Should I go dynamo, or stick with batteries? And if dynamo, DIY or commercial?

GlowBike

Newly Enlightened
Joined
Jan 22, 2003
Messages
24
Location
Portland OR
I'm going to be getting a new wheelset for my commuter bike, and am mulling over the possibility of going for a generator hub. Although I don't mind recharging my battery pack in the evening, there is some attraction for never having to do that and for always having lighting available no matter how long the ride.

What I'm using today: My current bike headlight is an XM-L using a P60 drop-in module from nailbender. Rather than putting it in a flashlight, I have it in an old TurboCat (MR11 based) bike light head, powered by an external 4.8v (4200mAh) NiMH pack. It has 3 modes (100%, 40% and 5%), which gives it a lot of versatility for various cycling situations. This light is GREAT for mountain biking and puts out plenty of light (supposedly 600-700 OTF lumens) but the beam is pretty wide for onroad cycling. Not as bad as the MagicShine, but I can tell it's wasting a lot of light above and to the sides, and the throw is not great: too much of the light is hitting the ground 30' in front of me, and not enough 100' away. Even if I didn't switch to a generator, I might keep this light for mountain biking but get something with more throw (XP-E based, maybe) for commuting. By the way, I have taped a "hood" (shroud) on top of the light to provide a sharp cutoff and avoid "dazzling" people when I'm on the MUP.

My existing light is not perfect, but it IS bright. If I switch to a dynamo system with one of the better LED-based lights (such as the B&M IQ Cyo or the Philips Saferide 60), will it light up the pavement at least as well as my current system? It sounds like the answer to this question is probably "yes", but it's hard to get an objective opinion from someone who's actually compared them with mainstream battery lights that don't have a shaped beam. The response I always get is that these lights are awesome, but that's always coming from people who were used to the old halogen genlights, which of course sucked in comparison. I do understand that although the best genlights have less total output (250-300lm from what I can gather) than what I'm running, they do a better job containing the beam and concentrating it where it's needed, with the optics concentrating more of the light further away for better throw. So I suspect they're going to light the roadway as well or better than lights such as I have now, but it's really hard to get an answer.

Next question: if I go dynamo, should I just get a commercial gen light and be done with it, or consider going DIY? I've read about how gen hubs can actually kick out quite a lot more than their rated power at higher speeds (and I do ride at high speeds in the dark quite a bit, since I have to descend 500-800' out of Portland's West Hills on the way home, and the drain of powering this much light from a generator is not a concern). I'm comfortable with basic circuitry but not too much in the way of electronics -- maybe wiring up a bridge rectifier and a capacitor or resistor or two, but nothing much more advanced than that. If I had fairly simple circuit and an array of, say, 6 XP-Es (I understand the generator can only provide 500mA so you need more emitters to fully take advantage of it), could that potentially light up the pavement at higher speeds better than a B&M Cyo, or only maybe about the same? (If I did this, I would probably at least go to the trouble of wiring up a bar-mounted switch so I could shunt out 2 or 3 of the LEDs for better low-speed performance). Unless the DIY system would be substantially brighter on the pavement, I'd probably inclined to save myself the trouble and buy a commercial light.

I know that's a lot of complicated questioning there, but ... well, it's a somewhat complicated problem. Thanks in advance for any help you knowledgeable folks can provide.
 
Last edited:

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
There must be a shop in Portland that has a bike sitting around with a good dynamo & LED lamp?? Doing a test ride will be difficult since the sun sets so late. Too bad it's not January. :)
Maybe something can be worked out so you can do a night ride with one of the shop guys on the dynamo/LED bike?

Personally, I'm not much help. Never had a chance to see or use a commercial LED light and have always built my own. A basic headlight with two Cree's is simple enough, needing just a bridge rectifier and a supercap for a standlight. Maybe use 3 Crees if you don't need to ride slow. Making your own light is a fair bit of work, and I've heard from other folks who built their own that they would just buy a commercially made light next time. If you don't really enjoy doing this sort of stuff, it's getting harder and harder to justify doing the DIY thing.

good luck,
Steve K.
 

Lynx_Arc

Flashaholic
Joined
Oct 1, 2004
Messages
11,212
Location
Tulsa,OK
when I got my first bike I bought a generator headlight that was a gem. It took batteries and had two bulbs in it and if you turned it on the second bulb lit up but typically shone in front of the bike mostly downwards not way out in front. When you started riding and got to a certain speed the battery light switched off and the primary bulb lit the road off generator power. I guess they still make lights like this but maybe not two separate bulbs with LEDs and fancier circuitry you could have one bulb and perhaps rechargeable batteries and charge it so you would not have to replace batteries on occasion.
 

GlowBike

Newly Enlightened
Joined
Jan 22, 2003
Messages
24
Location
Portland OR
Thanks for the reply!

Clever Cycles in Portland does have dynamo-equipped bikes, as well as an in-store display with a generator-equipped wheel and several generator lights that you can play with. Of course in a store it's pretty hard to tell how a light is going to be in terms of throw and illuminating the pavement. Also, the best light they've got is the B&M Cyo "R" model, which has additional nearfield illumination versus the standard "sport" model (with the penalty of ~30% less distance illumination). From absolutely everything I've been able to read, I want the "sport" version, and they don't have it.

They have offered to let me test ride their bikes, but the latest they're open is 8pm (on Thursdays) but as you've pointed out it's no longer dark at that time, so a test ride won't do me much good. Too bad indeed that it isn't January.

I might at least swing by there with my headlight setup so I can do an in-store comparison. Even if it's just lighting up carpet in a store, if that "R" model can even get in the ballpark of my light -- in terms of illuminating the showroom floor from the other end -- then knowing the "sport" model is considerably brighter might just be enough to push me into the generator camp.

I think you're exactly right about it being hard to justify full DIY, at least for me. I don't mind the semi-DIY approach I've done so far, and If I do stick with batteries I'll probably continue down that road with a dual Cree setup (maybe an XP-G for a broader beam close up, and an XP-E aimed slightly higher for distance). But that approach won't work with a generator hub, so in that case I'd likely buy the B&M and be done with it.
 
Last edited:

unterhausen

Enlightened
Joined
Jan 16, 2009
Messages
275
dynos are the best thing that ever happened to my cycling. I always have my lights with me, and don't have to worry about charging. I used to use a triple XP-E that I made myself. It had a voltage doubler. It was never all that satisfying, and then one of the emitters blew, seemingly due to water. I was riding with someone that had a Cyo, and it was so much better than my light that it was embarrassing. And even though my light wasn't that great, it still angered all the oncoming drivers. I bought a Supernova E3 pro asymmetric, and I'm really happy with it. Any of the high-end dyno lights will probably put as much light where you need it as your xml.

I like building lights, but I get in a state of indecision about what to do next. It drives me nuts. I have too many projects and not enough time, so the fact that commercial lights are better for the kinds of riding I do pretty much made me gave up on DIY.
 
Last edited:

tomgopher

Newly Enlightened
Joined
Nov 27, 2011
Messages
12
a) The West Hills commute rules.
b) Get a dynamo light. The set+forget convenience of a dynamo light is awesome.
As far as commercial lights, I have a B&M Lumotec IQ Cyo senso plus and while it's excellent for going slowly, I find it to be a bit lacking in the going fast in pitch-black department. I also find the fancy beam to be quite limiting when going on dirt roads at night. I think stock commuter lights are great if you want to save eyes of oncoming cyclists, but they are a waste of time for maximizing the amount of light in front of you.
I'm currently building an MC-E(4dies to maximize brightness at 0.5A) dynamo light to replace the B&M light for dark + fast rides(like werewolf hussle). I suspect that one can achieve a happy compromise by designing a light with a flip-down reflector half to shield eyes of oncomers at busy times.

I'm hoping my dynamo MC-E light with a 50mm reflector(http://www.dealextreme.com/p/alumin...-for-cree-mc-e-led-emitters-53mm-27-6mm-17384) turns out similar to http://www.dealextreme.com/p/tb-d7-...-5-mode-white-led-flashlight-1-x-18650-110262 (my current favourite bike light for dark paths). Bench-testing is promising so far.
 

jdp298

Newly Enlightened
Joined
Mar 22, 2011
Messages
115
Location
With your sister, somewhere you wouldn't like
Go with the hub dynamo, there's nothing so relaxing as knowing that if you're stuck in the office late or it's raining and cloudy, that you have something to be seen by. The total DIY thing eludes me, I re-use existing light housings & reflectors and fill them with a simple circuit and LEDs. Aft I use the old dynamo mudguard light and on the pannier rack an old motor-bike rear light stuffed with red LEDs.
At the front I run 2 old-style bullet dynamo headlights with a 1W LED in each. It's not really total downhill, pitch-black forest trail bright, but it gets me across the London parks where there are no lights and big deer. If I was minded, I'd probably advance the current set-up to a 3W front light dialed down to 2W or so (just a bigger resistor) to give better throw and some low power white LEDs as standlights to be seen by once stopped.
I've generally found LDOs (Low Drop-Out regulators) to be the right balance between useful voltage regulation and simplicity of circuit design. There are loads of circuits to choose from, but I'l send you mine if you're interested.
 

minisystem

Newly Enlightened
Joined
Jul 12, 2011
Messages
96
My vote is for dynamo as well, but that's no surprise. The main issue I can think of with a DIY light is that there are no off-the-shelf asymmetric optics similar to what the high end commercial dynamo lights use (that I know of, anyway). Much ink has been spilled about how the cut-off optics produce a beam that is superior for road riding, so if getting the light out onto the road where you want it is a priority (as opposed to lighting up a trail or being visible in urban traffic), then a DIY solution may be too much of a compromise.
 

GlowBike

Newly Enlightened
Joined
Jan 22, 2003
Messages
24
Location
Portland OR
Thanks again for all the replies. You've all been very helpful. I'm still not quite at the point of being able to make a decision, but I've learned quite a bit more and I'm closer.

On Friday I stopped at Clever Cycles to check out their generator display. I spun the wheel as fast as I could with a gen-light aimed across the room, and compared the intensity of the beam with my XM-L based battery light. I tried both a B&M IQ Cyo R (they did not have the somewhat brighter IQ Cyo Sport) and a Schmidt Edelux light.

The result? Not even close. Ignoring the much broader (and admittedly too broad) beam of my battery light, even at the center of their smaller beams the gen lights weren't anywhere near as bright. In fact when my light was already shining on our target zone, you could barely tell when the generator light was on or off. Even on its medium setting (with about 1.1A going to the emitter) it was quite a bit brighter than the generator lights.

As I was leaving, unimpressed, the store guy suggested that I really ought to look at the Supernova E3 triple. It puts out an inefficient conical beam like my light, albeit a somewhat more focused one. With 3 XPG emitters at 500mA it is plausible for total output to be in the 400lm range -- about the same as my light puts out on medium. Since the beam is considerably tighter than my XM-L's, it might be just about as bright on the road as mine is at 100%.

BTW, as I have done with my battery-powered light, if I got the E3 Triple I would definitely install a shroud (and mount the light down at the fork crown) to protect the eyes of those I'm sharing the roads and paths with.

Anyway, given the existence of the E3 Triple, I'm reasonably confident that I can get a generator light to meet my needs. Now it comes down to money. Building the dual-emitter battery system I'd been contemplating would be "only" another $100 or so (beyond the $100 invested in my XM-L system). But for the E3 Triple and the added cost of the hub, I'm looking at about $300. In my current house/kid-poor financial situation that's a ton of money. Dang, I still have some thinking to do.
 
Last edited:

znomit

Enlightened
Joined
Aug 1, 2007
Messages
979
Location
New Zealand
On Friday I stopped at Clever Cycles to check out their generator display. I spun the wheel as fast as I could with a gen-light aimed across the room, and compared the intensity of the beam with my XM-L based battery light. I tried both a B&M IQ Cyo R (they did not have the somewhat brighter IQ Cyo Sport) and a Schmidt Edelux light.

The result? Not even close. Ignoring the much broader (and admittedly too broad) beam of my battery light, even at the center of their smaller beams the gen lights weren't anywhere near as bright. In fact when my light was already shining on our target zone, you could barely tell when the generator light was on or off. Even on its medium setting (with about 1.1A going to the emitter) it was quite a bit brighter than the generator lights.

On a dark road at speed the Cyo significantly outperforms any single XML based light.
It is difficult to hand spin the wheel fast enough to appreciate it though.
 

GlowBike

Newly Enlightened
Joined
Jan 22, 2003
Messages
24
Location
Portland OR
Good point that hand spinning the wheel may simply not be fast enough to achieve reasonable brightness. Which makes the in-store display kind of pointless because it makes the generator lights look pretty embarrassing, at least if you compare them to battery lights in the store. The lights I was looking at would have had to be 5-10x times brighter than what I saw in order to equal the performance of my XML light.

So basically my in-store test was fundamentally flawed. Maybe I'll go down to the MUP near me some night this week and watch the generator-equipped bikes go by, to get a better idea of real-world output. It is Portland, so I won't have to watch for too long. :D
 
Last edited:

Savvas

Enlightened
Joined
Jun 11, 2010
Messages
222
Glowbike,
I'd recommend the dynamo approach for sheer set-and-forget value, BUT I would definitely be careful how much money I spent! It doesn't have to be an expensive or all-or-nothing exercise. A decent Shimano hub can be bought for around $70 (from Bike24) and don't forget the Planet Bike dynamo light is pretty reasonably priced. I haven't used it but it seems to get reasonable reviews. I am currently using the Philips Saferide (2 X leds - 60 lux or something like that, a;so from Bike24) and can recommend it! It's more than adequate for commuter riding after hours - an excellent beam and plenty bright. I'd suggest - in evaluating light usefulness - you consider questions of 'what is enough light' and beam shape issues rather than simply 'what is the brightest light'. My Philips is extremely bright from the front even though it's output is still way below many battery lights. It provides excellent near field and distance vision, again even though it is far less bright than many battery lights. I'm no expert but it's obvious to me that there's a lot more to night time visibility than sheer measured output!

DIY is fun and doesn't have to be complicated. You don't even need a housing and big heatsink really. I built a 2 XR-E-led light for my brother's bike that is on all of the time, has a stand light that works and is based on a simple bit of aluminium angle and a few squirts of sealer. You can see him coming in daytime from 100s of metres away and even though it's a symmetric beam it's not bad at night either. In short it works to see AND be seen and cost only about $15 to make and is so lacking in bling-factor that no-one would even think of removing it (even if they had a spanner on them)!
I also own some very nice custom-made battery lights (DesignShine) but it's horses-for-courses and for commuting, dynamo hubs/led lights are extremely useful I reckon.
My 2 cents worth anyway!
Savvas
 

GlowBike

Newly Enlightened
Joined
Jan 22, 2003
Messages
24
Location
Portland OR
Well, I've decided. I'm going to go dynamo. Will be ordering the wheelset today. About time too, because my usual commuter bike has been out of commission for 6 weeks while I've been dithering on this decision.

I am at least confident that a dynamo light will provide enough light for fast riding in dry conditions (on its 40% setting my current light is more than enough), which I guess is enough to get me to pull the trigger. Whether or not it'll be enough for nights when the pavement is not only wet but it's actually raining hard (which at least doubles the light requirements) is another matter, but that's not most nights and I'll deal with it later.

Although I may easily just end up with a commercial light, I think I'll try some DIY experiments first, starting with something simple as Savvas suggests and just an emitter or two. Not going to put a bunch of money into that unless it looks really promising, but I have to admit, from everything I've read here, on pilom.com and elsewhere I've grown a bit curious about the possibilities. Looks like you guys will be hearing more from me over the next few months! Again, thanks for all the feedback. When 7.5+ out of 8 replies are positive, it paints a pretty convincing picture.

[oh boy, guess I'd better update my sig. That is about 8 years out of date.]
 

fyrstormer

Banned
Joined
Jul 24, 2009
Messages
6,617
Location
Maryland, Near DC, USA
Dynamo paired with a NiMH battery pack should be pretty much impossible to beat. You'd have power when you're stopped and charging when you're moving. You wouldn't even need a complex charger circuit, just hook up the dynamo and the battery pack in-parallel with each other, and feed the power from both into the light's driver.
 

Trevtrain

Enlightened
Joined
Sep 28, 2011
Messages
407
Location
Melbourne, Australia
Are you sure about this fyrstormer? From what I remember of my cycling setups, most generators were AC not DC. (Mind you, that was in the '80s :sssh:)
I built a bridge rectifier setup about 25 years ago (using Schottky diodes IIRC) to run/charge a 6V gel-cell for 2 halogen front lights.

Of course, I suppose with LED stuff, polarity is important so perhaps the dynamos now have changed as well. I probably should get down to a bike shop to check out the latest tech.

(But I'd still double check the whole AC/DC thing before wiring parallel to a battery pack. :))


Dynamo paired with a NiMH battery pack should be pretty much impossible to beat. You'd have power when you're stopped and charging when you're moving. You wouldn't even need a complex charger circuit, just hook up the dynamo and the battery pack in-parallel with each other, and feed the power from both into the light's driver.
 

Lynx_Arc

Flashaholic
Joined
Oct 1, 2004
Messages
11,212
Location
Tulsa,OK
Are you sure about this fyrstormer? From what I remember of my cycling setups, most generators were AC not DC. (Mind you, that was in the '80s :sssh:)
I built a bridge rectifier setup about 25 years ago (using Schottky diodes IIRC) to run/charge a 6V gel-cell for 2 halogen front lights.

Of course, I suppose with LED stuff, polarity is important so perhaps the dynamos now have changed as well. I probably should get down to a bike shop to check out the latest tech.

(But I'd still double check the whole AC/DC thing before wiring parallel to a battery pack. :))
Actually my first bike had an alternator "generator" and back then I believe most bike lights used generators that were DC based not AC based (in the mid to late 70s) I can believe after the 80s they all went to alternators or AC "generators" though.
 

Trevtrain

Enlightened
Joined
Sep 28, 2011
Messages
407
Location
Melbourne, Australia
Now I'm curious Lynx Arc (or nostalgic?) - a quick Google search just found a few pictures of the unit I had, along with some useful info.

It was a Sanyo Dynapower that drove off the tread rather than the sidewall and supposedly was rated at 6V 3W. (500mA) Slipped a little in wet conditions and no use at all if muddy but was "cutting edge" at the time. Less drag than the sidewall "bottle" model I had before then but apparently no regulation. Some people complained of blowing bulbs at high speed but as I was mostly a tourer it never happened for me :), even before I wired in the rectifier and 6V cell.

I'm not sure about the bottle model I had, but was definitely AC on the Sanyo Dynapower. The filament bulbs didn't care of course.

But we may be getting a bit too far OT now.

Looks like GlowBike has already made a decision but to answer the original question...... I'd personally probably just mount a good flashlight or two on the bike now and carry spare cells. There are some nice AA lights with good throw for sane night speeds and Eneloops are cheap and lightweight enough to carry plenty of spares. Then I can easily take off one of these lights for other tasks if I need to. I can appreciate that a specialist bike light will have a properly designed beam profile though. It also depends on whether you mind carrying extra gear in a backpack, pannier or handlebar bag.

But regarding how to power it - these days I'd fly in the face of most of the posters here and just stick to a battery pack rather than muck about for 500mA from a dyno. For week-long touring trips I'd carry spares and go with a solar panel option. If you haven't heard of it, this thread http://www.cpfmarketplace.com/mp/sh...Solar-Powered-amp-USB-Chargers-2-panels-added has some very interesting stuff. I have just ordered a foldable solar charging kit with USB charger for Li-Ion cells. I can also use it with a couple of Sanyo USB travel chargers for my Eneloops.

YMMV

Actually my first bike had an alternator "generator" and back then I believe most bike lights used generators that were DC based not AC based (in the mid to late 70s) I can believe after the 80s they all went to alternators or AC "generators" though.
 

Lynx_Arc

Flashaholic
Joined
Oct 1, 2004
Messages
11,212
Location
Tulsa,OK
Now I'm curious Lynx Arc (or nostalgic?) - a quick Google search just found a few pictures of the unit I had, along with some useful info.

It was a Sanyo Dynapower that drove off the tread rather than the sidewall and supposedly was rated at 6V 3W. (500mA) Slipped a little in wet conditions and no use at all if muddy but was "cutting edge" at the time. Less drag than the sidewall "bottle" model I had before then but apparently no regulation. Some people complained of blowing bulbs at high speed but as I was mostly a tourer it never happened for me :), even before I wired in the rectifier and 6V cell.

I'm not sure about the bottle model I had, but was definitely AC on the Sanyo Dynapower. The filament bulbs didn't care of course.

But we may be getting a bit too far OT now.

Looks like GlowBike has already made a decision but to answer the original question...... I'd personally probably just mount a good flashlight or two on the bike now and carry spare cells. There are some nice AA lights with good throw for sane night speeds and Eneloops are cheap and lightweight enough to carry plenty of spares. Then I can easily take off one of these lights for other tasks if I need to. I can appreciate that a specialist bike light will have a properly designed beam profile though. It also depends on whether you mind carrying extra gear in a backpack, pannier or handlebar bag.

But regarding how to power it - these days I'd fly in the face of most of the posters here and just stick to a battery pack rather than muck about for 500mA from a dyno. For week-long touring trips I'd carry spares and go with a solar panel option. If you haven't heard of it, this thread http://www.cpfmarketplace.com/mp/sh...Solar-Powered-amp-USB-Chargers-2-panels-added has some very interesting stuff. I have just ordered a foldable solar charging kit with USB charger for Li-Ion cells. I can also use it with a couple of Sanyo USB travel chargers for my Eneloops.

YMMV
I remember my generator making the claim it was more efficient because it was an alternator. I think that one reason it had a rectifier is the circuit to switch on/off the battery light source when the generator was running too slow to light up well.
 
Last edited:

Steve K

Flashlight Enthusiast
Joined
Jun 10, 2002
Messages
2,786
Location
Peoria, IL
Dynamo paired with a NiMH battery pack should be pretty much impossible to beat. You'd have power when you're stopped and charging when you're moving. You wouldn't even need a complex charger circuit, just hook up the dynamo and the battery pack in-parallel with each other, and feed the power from both into the light's driver.

this is one of those areas where YMMV definitely applies!

My first design was a basic arrangement like this with AA NiCads. The NiCads got overcharged and died an early death.

My second design had a very nifty pwm'ed regulator for charging the AA NiCads, and a regulator for driving the headlight. It didn't kill the battery, but there just wasn't adequate capacity in the winter. This was a result of it being very difficult to charge a nicad in very cold weather and the almost constant need to power the headlight (dark in morning and after work).

Changing from keeping an incandescent headlight illuminated to just keeping a group of 5mm yellow LEDs illuminated allowed me to vastly simplify the standlight design. It was the origin of the fairly simple arrangement that I use now, where a AA NiCad is charged and a small boost converter drives the standlight LED(s) when the AC from the dynamo is gone.

But.... for a beginner project like this, the basic arrangement of putting a supercap across one of the headlight's LEDs might be sufficient. There's a long thread about standlights that is a good resource if a more complicated/sophisticated design is desired.

And of course, I always recommend keeping a battery powered small headlight on the bike too, since you may have a wiring failure or a headlight failure. It might be useful for fixing the flat tire too...

regards,
Steve K.
 

GlowBike

Newly Enlightened
Joined
Jan 22, 2003
Messages
24
Location
Portland OR
I think I will KISS and not have a battery in the circuit. I do also have a PT EOS on my helmet as backup (and also an EDC light on my keychain), so I've always got a light source for roadside repairs if something also goes wrong with the genlight. Also, this is America: I'm doing battle with texting teens and cellphone-yakking parents in oversized SUVs, so of course I'm running blinkies too, and the front blinky can still function as a last-ditch get-me-home light if all else fails.

I think my initial DIY experiment will just be a well-focused Cree or two with a bridge rectifier and a capacitor or supercap -- basically pilom circuit 2, but possibly with a second emitter in series. Nothing more complicated than that, and I'll see how satisfied I am with the performance of that simple setup before adding more emitters and/or complexity. I don't even really have much need for a standlight, since I've also got the blinkies.

Then ... if I'm satisfied with my DIY system's promise, my ultimate system might have more emitters (a total of 3-4 white Crees in front, a well-diffused red one in back, and maybe even a well-diffused amber "running light" in front), with a bar-mounted switch to shunt out 2-3 of the emitters at lower speeds. Possibly also the tuning capacitor(s) suggested in circuits 5/6, but that's about as far as I'd take it. I can't see myself bothering with the complexity of a voltage doubler (especially since this is a 20" wheeled folding bike and will develop higher voltages at a given speed than a big-wheeled bike) or switching back and forth between two different circuits. Circuits 8-10 seem like diminishing returns to me, and circuit 12 makes my eyes glaze over to say the least.
 
Last edited:
Top