car's tail lights over about 170 degrees where you can see the LEDs.

BrianMc
 

Savvas

Enlightened
Joined
Jun 11, 2010
Messages
222
Hi Friends,
Found this at the DX site and just received today. A small 30mm (approx) diam mpcb with 6 X 1W red leds on boards.
Specs (from the DX page): SKU 94083
High power red light 6-LED emitter
- Voltage: 6~8V
- Power: 6 x 1W
- Current: 700mA
- Luminous flux: 110~120LM
From the specs I assume this is a 3S2P arrangement. I thought it would be worth trying with the dynamo, a bridge and a simple diffuser as a running light. I couldn't find a dedicated lens but there may be one at DX somewhere. There's a 4 led/4W/90lumen/4-5V version as well.
I'll report results!
Savvas.
 

1 what

Enlightened
Joined
Jul 6, 2007
Messages
617
Location
Australia
Hi Guys,
I'm still following this thread but have been very busy over the last few months.
Brian - I'm impressed by your ongoing enthusiasm. I agree that some of these DX modules are bright! and maybe specific lenses aren't all that necessary.
It's more the angle of view and side on visibility.
Savvas - interesting find - why do you need optics? How about mounting it in an acrylic tube 6 or more cm long so you get good side visibility. There will be more than enough light going out the back and heaps directed to the sides of the tube.
I've got a new taillight to replace "The Enterprize" and will post details in the next few days when I finish putting all the photos etc together.
 

BrianMc

Enlightened
Joined
Nov 4, 2009
Messages
940
....Found this ... small 30mm (approx) diam mpcb with 6 X 1W red leds on boards.
Specs (from the DX page): SKU 94083 High power red light 6-LED emitter - Voltage: 6~8V, -Power: 6 x 1W, - Current: 700mA,- Luminous flux: 110~120LM
From the specs I assume this is a 3S2P arrangement. I thought it would be worth trying with the dynamo, a bridge and a simple diffuser as a running light. I couldn't find a dedicated lens but there may be one at DX somewhere. There's a 4 led/4W/90lumen/4-5V version as well. I'll report results!
Savvas.

Interesting and puzzling.

First output: The old '1 watt' spot MR16 bulbs I cannibalized for my lights test as about 100 lumen with my crude integrating sphere. That is close compared to the Red Zone 4's at half power. The 'one watt' led in the Turbo with fresh cells showed 90-100 lumens for a few seconds, too. There is a new XP-E Red rated 72 lumens at 350 mA, and 130 at 700 mA (1.6 W). If only 120 lumens total, that is 20 lumens per watt. The Luxeon is about 70 lumens per watt at 700 mA, and the XP-E Red is about 80. If 120 lumens each die, that is 150% of the best CREE. I don't think so.

Second resistance: Red power LEDs are typically around 2.1-2.5 Volts Vf at max current of 700mA. The Luxeon in my light s is actually about 1.4 W at 700 mA. A true 1 watt at 700 mA would be about 1.4 Volt Vf. Six in series would be 8.4 Volts. 3S2P only 2.1 Volt. So with approximately 6-8 volts to drive the array, it looks like they have a very low Vf for red LEDs and maybe a hair under 1 watt each giving an 8 volt max in series.

So we hope you will be pleasantly surprised.

BrianMc
 

1 what

Enlightened
Joined
Jul 6, 2007
Messages
617
Location
Australia
As promised my new improved taillight.
As much as I like my light described in post 177 and 186 I've continued to think about the issue of attracting the attention of other vehicle drivers (so that they might "see us" and not run over us).
Brian, Savvas and others have written here and in other posts about the perceptual challengers involved in cutting through an overload of fatigue and visual data and I agree with almost all that has been said. The question is "how do we safely make them aware that there is a smaller, vulnerable and slower moving vehicle hidden in the dark"?
Brighter lights? Flashing lights? Bigger lights?
Then I thought….How about lights that change colour and change size….That's not something you see every time you drive. That might register and attract attention.
Therefore I present for discussion the yellow/red size changing light. From the rear it simply flashes yellow and red but from the side (where most danger lurks) its' size changes with the red light being twice the size of the yellow light.

redyellow.jpg



In order to attract maximum attention from the side it is also larger than its predecessor.

size2y.jpg



Summary of the build:

assembly2.jpg


Like its predecessor it has a Fenix diffuser on the tail so that it can be diffused for night-time use and opened for full direct output brightness for day-time use as shown in the last photo of post 186.
The LEDs are driven at 750ma from an astable multivibe with a 3:1 Yellow:Red ratio.
It has good visibility from the oblique front of the bike both day and night.

daynightvis.jpg
 
Last edited:

BrianMc

Enlightened
Joined
Nov 4, 2009
Messages
940
Nice 'out-of-the-box' thinking,1 what

Starship Enterprise & Jedi Light saber. I sense a theme. Then I guess mine look like small versions of Super Star Destroyer ion drives. :)

So 8 1/4 watt yellow LEDs in parallel for 63 mA, and one CREE XP-E Red, at 500 mA? What circuitry are you using for the switching?

I see you are using another diffuser. I don't think mine are as bright from just before 90 degrees to the forward angles, but I had a bright white door messing up the camera exposure. They should be close on lumens. So I may try a diffuser to increase that a bit since I can sacrifice some out the back output. I need to get a photo of the red wash off my lights to compare. Maybe with and without the diffuser.

Since I would like a light to use the battery pack(s) I have, I am thinking Bobblehat's flashlight & orange traffic wand vertical light in the backpack (on my back) idea, might combine well with this idea in a dedicated bike light. If you are using 500mA, a dyno version is feasible. That would be interesting for my other bike.

BrianMc
 

1 what

Enlightened
Joined
Jul 6, 2007
Messages
617
Location
Australia
Thanks Brian.
Flasher/driver cct follows:


flasher.jpg
As you can see I built it the old fashioned way ...Microprocessors are for yuppies!
The battery pack is a 7.4V, 6000maH unit (ie 2S 2P)
Re the yellow LEDs. They are 0.3W and can be driven up to 100ma.
There are 9 of them in the light (2 at the top) in parallel and total drive current is about 666ma.
See;
Link to yellow LED:
http://besthongkong.com/product_info.php?cPath=19_63&products_id=534

The Cree is driven at the same level.
 

BrianMc

Enlightened
Joined
Nov 4, 2009
Messages
940
I saw this and I thought it would be clear when I revisited. Is the word 'multivibe' and is the 'other output' the 'LED driver circuit' where you duplicate the dotted outline circuit?

The 0.35 Watt 5.8 lumen LEDs are interesting.

BrianMc
 

1 what

Enlightened
Joined
Jul 6, 2007
Messages
617
Location
Australia
Hi Brian.
The answers are "Yes" and "Yes".
The multivibe (multivibrator) produces the timed signal that turns the 2 LED drivers on and off (ie one turns on when the other turns off).
The LED drivers switch the current (on and off) that goes to the LEDs.
The current flowing through each LED is determined by the resistor in series with it (the 1.5 ohm R in this case).
Since the supply voltage is 3.3V and there is about 0.3V drop across the switching transistor when its turned on and about 2V across the LEDs that leaves a 1V drop across the series resistor which gives a drive current of 1/1.5 (I=E/R) or somewhere slightly north of 666mA.
Purists and young circuit designers please note that I know this is a rough estimate only but thats how many fine things were prototyped back in the Dark Ages when something called a slide rule was the Lab computer!
 

BrianMc

Enlightened
Joined
Nov 4, 2009
Messages
940
I built electronic kits including Dynaco's AM-5 tuner, or much more recently built up Martin's Circuit 12 dyno board. I have not bread boarded any novel circuits. Familiarity with the concepts is not the same as hands-on experience. Like sex education versus being sexually active. :) My Hughes-Owens SR is waiting to be framed for display. A thing of beauty is a joy forever. It was an award for winning the high school Math contest. I am chelating Heavy Metals to regain my health and that drops the IQ a tad. So I appreciate the clarification.

Looks like using DX's 5 volt and modifying values would allow 2 series CREE XP-E Red-orange LED's and twice the number of .35 watt ambers or a or maybe 4 2S2P XP-E ambers (0.5 A max) should work.

BrianMc
 

1 what

Enlightened
Joined
Jul 6, 2007
Messages
617
Location
Australia
Should work "a treat" and emit an awful lot of light.
The transistors I used are universally available and dirt cheap (great pun for something made out of silicon!)
 

1 what

Enlightened
Joined
Jul 6, 2007
Messages
617
Location
Australia
The light in post 208 might be on its way to the History Museum!
I was in my car a week ago and noticed one of those led emergency traffic noticeboards (in full daylight)...thought "those leds show up ok in daylight!"
A little research and I discovered CREE Screenmaster 4mm x 5mm leds. Made especially for outdoor displays.
Cutter were kind enough to sell me a few and mixing them with the yellow leds from 208 gave me:
[video]http://img268.imageshack.us/img268/6353/k4ub.mp4[/video]
As you can see I weakened and used a microprocessor this time.
Now all I have to do is shift the prototype to the bike!
 

BrianMc

Enlightened
Joined
Nov 4, 2009
Messages
940
Interesting. I have my order on its way with three XP-E Red oranges for a dyno tail iight as 1S3P serial with the head light. That should be about 100 lumens, or the power of three Turbos, using a flare lens on each for a 15 x 70 degree FWHM beam on each, with the outside ones aimed at an angle (30-45 degrees) to the bike axis. Martin's circuit I am using was for about 10 Vf of Luxeons, and i will have about 10 Vf of XP-E and XP-Gs. If I change out the doubler board I can up the rear array to 3s1p and 300 lumens.

No fancy flashing though.

BrianMc
 

1 what

Enlightened
Joined
Jul 6, 2007
Messages
617
Location
Australia
Follow up re post 215.
While I thought it looked great close up (both the colour change and the clever flash pattern) it was a failure.
Even at 30 meters it was disappointing and would not attract a lot of attention since the detail of the pattern was lost due to the small size of the LEDs. Plus the busy colour change pattern was too fast and ended up "a twinkle".
"Twinkle twinkle little light, I don't want you upon my bike"
So….Back to the light in post 208.
I reprogrammed the microprocessor to give a more sedate and traditional flash/colour change pattern and I can assure you that is much more visible.
http://www.youtube.com/watch?v=nBVcdcifsWk

Lesson learned: The apparent size of the light source is important and complicated high speed flashing is not very effective.
 

Savvas

Enlightened
Joined
Jun 11, 2010
Messages
222
I reprogrammed the microprocessor to give a more sedate and traditional flash/colour change pattern...

Hi 1what,
Great light!!! But can you elaborate on your 'reprogrammed the microprocessor' statement please. Maybe I'm missing a 'humorous aside' but I thought you said 'microprocessors were for twinkies' or something similar! The idea of slowing down the rapid flashing makes a lot of sense to me...
ta,
Savvas.
 

BrianMc

Enlightened
Joined
Nov 4, 2009
Messages
940
That is an interesting light. If I calculate output to lumens and lumens per watt, the small LEDs are about 20 lumens per watt, while amber XP-Es are about 100 lumens per watt, so I wonder about using an acrylic tube as a fiber optic fed by an amber XP-E on each side as being bright top or bottom isn't as important. The vertical flashlight on the back so it is up over at least some SUV's using a similar setup may be much better than a turbo on the back of the helmet. I an old thread they said that the PBSF was the ultimate answer. Apparently not. :)

BrianMc
 

1 what

Enlightened
Joined
Jul 6, 2007
Messages
617
Location
Australia
In order of posting:
Savvas - Yes microprocessors are for yuppies but when I was flirting with a more complex flash sequence they were the only way to go since you have infinite and rapid control over the sequence of loght flash as well as colour change and I was in the mood to play!
Since I already had the LED driver built it was easy to reprogramme it for the new slower sequence.
As an aside it was the first time I've used them in a light and it was very interesting to see how different sequences looked. I now have no doubt that some sequences attract attention much better than others. The rule of threes seems to apply with 3 flash ptns being especially potent. Longer complicated ones just dont "grab".
Also when using 2 colours it was more attention getting if one of them looked brighter than the other. When the Y & R were equal it was nowhere near as efffective (it's easy to adjust by altering the current setting resistor in the driver). I chose to set red brightest since it is a tail light.

Brian - Care to upload a rough drawing of what you suggested?
 
Top