Designing good daytime rear commuter lights

[BrianMc]

First 'good' is relative. For many the Planet Bike Superflash or Princeton Tec Swerve, and others of the type are 'good' enough. Some have upped the ante to the Dinotte or Klite tailights.

Here is a video taken in daylight with three Superflashes. One steady two flashing, one of the flashing ones on the back of the helmet (Camera Sony Digital 8 Handicam):

<embed src="http://img515.imageshack.us/flvplayer.swf?f=Msuperflashday" width="640" height="380" allowFullScreen="true" wmode="transparent" type="application/x-shockwave-flash"/>

The darker parking lot pavement is about 50' (15 m) from the camera. This is the distance for cars to the right or left to yield. This is full behind. Side views are less bright.

Cameras are notorious for failing to pick up brief flashes of light if they are out of sync and the flashes are too short a duration. This camera adds a lack of resolution on top of that. So I repeated with all three Superflashes on full and a long duration yellow Xenon strobe behind the Superflashes on the seat stays. I carefully aimed the lights and it is a slightly higher camera position. This is 'everything I got'.

<embed src="http://img714.imageshack.us/flvplayer.swf?f=Msuperflashday2" width="640" height="380" allowFullScreen="true" wmode="transparent" type="application/x-shockwave-flash"/>

The helmet light was 'On' showing how beamy the Superflash is. Head tipped -not seen from behind, side yes. The lights show better on full. Mostly a camera effect, though the intensity of a flashing light is less on average. We count on the simulated movement of the flash to catch the eye. That works under lower light conditions. Without a high speed vieeo camera, I can't record it. I may have the only Superflashes in town, so I have no visual reference to how they work on bikes as a motorist myself.

Bear in mind I am riding away at under 20 mph (32 kph) the closing speed of a car going about 10 mph over a 30 mph limit (16 kph over a 50 kph limit). I ride roads where the closing speed will be up to 3 times as fast.

I first suspected that my three Superflashes were not doing what I thought they were for me when the addition of an ANSI vest. It has the same 'Limon' color of the Pearl Izumi vests and other clothing but has reflective stripes for night, too. You can see that vest for 1200 feet. Can't identify the bright spot, and will likely ignore it, but you can see it. I got a lot better respect and fewer issues with motorists after adding the vest.

Likely in sun, particularly in low sun, and with an upright or more upright riding position, the clothing is unsurpassed for maximum distance. I suspect we cannot do as well by any legally bright taillight. For cloudy days, shadows of buildings and other day situations where the sun isn't helping light up the clothing as much, I want to see 'brake light bright' lights. Hopefully, such a light shows up even with the sun low and shining on it. The goal is to give motorists on faster speed limit roads more warning time. I won't need that much power at night so two levels would be good. Side visibility to 180 degrees or a bit more.

Troutie allowed me to post these pics to add fuel to the fire (Many Thanks)::wave:

"I have had a few requests for a good tail light. 1 red XPE would make a good tail light and they can be run at 700 ma I am not too sure on daytime vis at that power setting.

Some pics: daylight beam shots. (Ha Ha!) (Brian: I leave them in the album for now, anyway)

http://s199.photobucket.com/albums/a.../REAR%20LIGHT/

The sun is just setting in our garden so it is still pretty bright. There are 3 red xpes @ 700ma with no optics in the first two so throwing light everywhere. The next 3 have a experimental botched optic. The last has three Carclo 10415 10 mm eliptical optics on the three xpes.

Here is a video of the 3 red xpes @ 700ma with Carclo 10415 10 mm eliptical optics:"

http://www.youtube.com/watch?v=vAjSIlZqz7c

I would say just 1 red XPE and a 20 mm eliptical optic in a housing with clear sides would be more than enough with Hi 700 ma and low 200 ma for day and night running."



So how good are your lights in the day? Any mods or 'beasts' out there already? Of course the Superflashes 100 hour runtimes on flash with two AAA's will have to go. Pity.



PS: Daytime 'beamshots' from as many angles as you deem fit to take are most welcome. Let's shed a little (or a lot) of (red) light on the problem!

 

[JustRidingAlong]

I had to view the videos two or three times before I could pick up the flashes.

One thing I'd like to know is how effective 45-degree LEDs would be when combined with a "straight-back" light.

 

[BrianMc]

I had to view the videos two or three times before I could pick up the flashes.

Exactly.

So we have three options:
1. The Camera was incapable of properly recording the actual flashes of the lights in either intensity or duration.

2. The lights are really that weak especially the 1/4 watt ones and in the daylight, ineffective.

3. Not quite that weak and the camera is putting them in a bad light naughty: couldn't resist, felt we needed to lighten up some, Oops, another one!)

On the other hand, Trouties' three XRE's in flashing mode (a Taskled driver?) was easily picked up by his camera.

So we have lumens and we have pulse interval and pulse duration variables. Plus the camera as a filter on what they really look like.

Are the best flashing patterns for people the very ones that are hard to record?

Here is another Superflash video:

http://www.candlepowerforums.com/vb/showpost.php?p=2562123&postcount=1

This video seems to show that the 1/4 watts are sometimes missed in night shots too. So the camera effect is real with the pulse width and frequency of these lights. Am I seeing the lights and their reflections?

So my camera may be especially poor in this application.

One thing I'd like to know is how effective 45-degree LEDs would be when combined with a "straight-back" light.

Aiming is another variable since these have fairly narrow best viewing angles. Seems to be a bad thing in a taillight IMHO.

We could give up the Superflash as a bad job in this application and forge ahead with a DIY design. However, with as many Superflashes in use as there are, it seems a good idea to make sure of our evidence.

1. Are my results repeatable with a better camera? OK suggestions on what to borrow or rent?

Or please repeat my test with your camera and bike. Even a single Superflash. Post the recording and report the camera make and model. Maybe we can find a decent tool for the job.

2. Does an angled array of Superflashes and the like work?

I can do three Super flashes and a Swerve at one time. Let's say the Swerve and a Superflash straight back and two Super flashes to the sides. One run at 45 degrees another at 90. But I need to resolve the camera issue.

3. Other people's testing or observations. Has anybody seen anyone riding in the daylight with three or four Superflashes splayed out with the side ones at an angle? If so how well did that work for you as a motorist? Anecdotal evidence has some value, here.

 

[trout]

OK just for research here are 3 vids

all the same led and optic a single red XPE and an unknown 20 mm eliptical optic

this one running at 700 ma and 15 metres from the sony handycam miniDV on auto in my garden with no other lighting
http://www.youtube.com/watch?v=YwCyt2e8nqg&feature=channel

this one is on the street with street lights the distance to the same camera is 65 metres or large strides again at 700 ma drive

http://www.youtube.com/watch?v=uuxLIYySS5s&feature=channel

this one is same distance but 200 ma drive curent

http://www.youtube.com/watch?v=YXyp1hG_7NY&feature=channel

I will do the same shots in daylight in the morning

this is the test rig

 

[BrianMc]

Interesting. I LIKE 700 mA. Not surprised there.

See you in the morning.

 

[1 what]

Hi BrianMc,
Have a look at:
http://www.candlepowerforums.com/vb/showthread.php?t=264640
I'm a bit paranoid about the vehicles behind me that I can't see and therefore can't avoid. Therefore I build very bright red lights and consider side-on visibility.
The tail light in the post makes plenty of light in the day especially when it runs on strobe.
Much to my relief I've not been picked up at night for a too bright tail light.
I'm in the final stages of a different design for another of my bikes and will post it here in a couple of weeks.

 

[BrianMc]

Hi BrianMc,
Have a look at:
http://www.candlepowerforums.com/vb/showthread.php?t=264640

I did. I found it interesting. It so hard to judge these up close. or straight on only from the back. Pe2er used the same LED AFAIK in his lights (hope you don't mind adding you to the debate, Erwin):

http://www.candlepowerforums.com/vb/showpost.php?p=3360038&postcount=12

What I'd love to see is daylight shots at distances and angles especially movies of flashing or walking through angles at distances or the bike moving away from the light method, that I used.

These Superflashes had me fooled. (Although the jury is still out on how poor they are, they are not enough.) I thought the motorists were being difficult (blind). They couldn't see me in spite of my best efforts until the vest. I use a mirror to watch upcoming, but sometimes I forget and one sneaks up. It woud be nice to help them not screw it up and run over me! Defensive cycling at its best! I get far more passing room taking the lane and wearing the vest.

QUOTE=1 what;3362113]I'm a bit paranoid about the vehicles behind me that I can't see and therefore can't avoid. Therefore I build very bright red lights and consider side-on visibility. ... I'm in the final stages of a different design for another of my bikes and will post it here in a couple of weeks. [/QUOTE]

Apparently we share a common interest: protecting our hides! Actually the motorists don't want the hassles of running us down, either.

I came across a video of the Dinotte on Youtube near Trouties' links above, but they only did on axis in a gym complex under the lighting.

http://www.youtube.com/watch?v=mjlDJ7pUc8Q

The side view and angled views are very important. The Superflashes are NOT directional in side view, they just aren't very powerful in those angles. Great at night, though.

The Luxeon is an older LED, so I assume with about 100 lumens, the red-orange XP-E at 700 mA is the brightest and most efficient LED for the task at present. Troutie will have daylight movies in a few hours from now with that LED.

Be fun if GreenLED and some of the folks in the threads of the Superflash/Swerve who are still active would drop by for a comment or three. I'd love their take and more about the winter morning ride conditions where the Superflash did so well.

 

[trout]

as promised but a bit early for the rising sun

straight on 65 metres

45 degrees to the light

90 degrees to the light

 

[Offroad'Bent]

So, someone going to come up with a nice, light, self-contained 3AAA or 2AA rear light with an XPG at 700ma and 200 for night, with a nice easy switch?

Or at least instructions for dummies as to how to wire and silicone the emitter and optic on to an existing LED blinkie housing?

I'd take a couple and throw away all my Superflashes and Mars 4.0.
:wave:

 

[pe2er]

... Pe2er used the same LED AFAIK in his lights (hope you don't mid adding you to the debate, Erwin): ...

I Don't mind In fact, I am curious to see if my tail light has any value in broad day light. Hope to be able to shoot some video of it in action later this weekend.

 

[BrianMc]

as promised but a bit early for the rising sun

If I'd stayed up another half hour I'd seen them before bed! All at 700 mA, right? Great dispersion. Looks like there might even be something to see at 120 degrees. A very good candidate.

1 what mentioned the "too bright" aspect. Firstly, it will be very much a local enforcement attitude matter. Here they are so happy to see a well-marked bike that I would be surprised at even a warning, unless the light was greatly affecting the vision of approaching motorists, singing paint melting the pavement, you know, the usual for lazer beam weapons.

I did Google the regs. The smallest light area with a specification is for
motorcycles and scooters turn signals. S5.1.1.25 Each turn signal lamp on a motorcycle shall have an effective projected luminous lens area of not less than 2258 square mm. (3 1/2 square inches). Or 1.9 inches a side for a square and 2.2 inches diameter for round. A bit larger than we are thinking. Does size matter (for conspicuity)? We may need a much brighter light because of small size. I asume SAE researched that to come up with the regs. No idea how to ferret that out without an Enginerring School library close to hand. It will be pre-internet by decades.

The next is from USDOT standards listing minimums and maximums for lights on vehicles.

FIGURE 1B-MINIMUM AND MAXIMUM ALLOWABLE CANDLEPOWER VALUES

LAMPLIGHTED SECTIONS___ 1_______ 2 _______3

Stop__________________ 80/300 ___95/360_110/420
Tail ___________________12/-_____185/ -__200/-
Parking_______________ _20/125___ - _____-
Red turn signal __________80/300__ 95/360_110/420
Yellow turn signal rear__ _130/750_ 150/900_ 175/1050
Yellow turn signal front___200/-____240/-___275/-
Yellow turn signal front _3500/-____ 600/-___685/-

Interesting that yellow rear turn signals are allowed max values 2.5 x as large as stop lights and more than red turn signals. The flashing either makes brightnees less of an issue, or more likely, you need a brighter light to compensate for the off part of the flashing mode. Maybe red stays on the retina longer or puches through fog and rain better. The tape of on versus flashing Superflashes supports the flashig idea. Candela measure density of light per unit area and we have light output in lumens. So if we spread the beam more we reduce the lumens. If we flash we can increase and may NEED to increase the lumens. Does that change if we have a larger area light?

I think Pe2er's comparison to a car light is a good enough crude estimate of whether we a grossly over the allowances for motor vehicles and so likely to attract unwanted attention. We should compare to the brightest. Brake lights and rear turn signals at 90, 45, and 0 degrees to get the best feel for whether we are approachjing "too bright". I would think it would have to be a real butt-head of an officer who would cite after you told him they had been evaluated as less that other legal lights on the road. Cost them a bit to prove it, actually Could be fun, if aggravating.

Further, Bike lights are specifically NOT listed in USDOT which are the same as Canadian regs. Europe and Australia? YMMV. So the regs don't appear to apply, as in the Pirate Code of Pirates of the Carribean, more of a guidline, really. So since safety is involved, close is likley good enough.

I'm not too worried about being too bright, actually. Nor overly fussed, if it is. Hurting a too-close motorist's eyes while stopped for a traffic light, versus having them recognize me in ample time to make good decisions and not hit me is no contest.

 

[pe2er]

Here is my commuter bike in broad daylight on a sunny day in Holland.

http://www.youtube.com/watch?v=IQXEMtPJxGk

Front:
Fork crown: 3 Watt XP-G steady on.
Handlebar: Magicshine MJ-808 on strobe.

Rear:
Below reflector: 1 Watt Red Dx spotlight with 120° Dx diffuser steady on.
To the right of the Dx spotlight: PB Superflash on strobe.

I think Pe2er's comparison to a car light is a good enough crude estimate of whether we a grossly over the allowances for motor vehicles and so likely to attract unwanted attention.

I Need to redo this comparison as the garage determined that the lights on my car were on reduced intensity some time after taking this photograph. The repaired something and they now should be brighter.


PS: Brian, I did a testride in this and think I am in love... http://i42.tinypic.com/25fll48.jpg

 

[BrianMc]

Here is my commuter bike in broad daylight on a sunny day in Holland. To the right of the Dx spotlight: PB Superflash on strobe.

Good video. It isn't as easy to do as it looks.

I had to watch four times to be sure of the Superflash. So either its flash mode gets sampled poorly (front or trailing edges if at all) by video cameras or it is pretty ineffective in daylight.

As you predicted (between the lines) and 1 what stated straight up, the 1 watt Luxeon from DX is doing a good job for you in daylight That was about 9 car lengths and say about 3 m per length so about 30 m or 100 feet. About 1.5 seconds of closing time at highway speeds and it still looked good for at least another 100. So 3 seconds plus. Not bad.

Well, unlike the road front beam lights, we seem to be talking readily availaible DIY technology, and only deciding which of two LED's (Luxeon or XP-E), which lenses, how big, and power level(s). The long life on three AAA's comes at a cost in the Superflash. They could use a Superflash Day model.

Headlight: Confirmed (IMHO): You want floody light in a daytime front strobe. A thoroughly controlled hard cutoff beam doesn't have enough side spill to be of use as a 'be seen' light other than to alert a head-on driver. That is something you questioned in the road beam thread, Erwin. Your surmise appears to be correct. The 500-1000 lumen range is fully up to the task.

So, I need a snoot and a non snoot bezel so I can change out for day/night use or different lights. On strobe, I would think the MJ-808 should run cool enough to be towards the 650-700 lumens on the pulse (900-20% seems about right for max.). It was NOT too bright even looking straight at it. Spreading the beam helps the angle of effectiveness and the brightness. (So much for the complainer you had, Erwin. You now have documented it is NOT too bright.) If a headlight is a bit bright head-on, that's fine by me. Alll the better to tell thje motorist emphatically to get back over!

 

[BrianMc]

I Need to redo this comparison as the garage determined that the lights on my car were on reduced intensity some time after taking this photograph. The repaired something and they now should be brighter.

Don't you hate it when benchmarks in testing prove unreliable.

Like the camera itself? :sigh:

PS: Brian, I did a testride in this and think I am in love...

As OP, I think bright orange-red velomobiles fall in the category of conspicuity and on topic. I'd prefer that yellow green hue of Pearl Izumi and my ANSI vest for maximum punch, but man, that has to show up better that oxidized aluminum. So after selling your refurbished VM, how many years of commuting will pay it off? Of course the safety aspect of the color and easier repair of damage also factors in. Especially safety. Unless they want to collect on life insurance spouses like safety a lot. Good luck! (Oh and it is better than a mid-life crisis sports car: you can't pick up chicks and I don't mean Coots, either).
:twothumbs

 

[pe2er]

... So after selling your refurbished VM, how many years of commuting will pay it off? ...:twothumbs

Pay it off? It costs more than my car so it will probably have paid for itself... never Should I let that stop me? (That's a rhetorical question...)

 

[BrianMc]

Pe2er: Can't take it with you.


Found this old thread and still going through it.

http://www.candlepowerforums.com/vb/showthread.php?t=147209

Only part way though it myself. a replaceable lens for side view and inevitable clumsiness remains a design issue.

Dinotte's 400L rear light is circa $200 plus shipping and the 140L with battery and charger about $150 and shipping. Good in a straight line. So not the answer.

The Nova Bull 12 v nominal (10-16 v) flashing light is three Red Luxeons with eight flash modes was about $80, first retailer I looked at had it for $40. There is a video of it compared to the Dinotte but the conditions were less tahtn taxing, but appeared close. My main battery is nominal 12 volt 10 min and the draw is 0.25 amp.

http://www.youtube.com/watch?v=sGdZPf69H84&NR=1

Note a higher (more frames per second) camera ws used. Much better at 45 degreees than the Dinotte and the new ones have a wrap around lens cover for decent 90 degree in theory. Dinotte pushes their R/O Lux III at 1.4 A for 140 lumens according to this thread:

http://www.candlepowerforums.com/vb/showthread.php?t=177631

I assume about a 3.7 Vf so at 1.4 A about 5 watts. The XP-E R/O is 700 mA at 2.3 Vf for 1.61 Watt and 120 lumens (67.2 at 350, 1.8 x at 700)

More reading to do and riding!

But the sale price on the Nova Bull looks tempting.

 

[pe2er]

Nice to see you reached the same conclusions I had about my movie.

Here is another video. Same setup, one hour before sunset (20:00 local time). MJ-808 on full bright:
http://www.youtube.com/watch?v=kTtnRk1D33g

Same video, with MJ-808 on strobe:
http://www.youtube.com/watch?v=ilusHTg8uD4

Previous video was from around noon.

 

[BrianMc]

Nice to see you reached the same conclusions I had about my movie.

It's obviously two great minds! :rock:

Here is another video. Same setup, one hour before sunset (20:00 local time). MJ-808 on full bright:
Same video, with MJ-808 on strob...from around noon.

You show up much better (lighter clothes, lights), your bike fits better, and you ride better than the preceding cyclist, though I like watching her better, but I digress. Where are those pretty scantily dressed cyclists when you have a boring video to record? :naughty:

No issue with light for the headlights other than no side visibility. I used the helmet light exactly the way people mention: looking right at a car about to launch into me. So IF you are paying attention the front side lighting is moot if you have a helmet light and aim it when and where needed.

PBSF seems a waste of space in the daylight, actually. Hardly notice it side on. I have three. Makes me sad. So the drivers haven't been as dumb as I thought. Very encouraging when you consider their driving from that point of view. Back to the drawing board.

Is the DX the red XR-E P60 module? And how much power is the Luxeon getting? 1.4 amp like the Dinotte?

The side lighting of the tail lights isn't much help until you are almost dead ahead. Not a much help that close. Likely better at a distance with bike and car approaching an intersection with clear line of sight to each other.

No helmet? Oh, and I have a bigger handlebar...mustache!

But a bit more spare tire. :sigh: A work in progress. A 20% reduction is progress.

 

[pe2er]

... though I like watching her better,...

Now you hurt my feelings

...No issue with light for the headlights other than no side visibility.

Same problem here. I Am wondering if it is better to use separate front lights for daylight visibility. Look at cars. The trend is to not use the main lighting in daylight, but instead use separate LED lights. Most retrofit lights are simple LEDs without optics, providing better visibility than the main lights that are designed to give a minimum of glare.

Well, you noticed the result of that with your main lights.

Is the DX the red XR-E P60 module? And how much power is the Luxeon getting? 1.4 amp like the Dinotte?

No, it is not. It is a 12V 1 Watt LED spotlight, Dx sku 05325.
It consumes a bit more than 1 Watt (do not recall the exact value).

In my experience, 1 Watt is about the maximum limit for tail lights in night time. It is, judging by the video's, barely enough for daylight visibility. I Think a 3 Watt RED LED would do better (in daylight). Or maybe add a second 1 Watt LED to the same tail light - easier to do as I have a second one in stock.

The side lighting of the tail lights isn't much help until you are almost dead ahead. Not a much help that close. Likely better at a distance with bike and car approaching an intersection with clear line of sight to each other.

Also of help in situations with idiots like this: http://www.youtube.com/watch?v=lC4UVQVWA3I

No helmet? Oh, and I have a bigger handlebar...mustache!

Well, you beat me there, yours is bigger Normally, I do wear a helmet. But only when commuting.

But a bit more spare tire. :sigh: A work in progress. A 20% reduction is progress.

Well, you beat me in the spare tire department then also

 

[1 what]

Interesting discussion:thumbsup:.
I can only do low quality videos but this still is of the light in my earlier post at about 120 meters (in daylight).

As you can see the light is clearly visible.