World's brightest topics on this page include:

World's Brightest Bike Lights  2010
I ride a bike to work.
I ride a bike home at night on an unlit country road.
I use a 5 watt LED bike light which gives a feeble spot of light ahead of my bike.
So, I think to myself "what if I had a brighter light on my bike"
I think to myself "what if I had a really, really bright light on my bike"
Hey, I'm Tesladownunder,  "what if I had the world's brightest light on my bike"
What would it take?
I have experience with a huge 100W LED that is brighter than a cars headlights.
But that is simple to do and I have already done that almost as a joke.
But what extreme can I take this to?
Lets make a real project of this and give it a real budget - say $1000 (later blown out to $8,000).
So....here it is, ladies and gentlemen; welcome to the Worlds Brightest Bike Lights.
How bright?  100,000 lumens. Remember that figure.

Gallery

OK.Step one. Take a ladies bike with front tray and rear carrier.   Add a big round thing in the front and some tacked on stuff and you can get some idea of where this is going.

So here it is in its full glory.

Lighting a lighthouse!  Here is ONE 100 W 6000 lumen LED focussed to a 5 degree beam. I will be using 15 of these in the front of the bike. Check out my 100W LED page.   The Bunbury lighthouse is on but you cant see it's beams easily in the photo from this angle. The 100W LED runs on three 1.3 AH, 12 V SLA batteries attached to the LED which is resting on on the wooden fence here. It is fan cooled and runs around 3 A 34 V via a 0.4 ohm resistor, if I recall.  This is 6000 lumens - remember the 100,000 lumen total figure.

 The World's brightest flashlight shown above lighting a lighthouse with different colours.  The last photo is with lights on full with much shorter exposure.  This uses 15 of the 100 W LEDs.

First light for the rear lights (above). These are separate red, blue and green and are mostly decorative at a "mere" 300 W.  Shown here with blue (left), red plus blue (center) and red, blue and green (right).

Specifications

Front lights: 
    15 x 100 W LEDs (12 white, 1 red, 1 blue 1 green: all approx 6000 lumens)
    1500 W total
    90,000 lumens
    Beam 5 degrees
Rear lights:
    3 x 100W LEDs (1 red, 1 blue 1 green)
    300 W total
    18,000 lumens
    Beam 20 degrees
Total:
    18 x 100 W LEDs (12 white, 2 red, 2 blue 2 green)
    1800 W total
    108,000 lumens (range 82,000 - 115,000)  
    ie approx 100,000 lumens total
Power:
    3 x 12 V deep cycle 33 AH batteries
Run time:
    est. 10 mins at 60 A 33 V                   

LED data from  Led-world2007 are limited and to my mind insufficient and at variance with other similar LEDs. The lumen question marks in the tables below are for the quoted lumens from similar LEDs from a different manufacturer which might be closer to the true values.

100 W Cold white LED
Parameter	Min.	Max.	Unit
Luminous Intensity	6000	6500	lm
Color Temperature	9000	11000	K
Forward Voltage	32.0	36.0	V
Forward Current	3200	3800	mA

100 W Red LED
Luminous Intensity		6000 (? 3500)	lm
Wave Length	625	630	nm
Forward Voltage	20.0	24.0	V
Forward Current		4000	mA

100 W Green LED
Luminous Intensity		6000 (? 5000)	lm
Wave Length	520	530	nm
Forward Voltage	30.0	36.0	V
Forward Current		3200	mA

100 W Blue LED
Luminous Intensity		6000 (? 1200)	lm
Wave Length	460	470	nm
Forward Voltage	30.0	36.0	V
Forward Current		3200	mA

Concept
 In designing this bike at this time I am taking advantage of two things.  Firstly, my experience with a single 100W LED last year (when they cost $500) and secondly that pricing of the 100 W LEDs have dropped to around US$100 which makes a multi - LED project feasible. So with some expense I can make a leading edge project with massive light output that was not possible only a few months ago.

Some of my original concepts are shown above. Things have evolved a bit since.
Note that for static displays or media it will work better if it is not just a "one trick pony" but has other aspects of interest. So my primary goal is massive light output but in addition I want it to be a light that I can strobe, pulsate, dim, sync to music, or colourise to any colour at reduced power.

This is not meant to go on the road any more than a monster truck or an F1 racer is meant to go on the road. They are the expressions of the strongest and fastest extreme versions of the 4 wheel "car".
But we still enjoy watching them in their place.
These bike lights should never be shone at oncoming traffic. Nor the brake lights either. This is a display bike. I will have a very low power setting of a few watts that will suffice for road use and maybe a standard bike light set for emergencies.  So, this is definitely NOT a mountain bike setup. Apart from TV and print media, I envisage this being used in static displays, at bike shops and shows. I won't get many miles on the clock. It has 3 internal gears but is nothing like a proper bike to ride.
HID's/Xenons are a bit more efficient than the best LEDs. They are also more compact and are much closer to a point source and can be focused tightly. So for that tight beam they would be great.
However, I don't have any special expertise with "glass" but do with 100 W LEDs so that seemed the natural way to go.

"Ohh really .... someone else claiming the biggest and best".
Well yes, but this here is the data.  My bike has 100,000 lumens.
Compare:

Construction

Bicycle: Electra, "Townie" ladies bike, cream AU$1300. Chosen by my wife. Ladies bike??? Big advantages of being able to step through and sit on the seat with feet flat on the ground to support the high center of gravity of the battery weight. The front tray is a big plus as well. No point in trying to make a heavily loaded racing bike or mountain bike I reckon. Old lady retro feel styling in cream. (I wanted the purple flowery one but my wife was firm).  It has 3 internal hub gears and back pedal brakes. And it has a bell...

Our cat and it's WTF expression on first ride of the new bike.  Cycle outfit is from my daily ride home on my "normal" bike.

Batteries: Three 35 AH Absorbed Power deep cycle batts . These are like car batteries but with better construction to allow greater discharging.  From the data sheet, these are good for 12.2 V 35 A dropping to 12.0 V at 12 mins. At 70 A, 11.8 V dropping to 11.7 A at 6 mins. Need 3 batteries to run 34 V at 52 A peak (1.8 kW) delivered to LEDs ie need to drop from (12.0 x 3=36) 36 V to 34 V. 2 V drop at 50 A is barely enough for heavy duty cabling and some big MOSFET's to do pulse width modulation (PWM).  This is all a bit on the edge and will call for some tricks.

The left photo shows a battery close up, center photo shows shows what happens if you accidentally short circuit them while using a screwdriver to tighten a terminal and the right photo shows the three 2.7 A chargers.

Front LEDs
From eBay of course. Led-world2007  is the place I use.

My original plans were to use 11 x 100 W LEDs.  Rated at around 34 V (32 - 36 V) at 3.5 A (3.2 - 3.8 A) with luminous intensity 6000 - 6500 lumens from a 1.8 cm x 1.8 cm surface. Roughly 10 Watt light output by my calculations and hence have to lose the rest (90 - 110 W) as heat hence the decent fans.
In addition to the forward facing 11 white LEDs there will be 3 coloured LEDs (Red, green and blue all 100 W 6000 lumen) facing forward to a total of 14 LEDs facing forward ie 84,000 lumens. The 3 coloured LEDs in total will be white but will allow me to run them separately to generate any colour.
For info on white 100 W LEDs see 100W LED page about it. 

These will be cold white 9000 - 10,000 K colour temp like those annoying expensive blue Xenon car headlights.

The LED ballasting is a complicated issue. LEDs are best driven by a constant current driver since the voltage drop across them is fixed and any attept to use a higher voltage will raise the total current dramatically. On small LEDs a simple resistor is adequate but this is not really suitable here.
While a switchmode buck convertor is ideal, there were cost and design considerations and a short timeframe.  Accordingly I have chosen to use a resistive ballast as is used with small LEDs.  Rather than a straight resistor, I have used 6 V 18+18 W globes as a combination ballast, current stabiliser and fuse all in one for each LED where it has to drop 6 V max. Globes are a non-linear resistor that has a very low resistance at startup when the filament is cold. Think about it: 3 amps almost constant current, no heatsinking problems. Anyhow, much better than a resistor. Now I just have to go and raid a lot of Kombie taillights... Actually sourced these from Delhi (6 V vintage motorcycle globes).
As an example, I recorded a peak current inrush of 6 A into my green LED today (rated 3.4 A) then settled nicely at 3.1 A using a 12 V 100 W globe which only glows to red heat. How ironic to use that as ballast for a 100 W LED!  Peak pulse rating for a 1/10 second pulse is 5 amps which is about 50% over continuous rating.

The left photo above shows the output of a photodetector taken at switch on of the white LEDs in the final array using a 6 V 18 W filament as ballast.  Note the 5% increased light output at switch on which tails off in the first 100 ms or so as the light globe ballasts heat up.  I had thought it might be greater. The right photo shows the current at switch on is almost 100% greater at 7 A reducing over the next 200 ms as the filament heats up. It suggests that I am running the LEDs (at least a single LED with a full battery charge), close to its maximum.

Here are the current and voltages at switch on of all 18 LEDs on the bike. Peak is 68.6 A (3.8 A per LED) which settles to 52.9 A (3 A per LED) 2.0 seconds later. Peak power is 2682 W settling to 2004 W.

The red LED uses both filaments of the 6V 18 W+18 W to drop 12 V, plus a 1.5 ohm 15 W resistor to drop a further 4 V, whereas the blue and green use one of the filaments only. I have a later option to switch in the second filament if the main voltage sags too much under load.

Here is an alternative source for 100 W LEDs and specs. Snowdragon.

The left photo above shows me starting the front array construction, center photo shows how the fans will sit and the right photo shows first light of a single 100 W LED. Note the weak 150 + 150 W shed lights.

The left photo above shows the front array construction with red, green, blue and white LEDs at low power. Center photo shows the rear view with the rats-nest of wiring becoming evident, and the right photo shows first light of the partial front array on the bike. "Only" drawing 700 W at this stage out of target 2000 W.

Controls
The rear control box which includes a 100 A keyswitch (1000 A peak) plus 6 switches for switching the rear lights (3 to control RBG independently plus one for road safe red stop light plus one for PWM dimming plus one spare). There are another 5 switches to control other 12-36 V functions as I might decide later.

So far only RGB switching basic capability at present.

Rear LEDs
Three coloured LEDs in the nominal 100 W and 6000 lumen range. If you believe the source data.  Or if you believe a similar 100 W LED manufacturer then red 88 W 3500 lumen, green 100 W 5000 lumen and blue 100 W 1200 lumen average. Bear in mind that lumens relate to the eyes sensitivity at that wavelength - greatest for yellow, least for blue.
These are facing backwards. ie approx 10,000 - 18,000 lumens to the rear as well. 
The colour gives capabilities of, for example, a red brake light or a flashing red and blue police light, or a smoothly colour changing light etc. This gives it more variety by far. And it gives crazy shadows.......

The red green and blue lights above left give really colourful shadows as the primary colors mix.  The basic LED mounted onto the heatsink is shown center and the effect of the lens shows the fine detail of the 1W 10 x 10 array of LEDs on the right.

Above shows the mounted LEDs in the rear array which can be tilted vertically (for the Xmas tree project).  The plastic toolbox is an excellent fit for the red, green and blue rear lights fans and optics. It is compact and I will be able to swivel it as well.

Above shows the as yet unwired rear array.  The center photo shows current of 9.3 A 37 V = 344 W. Love this clamp meter with 40 A DC scale - bought it especially for this project. Ballasting of the Blue and Green LEDs is via 6V 18 W globes. The Red LED runs a lower voltage so has 6 V 18 + 18 W plus a 1.5 ohm, 15 W resistor in series.

Above shows the low power (2 W) tail light riding mode with a normal tail light for comparison. Keeps it street legal (maybe).  You can see each of the 1 W LEDs in the array.  I also have a low power 10 W front light mode.

Electronics
At the moment I have simple switching but my plans are for more electronic control.  I have a PIC driven MOSFET motor drive kit for 40 A at 24 V which I will adapt. to rate 70 A MOSFETS's x 4 in parallel are IRF1405's. 55 V 169 A. Peak voltage is a bit close but nothing much inductive here. The front RGB LEDs will be switched into the main PWM or onto an individual circuit.

Left photo above: Every man needs his shed... Right photo shows a battery power test.  The meters above, read 55 A, 33 V and water coolant 65 degrees C during a power test.  It took me a bit by surprise that the bucket of water reached almost 80 degrees C and I had to dash out to get more cooling water.  That's what 1800 W does.

Heatsinks, Fans and Lenses
Each LED needs to dissipate about 90 W of heat, similar to a modern CPU in computer.  These fans and heatsinks are low weight but effective.  The bike uses 15 of these in front and 3 in the rear. 

The 10 cm diam glass lenses need to be about 10 cm from the LED to focus to 10 degrees so you will lose some of the beam to side scatter. I want a tight 10 degree beam in front and a wider beam in the rear. Again from Led-world2007 .   Later I hope to try a large fresnel lens to see if I can tighten the beam more.

Mechanical supports

I made an outrigger stand in the left photo, as the inbuilt one would not handle the 33 kg battery weight so high up. I cut up a tri-bar set that I got for a few dollars from a tip years ago. These are extensions to the handlebar of a road or mountain bike out forward for a better racing position for triathletes in particular. Add a PVC conduit and these can provide a good support and can be flipped backwards out of the way when the bike is being ridden.  For trailer use a more solid setup was needed in the right photo.

Reminds me of the pun:        A bicycle can't stand on its own because it is "two tired". 

Then I made a battery tray out of perforated angle iron and particle board. Simple and effective but quick. Downside is the centre of gravity is high but it is much more compact and unobtrusive than some other ideas I had.  Also able to do the heavy cabling to the front of the bike - needed as can't afford much voltage drop at all at 55 A+. Black cable had 85 mV and red cable 235 mV including switch ie a total of 0.3 V at full load.

It handles like a tank but is still rideable. But don't I look stylish riding it! Don't forget that this is a show bike and not really for much road use.

See it from space?
Surprisingly it should be able to be seen from space. Let's consider the International Space Station (ISS) which typically passes by at a mere 230 miles (370 km) overhead. Note that the front array of the bike can be reconfigured into a flashlight and pointed upwards.

  (click to enlarge)

The ISS has a nice recently installed viewing cupola (above left) where astronauts can take happy snaps like Las Vegas (above right). This has the reputation of being the world's brightest city and  the brightest lights are probably beams facing upwards highlighting buildings like casinos etc. Individual lights can easily be seen as well as yellow low pressure sodium street light lighting the roads in the "suburbs". 
Conclusion: The ISS can see bright lights on the ground, even if the lights are not aimed directly at it.

Now, can we calculate whether the flashlight can be seen from space?
See the discussion on 4HV forum by Chris Russell for calculations based on a 14 LED array fully driven and with all light fully focussed into a 5 degree beam. My modified version below accounts for a broader beam of 10 degrees and perhaps 50% light loss to "flood" rather than beam.  Also there are 15 LEDs although with reduced output of the coloured LEDs. The LEDs are driven to perhaps 80% of rated.
Flashlight nominal output: (6000 x 12 + 5000 + 3500 + 1200) = 81,700 lumens nominal.
It is driven to perhaps 80% with perhaps 50% flood losses = 32,000 lumens in beam.
ISS optimum viewing orbital altitude( >85 degrees elevation) : 370 km (230 miles)
Flashlight beam width diameter at 370 km:      2 * tan (5 degrees) * 370 km = 64 km
Total illuminated area (assume circle) = 3200 km² = 3,200,000,000 m²
Total illuminance (lux) in beam at ISS: (32,000 lumens) / 3,200,000,000 m² = .00001 lux
Apparent magnitude: = -[log(50,000,000 * lux/127)]/log(314/125) = - 1.48
This is the same as the brightest star (Sirius)which is also -1.47
For comparison, the Sun (130,000 lux) is App. magnitude = - 26.7
Conclusion: Calculations show that it will be easy to see my light from the ISS with the naked eye.

Now let's compare that with an actual measured reading using a lightmeter.

   (click to enlarge)

Above is the lightmeter recording 111,000 lux in afternoon sunlight. Meter range is up to 400,000 lux.
I have measured flashlight  values of 13 lux at 120 m from the Flashlight (200,000 lux at the lens).
This gives an illuminance at 370 km of 1.3 x 10^-6 lux = Apparent mag - 0.7 which is the same as the second brightest star, Canopus at - 0.72. 
There is a discrepancy between calculated and measured (-1.47 vs -0.7 magnitude) likely due to assumptions about amount of loss in flood light.
Conclusion: On measurements, the ISS should see my light by the naked eye like a bright star in intensity. 

 But how will you know it is my light when seen from space? Easy.  Just modulate it on and off.

    (click to enlarge)

Above is a model  I made with a lot of LEDs to demonstrate the concept. The still photo gives no indication which is "my" light. The moving photo image, however, makes it very obvious and shows a row of dashes as my light is turned on and off electronically. In addition, the modulation can be smarter than simply on and off.  How about the world's first flashlight to satellite messaging using Morse code?  Morse code is a largely obsolete messaging system using long "dashes" and short "dots".  The best Morse code users are faster than SMS texting.
Conclusion: My light can be differentiated from other adjacent lights by on-off modulation.

 (click to enlarge)

Above shows a PIC microcontroller programmed as a Morse code transmitter.  It uses a PICAXE 08M and details are here.

 (click to enlarge)

Above shows the preprogrammed message "PITTSBURGH" named after a long running Morse beacon there. I have modified the program to give a 2 second delay then a dot then the message after a further 0.5 s. The Morse speed was doubled from the default of 10 WPM but needs to be faster.  This is a 6 second exposure hence message is about 4 seconds. Needs to be a lot faster.

"TDU to ISS" in Morse dots and dashes would be: - -.. ..- - --- .. ... ... 
Or in more detail including spaces:  ===...===.=.=...=.=.===.......===...===.===.===.......=.=...=.=.=...=.=.= "
Remember SOS is dit, dit, dit,      dah, dah, dah,      dit, dit, dit

 (click to enlarge)

I took the above photo of the International Space Station in a 4 second exposure on 4th March 09. The moon has a lens flare and there is a lot of extra light from the moon and the time around sunset. The ISS image passes about 3 moon diameters (3 * 0.5 degree = 1.5 degrees) in 4 seconds although I didn't record the elevation which was not immediately overhead. Apparent motion is from west to east - opposite to the sun.  
Now they were using a 400 mm lens on the ISS so the field of view at full magnification will be restricted to 3 x 5 degrees. With their camera not tracking, I will pass through their 3 degree minimum camera field in 2.5 seconds. This assumes it is 90 degrees (vertically overhead) travelling at a velocity of 7.8 km/s at 370 km. 

 (click to enlarge)

Note that if the ISS view is only 3 degrees wide, then there is no requirement for my 10 degree beam to move at all. Interesting. However, that assumes the ISS camera is aimed exactly vertically but this may not be possible in what has usually been a handheld camera.  In reality the camera would be pressed up against one of the cupola windows which would be unlikely to be at 90 degrees.  In fact they would have to aim it carefully to make sure the path includes my position and they have only seconds to be sure. So I would be most likely to get best results with tracking it after all.  It will be easier for the ISS to have a wide field of view and for me to have a narrower tracking beam.
Bear in mind that the main cupola window is 31 inches wide and the cupola was designed for two astronauts. Hence even though the camera view at full magnification would be brief, it may well be possible to have one or two astronauts actually looking directly for the flashing light with a potential visibility for perhaps a minute. Tracking may well be appropriate in that case. The high speed morse code may be too fast for visible flashing so may have to add longer on/off periods outside the main code transmission. My beam will be naked eye visible in a 40 mile path ie 8 seconds in the ISS.
Conclusion: The beam width and relative passage times have been calculated. Tracking by me is preferred but not essential to prolong the ISS view and reduce the requirement for ISS aiming accuracy.

Just imagine that I am on the ISS.  Would I be able to see the light or a streak with flashes of Morse code with a 400mm lens?

    (click to enlarge)

Here is Alpha Centauri which is one of the Pointers near the Southern Cross and the third brightest star at +0.01 apparent magnitude. The left photo shows clear images of the star (click to enlarge) and others with my 180mm lens and a Nikon D300 with a 5 second exposure.  The right photo shows the streak of manually swinging the camera (faster than ISS transit time).  The star remains clearly visible although it has a lot of "twinkle" as it was not that far off the horizon.  It would be even better with a 400mm lens and slower transit and the actual flashlight should be 0.7 magnitude brighter anyway than Alpha Centauri. So, yes, point the ISS camera in the right direction and they will see it with magnitudes to spare provided the cupola is in the right viewing orientation.
Conclusion: The ISS should be able to record the light from the flashlight with a long exposure from a fixed camera rather than tracking one.

Also getting the "TDU to ISS"  message across needs to be done twice in 2.5 seconds to ensure that at least one full sequence is received. That's too fast for human generated Morse code so it will be electronically generated with a small computer chip (PIC microcontroller) and ideally will be seen in full across one picture of duration of just over 2.5 seconds.
Conclusion: I can send a Morse code message to the ISS.

Where is the ISS?   At 15 orbits per day, the ISS often comes close to my home town of Bunbury in Western Australia in any given 24 h even if not visible.  Using NASA Skywatch, one can plot passover times and tracking data.

How can I be sure of the best ISS view. The brightest view will occur when the ISS is immediately above and closest. Realtime views on Google maps are here for example. If I travelled 125 km to Arthur River (33.396 S, 117.035 E) in dark rural countryside and faced the light exactly vertically upwards on Jan 12, 2010 at 22:43:20 hours (14:43:20 GMT) then the ISS would be directly above.  I could flash it a few times to say "Hi".  Easy as that. But a combined Morse code and prolonged visual contact with tracking would be better and have a photographic proof of the contact. Note that I have to make sure this is wake time for the astronauts which is 06:00 to 22:00 if I recall. They could pop over to the cupola after lunch at 14:00 GMT.
Conclusion: I can travel to get the best view for the ISS hopefully at exactly overhead at exactly the right time and know tracking details in advance.

But what could go wrong? Lots.
Clouds, for example, but in summer here November to March are not that common - perhaps less than 10%. Weather forecasting and satellite imagery can give enough time to get a message to ISS to cancel and "rebook".
ISS camera aiming. I have no control over that and I am concerned that a 400 mm lens has too narrow a field for a handheld camera to be accurate with a margin for error. That will be a matter for discussion. Backing off to 200 mm might be safer.
My mechanical aiming error. With 10 degrees, there is a big margin of error. The mechanical tracking just needs to pass through the vertical at an appropriate speed perhaps from 30 degrees either side with known directions from and to. I won't need to set astronomical coordinates. I could have an astronomical colleague confirm accuracy as needed in test runs perhaps using a laser guide on my flashlight.
The cupola orientation might not be optimal, but I don't have that information.
Time to complete electronics and mechanicals. No doubt I will be energised if given the go ahead.
Conclusion: The predictable potential threats to success are hopefully small.

Overall this project seems viable and would likely generate some positive publicity for the ISS (ahem.. and me). The simplicity of "seeing a flashlight from space" is one that the public can understand.
I'll drop Mr NASA a line.....

Distance shots
Calculations show similarly that at 10 km the light should be as bright as a half moon (apparent magnitude of full moon - 12.92) . So let's test this.

       (click to enlarge)

The three photos above are from my first distance shot. 2.5 km along the straight stretch of Lillydale Rd out of Bunbury.  
The left photo shows the 1/4 moon for brightness reference and a car heading down that stretch about 100m away. My son's car taillights are in the foreground.
The center photo shows the flashlight firing the white LEDs 1200 W from 2.5 km. The moon has moved to be a bit obscured by trees (long wait for a car free period).
The right photo shows the green light alone to confirm the source. In fact this green brightness (about 1/16th total power) would be similar to a 10 km brightness ({10 km/2.5 km} ² =16 ) for the whole array and indeed looks like it would be similar to a half moon intensity. So the maths seems to be broadly holding up to experiment.
I need to find longer distances to do further testing.

Here's the light seen from 9 km over a good part of the Australind Estuary (Bunbury, Western Australia).

      (click to enlarge)

The left photo shows my light pointing towards the suburb of Australind across a long estuary. The center photo shows how the light appears from 9 km away with car taillights and streetlights for comparison. Beam width would be around 1.5 km.  If shone at the ISS at 370km, the beam would be 1700 times less bright [(370/9)² ] but with a dark field and dark adaption of the eyes this should be easily visible. The right photo shows the observation path from Google Earth. I also did a 12 km shot but got partially obscured by the geography.

Media
The website images above are available to be used with acknowledgment .
There are also two professional videos in HD with overlays, interviews etc available through Barcroft Pacific Media covering  the Worlds most powerful bike lights and Xmas tree as well as the Worlds most powerful flashlight.

The Bike and Xmas tree were covered by GWN regional TV on December 23rd 2010. Video here
Channel 7 covered this nationally in Australia on December 24th 2010.

The SouthWest Times did an article on 30th December 2010 for the World's Brightest Bike Light.

This news photo (left and center) is one of the only "photo shopped" images on my site. It was presumably done by the paper to remove the lens flare over my trouser area and to tighten the beam.  Did I tell you I didn't like photoshop? The right photo is not retouched (apart from being flipped for comparison) and is one I took at a different time which looks very different. Two points. Firstly there is no "beam" in the real unretouched photo. A beam arises from the light through the air which is currently is dry, smog free and clear. Even a laser of medium power would be hard pressed to show a beam from the side, let alone a beam arising from outside the light source as on the left.  So what is the bright haze? It is partly lens flare (with lens flare copy diametrically opposite over back wheel) plus the effect of light scattering from dust particles (largely diffraction) at low incident angle. The scattering should diminish evenly around a point source ie is typically circular.  So which is the better photo of the front array lights? You judge.

Online newspaper articles in major papers of Sydney Morning Herald , Brisbane Times and The Age plus regional papers including Areanews Banyuleandnillumbikweekly Barossaherald Batemansbaypost Baysidebulletin Bendigo Advertiser Bombalatimes Boorowanewsonline Bordermail  Braidwoodtimes Busseltonmail Camdencourier caseyweeklyberwick Centraladvocate Centralwesterndaily Thecityweekly Coastaltimes Cobarage Colypointobserver Coomaexpress Cootamundraherald The Courier  Crookwellgazette Dailyadvertiser Greaterdandenongweekly Devonporttimes Donnybrookmail Esperanceexpress Eyretribune Farmonline Theflindersnews Forbesadvocate Frankston Weekly Greatlakesadvocate Goulburnpost Gloucesteradvocate Hepburnadvocate Humeweekly Illawarramercury Theislanderonline Islandofcontrast Knoxweekly Lakesmail Latrobevalleyexpress Lithgowmercury  Macarthuradvertiser Macedonrangesweekly  Macleayargus Manningrivertimes Margaretrivermail Maribyrnongweekly Meandervalleynews Melbourneweeklyportphillip Melbournetimesweekly Melbourneweeklyeastern Melbourneweekly  Meltonweekly Merimbulanewsonline Monashweekly Mudgeeguardian Murrayvalleystandard Muswellbrookchronicle Myallcoastnota Nambuccaguardian Narrominenewsonline Northernargus Northernmidlandsnews Thenortherntimes Northernweekly Northweststar Nynganobserver Oberonreview Penrithstar Portlincolntimes Portpirierecorder Portstephensexaminer Queanbeyanage RHSGnews Theridgenews Riverinaleader  Roxbydownssun Sconeadvocate Southernweekly  Southwestadvertiser  Sunraysiadaily Summitsun Tastamartimes Tenterfieldstar Townandcountrymagazine Transcontinental Ulladullatimes Victorharbortimes Waginargus WAtoday Wellingtontimes Whyallanewsonline WimmeraMailtimes Westernherald Winghamchronicle Wollondillyadvertiser and the West.
Websites include Stuff.co.nz Chairforcengineer Topix  New-zealand Treehugger
Treehugger.com/quote Daylife Hdlns  Actualtechnologydot Congoo Guinness.firstblogfirst  twitter Connect.in  Examiner Allvoices Optuszoo  Onenewspage Ozcrunch  Kiwi247 Allvoices Jorbit Wotnews Businessinsider Friendfeed Terrystechnologypage  Mybiz.optus Endless-sphere Cmblog  Silobreaker 24dunia 360reports 7bay  Connected-community-hackerspace NZcity Sydneycyclist Nobmob Bicycles Mybiz  WD6ezc Cultshit  Iplextra Bicycles  Glosnet Photographymethod Pacc Energymatters LCDdisplayvideos Manlydig Ledstriplights Cyclingcrowd Ata       
      

There was a radio interview with 4BC Queensland Australia on December 26th 2010 as well.

World's Brightest Flashlight  2010
(Note that some images are removed for now pending TV and print media release)
Well this is really an offshoot of the bike project but will have more appeal to the non cyclist.  Using the front array of high powered LEDs to 90,000 lumens attached to a battery pack in a (sort of) flashlight body gives the "muscle" flashlight here.

The front array of high powered LEDs of 90,000 lumens attached to a battery pack in a (sort of) flashlight body  gives the "muscle" flashlight here.
(The flaslight body was empty - you won't pick it up easily at 57 kg when full with batteries included)

 Gallery

Specifications
Front lights: 
    15 x 100 W LEDs (12 white, 1 red, 1 blue 1 green: all approx 6000 lumens)
    1500 W total
    90,000 lumens
    Beam 5 degrees
Power:
    3 x 12 V deep cycle 33 AH batteries
Run time:
    est. 10 mins at 50 A 33 V             
Weight: 57 kg      

Concept
This will have a total light output of 150 W (10% efficiency at 1500 W input). So like a 150 W laser unfocused to whatever beam size so a lot of total output. Sounds dangerous.
However,
A single 10 W light output from a single 100 W electrical input LED focused at 5 degrees is perhaps 20 cm diam (= 0.04 m2) ,at 1m. i.e. 250 W/m2. Sunlight is 1000 W/m2 of which only 20 % is visible. ie 200 W/m2. It comes from a 1 degree source.
Hence at 1m, using back of an envelope figures you will see 11 large (5X) weak suns in a circle and a red, green and blue sun in the middle. You will blink and look away reflexively . You will not get an excessive dose as in a laser which will focus to a point and burn in a short time.
It should be fairly safe but still should not be used irresponsibly.

It is bright but not focused in a manner to cause damage.
I am much more concerned about my 40 mW Blu-ray laser as an eye risk.

It's easy to make a claim of being the "World's best" at anything and rather hard to refute particularly if there is no independent arbiter such as Guinness World Records, particularly if it has not appeared on the net. Of course, sometimes it's comparing apples and oranges. Like mine is not commercially useful, theirs is. Mine can't be focussed well, theirs can. Hence theirs will have higher peak intensity at a distance (candlepower) due to better focus. But mine has more total light output (lumens).

I have now applied to the Guinness World Records as world's most powerful flashlight.  However on further perusal, there is no online reference to any flashlight as being brightest or most powerful. Why might this be? I speculate that the "brightest" handheld light is going to be a laser with unrivalled intensity at 1 mile in a small spot. The beam divergence is so low that it will be hugely bright at a distance if looked at or measured.  So brightness at a distance is not really a good measure of what a flashlight is all about. What about total light output measured in lumens?  This is the best measure of light power output and is in common use, however, it is very hard to measure with a non uniform beam.  Sure it is easier with a source projecting evenly throughout 180 degrees but few light sources are like this. My 90,000 lm is the summation of LEDs derived from manufacturers information when the LEDs are driven to specification.

So what to do?

Construction
Some construction shots.

Take one domestic kitchen bin with defunct automatic lid opener removed plus the top of a domestic rainwater tank. Presto!

Media

x 10^{6 3 4 5 6 7 8 9 10   2  1  2}