Dyno-powered spoke lights

[wednesday]

Hi there, I think having spoke mounted lights for side visibility is a Good Thing. I've tried out the little battery ones but they don't last very long and are generally too fiddly to turn on/off, etc, while the POV stuff like Monkeylectric seems overkill to me, and still needs batteries. Feeding something from a hub dynamo which in exactly the right spot would seem smart, but the only way I can see to do it is with some sort of rotating contact/brush on the forks, which seems riddled with problems. I just had a bit of a mind-explosion on this while procrastinating and thought I would give it a try. So my thought: Wind a small parasitic coil onto the outside of the hub (in the gap between the spoke flanges) to capture some of the energy whizzing about inside (from the AC coil if nothing else) and feed it into some LEDs on the spokes. I don't imagine it would be very efficient, but it would only need a few mA to light up some 5mm LEDs. I've just tried this with 5 turns of fairly thick wire, and spinning the hub axle with my power drill (with a dummy load attached across the main terminals) registered 0.0V on the 200V AC position of my multimeter (the smallest setting it has in AC). I really haven't a clue here, so I thought I'd turn to the collective wisdom of CPF and see if anyone has any bright ideas. Can anyone suggest what I might do to make this work, or if it is even be possible/a good idea? Cheers, Jem

[swhs]

Hi there, I think having spoke mounted lights for side visibility is a Good Thing. I've tried out the little battery ones but they don't last very long and are generally too fiddly to turn on/off, etc, while the POV stuff like Monkeylectric seems overkill to me, and still needs batteries. Feeding something from a hub dynamo which in exactly the right spot would seem smart, but the only way I can see to do it is with some sort of rotating contact/brush on the forks, which seems riddled with problems. I just had a bit of a mind-explosion on this while procrastinating and thought I would give it a try. So my thought: Wind a small parasitic coil onto the outside of the hub (in the gap between the spoke flanges) to capture some of the energy whizzing about inside (from the AC coil if nothing else) and feed it into some LEDs on the spokes. I don't imagine it would be very efficient, but it would only need a few mA to light up some 5mm LEDs. I've just tried this with 5 turns of fairly thick wire, and spinning the hub axle with my power drill (with a dummy load attached across the main terminals) registered 0.0V on the 200V AC position of my multimeter (the smallest setting it has in AC). I really haven't a clue here, so I thought I'd turn to the collective wisdom of CPF and see if anyone has any bright ideas. Can anyone suggest what I might do to make this work, or if it is even be possible/a good idea? Cheers, Jem

Why not just use tyres with reflection?

[wednesday]

Why not just use tyres with reflection?

I do, but they quickly get grimy and no amount of scrubbing restores them. They are visible, I agree, but not like a spoke light.

[swhs]

I do, but they quickly get grimy and no amount of scrubbing restores them. They are visible, I agree, but not like a spoke light.

Well, I almost never clean my tyres and even grimy ones reflect well enough. Shine a flashlight on it or make a picture with flash which are both similar to what happens when a car's headlamps shine on them. I think spoke lights powered via anything else but batteries are just too complex and fragile to give better side visibility, which I think isn't needed anyway. I mean, it's the same as with some other issue in bike lighting: The people who won't see you even when you ride with proper bike lamps and reflection are almost certainly the same people who ride into fully lit ambulances, police cars etc. claiming they 'didn't see them'. Nothing will help against those people (except taking away their driving licence).

[Steve K]

It seems that just sticking a couple of AA nimh's in the wheel would be the easiest solution, but not the most fun.

In the search for how to couple energy from a stationary part of the bike (such as the lower fork blade where the dynamo power is picked up) to a rotating part, I'd suggest looking at all of the fundamental ways to transmit power. The ones that I can think of are to use magnetic fields, electric fields, optical power, electrical current (through brushes as mentioned already), ... did I miss any?

For magnetic fields, you could essentially build a transformer where the primary winding would be a coil of wire coaxial with the dynamo axle, and attached to the fork. The secondary winding would be attached to the dynamo hub flange, close to the primary. The axle would carry a lot of the magnetic flux, but the air gap is huge and the flux would be fairly small. Might be enough to light a few small LEDs.

for electric fields, you'd do something similar, but with capacitors where one plate was stationary and one was rotating. I think you'd need two; one for the "positive" and one for the "negative" wire. Again, the size of the plates will be small and the gap will be big, so you'll get very little capacitance. I'd suggest driving them with a very high frequency AC signal so as to reduce the impedance of the capacitors.

Optical: I envision a clear plastic ring attached to the hub flange. A power LED is attached to the fork and uses an optical waveguide to shine the light into the clear plastic ring. There would be arms branching out from the ring and extending out away from the hub. Surface features would be added to the branches to let light out at desired locations. Challenges: getting light coupled into the ring instead of reflected off of its face, or passing right through.

Obviously, all of these methods have challenges, and I'm too lazy to attack them.

Personally, I have a headlight that is mounted mid-fork, and the spill light does shine on the spokes and rim/tire. It's not a rotating light, but it does enhance my visibility from the side.

have fun!

Steve K.

[BrianMc]

Shine a flashlight on it or make a picture with flash which are both similar to what happens when a car's headlamps shine on them.

Yes, they work great when the line of sight is the same as the reflected light. Unfortunately, they don't become obvious until about a second or even less if a cyclist rides fast, before he/she is directly in front of a vehicle at a right angled intersection because he reflected light doesn't return to the driver. Most of the intersections here are right angles. If the road is angled toward the cyclist, the reflectors and tires show up sooner, and if angled away, later. My opinion is they tell some drivers what it is they are about to hit rather than being a great deterrent. Especially it they turn their head to check the other direction just before you roll into view. Better than not having them, but not by a lot for some intersections. Barely worth their extra weight. So you need active lighting to reach them sooner at night or under low light conditions.

I think spoke lights powered via anything else but batteries are just too complex

I agree, given current dynos. I have been thinking of NiMH AAAs or a single Li-ion at the hub for low rotational mass to power them and get away from the coin cells.

and fragile to give better side visibility, which I think isn't needed anyway.

No reason for the finished system to be overly fragile, though the prototypes will be. Maybe if a problematic method is used like slip rings. A redesigned dynamo with a connector in the wheel would do it.

I respectfully disagree about need. I assume you speak for yourself. Depends on your drivers and what they expect to see. There are those trying not to screw up who can use the help and I would rather give that to them than be hit by them. Many are my neighbors who don't want to injure me any more than I wish to be injured. A lot more slightly incompetent drivers here given our licensing procedures, I'm afraid. ;(

I mean, it's the same as with some other issue in bike lighting: The people who won't see you even when you ride with proper bike lamps and reflection are almost certainly the same people who ride into fully lit ambulances, police cars etc. claiming they 'didn't see them'. Nothing will help against those people (except taking away their driving licence).

I agree there is that subset of completely incompetent motorists briefly, or all the time who are 'beyond help". A group for whom "I did not see you/it/them!" = gross incompetence behind the wheel. We see what we expect to see. If talking on a cell phone they are hardly seeing anything let alone cyclists. So that is where defensive riding and in my case, a powerful aim-able helmet light kick in. I hope I see them in time.

In short, I applaud the desire, but as Steve K said it won't be easy. Batteries for me, I think.

BrianMc

[2xTrinity]

My kludgy but effective way to create side-visibility on my bike is to strap a very bright red flashlgiht to my seatpost aimed toward the saddle, and bounce all that light that off a series of faceted reflectors I've attached to the seatpost rails -- I made the reflectors by taking the plastic fluorescent panel diffuser sheets available for $5/sq-meter at Home Depot, and back-coated with aluminum foil tape for heater ducts ($7 for a big roll at home depot...), then cutting those pieces into trapezoidal shapes, and mounting the pieces with small machine screws to a rigid frame made out of a bent strip of sheet aluminum with tapped holes. This gives me something analogous to a shaped plastic vehicle taillamp diffuser -- which emits a considerable amount of light sideways as well as backwards.

The source is bright (lots of lumens) but diffuse/not glaring like a direct line of sight to a bare LED. It basically looks like a motorcycle taillight with the brake-lights stuck on.

I also rigged a diffuser/shield on my handelbar mounted light so that most of the rays which would normally go above-horizontal and blind oncoming drivers are diffused and reflected sideways.

I'd pretty much given up on the idea of spoke based lights -- my commuter bike (a steel singlespeed road bike) is often locked outside and loose accessories will get stolen. That's why I've gone for self-contained flashlights which can be taken with me, and cheap plastic reflectors/diffusers/secondary optics that "live" on the bike itself... and dynamo hubs and spoke lights with kooky power coupling schemes would just be wrong on my nice 17lb carbon fiber road bike

It seems that just sticking a couple of AA nimh's in the wheel would be the easiest solution, but not the most fun.

In the search for how to couple energy from a stationary part of the bike (such as the lower fork blade where the dynamo power is picked up) to a rotating part, I'd suggest looking at all of the fundamental ways to transmit power. The ones that I can think of are to use magnetic fields, electric fields, optical power, electrical current (through brushes as mentioned already), ... did I miss any?

For magnetic fields, you could essentially build a transformer where the primary winding would be a coil of wire coaxial with the dynamo axle, and attached to the fork. The secondary winding would be attached to the dynamo hub flange, close to the primary. The axle would carry a lot of the magnetic flux, but the air gap is huge and the flux would be fairly small. Might be enough to light a few small LEDs.

for electric fields, you'd do something similar, but with capacitors where one plate was stationary and one was rotating. I think you'd need two; one for the "positive" and one for the "negative" wire. Again, the size of the plates will be small and the gap will be big, so you'll get very little capacitance. I'd suggest driving them with a very high frequency AC signal so as to reduce the impedance of the capacitors.

Optical: I envision a clear plastic ring attached to the hub flange. A power LED is attached to the fork and uses an optical waveguide to shine the light into the clear plastic ring. There would be arms branching out from the ring and extending out away from the hub. Surface features would be added to the branches to let light out at desired locations. Challenges: getting light coupled into the ring instead of reflected off of its face, or passing right through.

Obviously, all of these methods have challenges, and I'm too lazy to attack them.

Personally, I have a headlight that is mounted mid-fork, and the spill light does shine on the spokes and rim/tire. It's not a rotating light, but it does enhance my visibility from the side.

have fun!

Steve K.

Interesting ideas the capacitance idea sounds the most clever to me. Most DC-DC converters are based on chopping DC into a high frequency square-wave, then filtering that back to DC at a different voltage with inductors and capacitors. Half of a DC-DC converter circuit would basically be capacitively coupled from the front fork to the wheel. The amount of RF interference radiating from the wheel would have the effect of jamming the hell out of every wireless cycle computer within a few bike lengths

The optical light guide idea might actually work -- but not efficiently light pipe systems like that are pretty much how vehicle instrument lighting in cars have been designed for decades -- stick a couple lightbulbs inside plastic "donuts" and use total internal reflection to guide the light around. But the light-guide based spoke light, like the vehicle light guides, would have EXTREMELY low overall efficiency. That doesn't matter for back lighting a "check oil" indicator light, etc. but it would be pretty lame to use 5W of LEDs and only get a tiny fraction of that that into the spokes.

TBH the low tech "brush" contact approach, ala old school DC motors, is probably the most efficient and easiest to implement, assuming one was going to create a hub from scratch. I don't see any easy way to retrofit something like this at all though.

[wednesday]

Interesting ideas the capacitance idea sounds the most clever to me. Most DC-DC converters are based on chopping DC into a high frequency square-wave, then filtering that back to DC at a different voltage with inductors and capacitors. Half of a DC-DC converter circuit would basically be capacitively coupled from the front fork to the wheel. The amount of RF interference radiating from the wheel would have the effect of jamming the hell out of every wireless cycle computer within a few bike lengths The optical light guide idea might actually work -- but not efficiently light pipe systems like that are pretty much how vehicle instrument lighting in cars have been designed for decades -- stick a couple lightbulbs inside plastic "donuts" and use total internal reflection to guide the light around. But the light-guide based spoke light, like the vehicle light guides, would have EXTREMELY low overall efficiency. That doesn't matter for back lighting a "check oil" indicator light, etc. but it would be pretty lame to use 5W of LEDs and only get a tiny fraction of that that into the spokes. TBH the low tech "brush" contact approach, ala old school DC motors, is probably the most efficient and easiest to implement, assuming one was going to create a hub from scratch. I don't see any easy way to retrofit something like this at all though.

I agree, great ideas from SteveK, but they fall back to the old air gap between frame and spoke. I'd hoped to harvest some energy direct from the hub before it gets to the frame side, but it seems like this isn't a goer. After a bit more research, it seems to me the people who have best crossed that frame/spoke air gap are reelight/freelight (e.g. http://www.freelights.co.uk/howmake.html). I wonder if, forgetting about the hub dynamo for a second, a mini freelight that is inverted, with a north/south magnet arrangement on the fork and a tiny ball bearing magnet inside a small coil connected to a couple of 5mm LEDs on the spokes might be a workable arrangement. I imagine on the front wheel it would put out a (shortish) arc of light as it passed the fork each revolution, which would probably stand out pretty well. I've ordered some small ball magnets and will post again if the results are at all promising. I also wonder if putting a flat magnet on an axle inside the coil would be more efficient, although being more bulky. Does anyone have any thoughts on the best layout for this?

[Steve K]

I thought about the Reelight too, but had misgivings. If it flashes each time it goes past the magnet, then it'll appear to be a fixed flashing light, which is not the original intent of a spoke light. OTOH, if it flashes continuously as it spins, then it might be okay. The flash rate is fairly low, I understand, so it's still a long way from a semi-constant light that would trace out a circle as the wheel spins.

All of these methods are limited by the difficulty of coupling energy across a significant air gap, especially when the wheel needs to be easily removable and doesn't have a precisely controlled location when it is installed. The optical method isn't really limited by the air gap, but is susceptible to dirt and dust that could block the transmitted light.

Are the AA NiMH's starting to sound more attractive??


Steve

[wednesday]

Are the AA NiMH's starting to sound more attractive?? Steve

Ah, but where's the fun in that? But seriously, glad you think the reelight approach might be viable. I'm hoping for a short burst of light, perhaps helped by a cap, that will last at least a quarter revolution. I imagine this would be pretty useful, especially if there were two magnet positions, one on the fork and another on a mudguard strut. As I said, I'll report back if I make any progress.be

[panicmechanic]

Wonder if you know LED scooter wheels like these? There is a miniature claw pole generator inside, like a tacho signal generator. Coil and claws as well as LEDs are cast into the transparent wheel, a multi-pole magnet ring sits on the axle between the bearings. So, it's an inverted hub dynamo, and should be almost ready for the task. Only needs a little higher output in lower rpm.

[wednesday]

Wonder if you know LED scooter wheels like these? There is a miniature claw pole generator inside, like a tacho signal generator. Coil and claws as well as LEDs are cast into the transparent wheel, a multi-pole magnet ring sits on the axle between the bearings. So, it's an inverted hub dynamo, and should be almost ready for the task. Only needs a little higher output in lower rpm.

Wow, fantastic! There isn't really a conversion option to put these on a bike though, is there? Or am I missing something? Trainer wheels?

[post tenebras]

Wonder if you know LED scooter wheels like these? There is a miniature claw pole generator inside, like a tacho signal generator. Coil and claws as well as LEDs are cast into the transparent wheel, a multi-pole magnet ring sits on the axle between the bearings. So, it's an inverted hub dynamo, and should be almost ready for the task. Only needs a little higher output in lower rpm.

I've always wondered why nobody has adapted this technology for a spoke or valve light. I have a 12 year old Razor scooter with these wheels and they still work perfectly.

[panicmechanic]

Bump
Just found the Kinetron website. They make miniature generators like a pedal light, they might have a solution for a contactless wheel-integrated light.

[wednesday]

Yowsers, thats a great idea. A pedal seems like a smarter place to put a generator for rotating side lighting than the wheel (assuming the wheel hub is already taken). Unfortunately I haven't got very far with the spoke generator. I've got some tiny ball magnets and am planning to wind a coil around something appropriate and stick it in a spoke reflector, but haven't tracked down any magnet wire yet. My ability to plan new projects always far outruns the opportunities available to me to complete them. As for why we don't have flashing bicycle hubs yet, as far as I can see the entire bicycle industry is focused on catering to a) Cadel Evans or b) people that only ride in the park. Good ideas wither on such stony ground.

[panicmechanic]

Magnet wire can be found in discarded electric appliances of all kinds. Small electric motors, bike sidewall dynamos, vacuum cleaners and whatnot. For motors, the rotating part is often not useable as a donor, since the windings are glued.

If you cannot find something in your household, a scrapyard will have a collection of motors, and someone selling household items like mixers will likely be happy to give defective items away.

Of course, magnet wire is also available in shops, but that would be boring, wouldn't it

Btw, my dynohub features iron from a toroidal transformer and copper wire from a washing machine.

Whatever you find, keep track of the number of windings you use as well as the effective wire diameter (without the isolating paint). This will help evaluate what works and what should be altered.

Have fun!

[fugue137]

First, as for the need for this, there's an article on Sheldon Brown's site in which John Schubert explains why he thinks wheel reflectors don't work. Because I'm new here and the moderators sometimes don't approve posts from new members that contain links, I will just say that the search "why reflectors don't work" will get you there immediately. (Nonetheless, by the law of safety in numbers, you _should_ at least use reflectors: even if they're completely useless to you in any given collision scenario (which probably overstates the case a little bit), they will be seen by cars that are not about to hit you, and raising awareness of cycling is a huge factor in cycle safety.)

Yowsers, thats a great idea. A pedal seems like a smarter place to put a generator for rotating side lighting than the wheel (assuming the wheel hub is already taken).

It IS a great idea, but I'm also having less enthusiastic thoughts:

• Difficult to retrofit into your favourite pedals. How many people here are members of one of the following cults? Shimano, Crank Bros, Bebops, Frogs, TIME, etc... Would you be willing to switch pedal types just for these lights?
• Something that still illuminates when you are coasting to an intersection (the most likely time to need a side light, methinks) would be... a lot of electronics to cram into a pedal. It needn't go in the shaft, of course, but still, I guess it would require a platform-pedal design.
• I think that when judging collision courses, a spoke light gives a decent indicator of vehicle speed, whereas a pedal light probably does an absolutely terrible job! (Because the speed of rotation looks like it _should_ be correlated with vehicle speed, but is not, which is worse than giving no indication at all.)

As for why we don't have flashing bicycle hubs yet, as far as I can see the entire bicycle industry is focused on catering to a) Cadel Evans or b) people that only ride in the park. Good ideas wither on such stony ground.

*sigh* yes. At least in the USA. But things are changing, slowly... and honestly, a startup can go a fair distance by selling to the entire populations of Seattle, Portland, and Boulder.

Finally: I too was considering hacking a Reelight to mount the light on my spoke, but haven't gotten around to it yet. I'm glad I'm not the only one thinking along these lines!