Anyone has them? Couldn't find them online for my DIY dyno light project.

Are you talking about sensing if the tail light is operating? If so, all you need is a current detection circuit. Really easy with a small micro-controller (which I suspect the Luxos use) or very doable with old school analog.

I'd like to have the circuits to start off with something tried and true. At present I have just a simple bridge + capacitor + led.

Okay, what specifically do you want the circuit to do? What dyno, headlight, and taillight will you be using?

Dyno is Shimano DH-3D72. I'd like to make my own head&taillight, better with standlight feature. Intended use is commuting/road, at moderate (up to 30-35kph) speeds.

Okay, this should get you started:
https://pilom.com/BicycleElectronics/DynamoCircuits.htm
https://www.bikeforums.net/showthread...namo-diy-setup

Thanks, I'm using pilom's C2 with a single 3w LED and a taillight in series right now. I tried C8 but didn't find it much better than C2. Read also mrbubbles' setup, but there was some controversy there. I thought that Saferide/Luxos circuits were currently some of the best, that's why I wanted to know if there was something different about them.

You are unlikely to find actual Luxos/Saferide circuits online, they are proprietary to their respective companies.

Since you wanted a specific circuit, I am assuming your relatively inexperienced with electronics (at least the design aspects). If true I hope your not doing this to safe money on a dyno lighting set-up. Even an experienced person with a large junk box of parts is not likely to be able to create a decent tail light for less money than they are commercially available. Headlights are a better prospect, but it doesn't take too many mistakes/dead ends before you end up spending more than a good commercially available product.


If your doing this for fun/to learn, then disregard this post.

PlanoFuji, thanks for your advice. You are right, I'm no electronics expert, just can solder some components together. Actually I'm doing this for both reasons - I like DIY projects and thought it could save some money too. Maybe I'll reconsider it.

I'd go with martin (pilom) #7 ,8,9, or 10 circuit and wire a 5.5v 1uf supercapacitor and an appropriate value resistor between the led(s) that you want standlight on.

An extremely bright dynamo taillight diy:
$3 Cree XR-E led.
$5 supercapacitor for standlight.
$1 for optics.
$1-5 for wiring, resistor, housing.

You just built yourself a dynamo taillight that's brighter than any dynamo taillight* on the market with more side spill. *largely thanks to German regulation.

Hardly a decent tailight. For those prices (which don't include shipping, taxes, duplicates mistakes when soldering, etc... all of which add up fast), not waterproof. No connectors (or a means of passing those connectors through the case in a waterproof manner) to make removal from the bike. A a big miss is no heatsink, the CR-E needs to dissipate a lot of heat. And you conveniently ignore that tools are needed for soldering, prepping the case, etc.. All of this adds up to a great deal more than the $14 you listed (which is already nearly 50% of the cost of good retail tail light for dyno)

An excellent commerically available dyno tail light can be had for $30-$50. When discussing the tail light, it is highly unlikely someone who doesn't routinely tinker with such things and have a large collection of parts is going to beat the cost of the commercial unit. At least in anything that will work reliably for more than a short time. The head light has a little more room to be able to do so (though I haven't seen a single diy headlight that seemed decent)

If the OP wants to do it as a learning exercise or hobby, great. But it is unlikely they will end up spending less than they would for a commercial product, especially for a one off.

Oh, and I assure you the circuits on the pilom site (they are fairly inefficient) you reference are far less sophisticated than the ones the OP wanted (Luxos/Saferider).

I'm working under the assumption that you already have the tools (you did say an experienced person with a large junk box of parts).


This is a light I just recently finished.



This battery version cost $20, it's silicone sealed so it's waterproof, has connector. On medium mode it's much brighter than 5 of my superflash put together. For comparison sakes, the superflash is brighter than most "good retail taillight for dyno". I can get the same brightness out of this setup in a dynamo rig by removing the $4 pre-made regulated circuit and putting in a supercapacitor. Btw, shipping is free for these parts from Asia, and the electronic components I get from my local shops (who happens to charge the same as digikey, so I avoid shipping there as well).

The first one I've made 5 years ago linked below (the switch is different), still running strong. Cost $30 back then when the technology was quite new.

https://www.bikeforums.net/showthread...n-2?highlight=

(That threaded stated $17 because the red leds and circuit was all I had to buy, I already had the switch, optics, and housing)

If a person doesn't have any tools and just starting from scratch, yeah, that person shouldn't be doing this (unless they're really adamant about making something better than what's commercially available).

Efficient or not, they deliver WAY more in the real world compared to the Luxos/Saferide. Unless someone is willing to cook up a BETTER circuit that can drive 6+ leds at 1000 plus lumen with low mode of 500-600, I have to use what's available to my knowledge.

Let me remind you that the Luxos isn't as bright as the Magicshine, and the low mode of those circuits I referenced are as bright as the Magicshine or more when paired with the appropriate leds. You can argue beam patterns all you want (and the beam patterns are customizable by optics) but the users of 500+ lumens demand something just as bright from a dynamo as their battery setup, and very few commercially available lights match those requirements.

You'll need to define "decent". So far, the only thing that seems to fit your bill is if it's made commercially, which is funny because some DIYers have gone commercial.

I doubt your including everything required in the cost of parts for that $20 light. Like most diy's your likely making assumptions that people will have the 'miscellaneous' stuff lying around. As just one example, just how much do you pay for a tube of silicon sealant. The last time I purchased a tube it was $6+. And then most parts need to be mail ordered. Such companies typically have minimum orders. They also charge shipping and possibly sales tax. None of that ever gets included in the 'cost' of the device.

It isn't just tools that need to be included in the costs. For instance, if a project needs a single resistor, there aren't very many places that will sell just one. Many of the components and needed hardware are like that. As I said, if one is doing to learn or as a hobby then none of that matters. But if one simply wants to get a less expensive device, they are likely going to fail.

I seriously doubt your claims about superior performance, particularly as I posted in your thread; you didn't actually perform measurements of light output. And I suspect from your post you didn't perform measurements of actual power supplied to the LED's. Those LED's require much more sophisticated driving circuits to achieve 100 Lumens per watt. And the first indication is the 1000 Lumenclaim. The dyno's max at about 10W of power supplied. Your circuits are not supplying 10W of power to the LED's A fair bit of that power is being consumed by the inefficient circuit itself. Also the heatsinks your using are not likely keeping the devices as cool as they need to be to provide maximum output.

Yes, many battery lights produce more output than any dyno light is capable of. The reason is simple, batteries are capable of supplying more power than any bicycle dyno is ever going to be able to match. That is even more true when misrepresenting the dyno output by only considering its output at maximum power from the dyno (~10W) which only occures at speeds of about 16-20mph and up... If someone feels the need to light up a small section of their path like daylight, they need battery lights.

Heck, even the lumen claims of such lights are a farce. Like you, it is unlikely the manufacturers are actually measuring their light output. They are relying on the device data sheets, without regard to whether they are actually providing the environment those specs are reported around. In short those claims are marketing. The lux reporting by dyno lighting manufacturers is a much better yardstick. It is also much more straight forward for even a diy'er to produce. Those light meters are available for <$100 and that a test stand and a ruler, and even a diy'er can produce the data/graph of the light output and display pattern of their device. Show me one of those and I'll reconsider your claim that your device is superior to the Luxos, EDELUX, etc...

1) Reliable
2) Well designed from a safety perspective
3) Reliable
4) Efficient (I really don't want to waste any more of my effort than I need to)
5) Easily mounted securely in the proper locations.
6) Reliable

And those DIY's that have gone commerical have produced products that were much better implemented than their DIY prototypes.

The silicone was a leftover from my bathroom and kitchen chaulking. They are available at home depot. The cost was already sunk.

Mail order from Asian suppliers do not charge shipping or tax. A single cree xr-e red from dealextreme.com is $3.30 with shipping, everything is included .
If you can't find a local electronic hobby shop and have to buy everything from digikey, yeah it adds up.

Once again, from dx, or fasttech, free shipping, $7 for the cree leds, $2 for a pack of 5 optics, $4 for the driver, $3 for the cable, switch, and wires, maybe another $6 for a blinkie that's hacked up (that was actually free throwaway I found), plus miscellaneous (like silicone....). As for your doubt, you can do the shopping yourself online on fasttech or dx.com, you realized that experienced diyers don't question my estimate because they already know how much it cost (they can easily say BS if I'm wrong), whereas you, well, what I can say.

Resisters are sold in 10-20 package for a $1.


1000 lumens require me to be riding at 25km/h and up. Dynamo lights can supply more than 10w. I've ridden besides commercially available lights so I have visual clues for judgment.


No they don't need battery lights, a Supernova E3 triple can light up a bike path at 10mph, so can any light that pumps out 600 lumens at 10mph.

Once again, I've ridden besides commercially available lights so I have the necessary real life comparisons. And please, I'm not going to go out and buy a light meter to satisfy your ego when I know my light outshines my magicshine, and magicshine outshines the Luxos, Edelux. I've ridden besides an edelux at speed and I know how bright it is, I've yet to ride besides a Luxos but if the visual account from people who have 700 lumen lights are true, then my experience speaks for itself (you can disagree all you want, I don't care).

Check over at mtbr.com (which I'm guessing you haven't), their commercial products are identical to their prototypes, and every components (drivers, leds, optics) they buy are commercially to you and I. The housings for "commercial" looking products can be purchased as well, like the example below.

https://www.nicobaumgartner.ch/blog/?p=198


Everything on this light is a diy, the person didn't do anything machining, just soldering. I can build the same headlight by buying all the parts, (it cost around a $100 or less (including shipping) depending on the configuration you want, and it beats anything $100 on the market (other than chinese lights and flashlights). That light is 2 times smaller than your Luxos and 2 times brighter (depending on drivers), power by 3 cree xp-g, you made a point about overheating, but you can buy drivers with thermal dimming protection.


Battery light though, the best bang for your buck is still the chinese lights, no comparison. As for your preference to German law lights, German laws suck so I like my dynamo lights brighter than what's offered commercially (with a beam that's wider and throw farther). I built my latest dynamo setup for 60km/h+ downhills mountain road rides (and it might rain) so I want all the lights I can reasonably get.

So we get you to admit that at least one of those 'miscellaneous costs wasn't included in your price estimate. Given that you started at 50% of the $30 price of a commercial unit, it doesn't take too many of those oversights to prove my point that cost savings are unlikely when diy'ing a tail light.

As to fast tech, I have never ordered from them I have ordered from Futurelec before (another cheap over seas supplier). There were two issues; I always ordered at least 3 of any part I needed since those 'cheap' places had a tendency of shipping faulty parts, though that go better in recent years. And my last order from Futurelec took over 6 weeks for me to receive it.

FYI, I did my first DIY project in the late 60's. It was a home computer capable of performing sophisticated mathematical calculations. I was also from a poor family, so my parents ensured that I understood what the actual cost was for my projects, which was always more than the cost of just the actual parts in the final gadget...

10W was a fairly generous upper limit of the power available from a dyno hub on a bicycle. You (and the site you based your circuits on) clearly were not measuring power under load. Here is a typical power output curve for the SON series dynos (I would be surprised if any other brands were much better)



AS you can see the SON 28 maxes out at around 4W of power when measured under load. In short about 60% less than your estimated available power and even that is more than the LED's in your inefficient circuit are actually receiving. You do realize that by inefficient I am explaining that your circuit is wasting a lot of the available power? In other words your LEDs are receiving a fair bit less than 4W ( I would guess somewhere between 2 and 3W).

Human perception of light levels are notoriously imprecise, so your judgment is likely simply wrong. Compound that with the fact that your taking a commercial products claims that it is producing 600 lumens at face value. As I have said, and that can easily be verified if you desire, those claims are made by marketing people using the same flawed approach you did by relying on light output curves from the device specification sheets.

I have never actually seen a E3 in person. I doubt their and your claims for a couple of reasons. One, the power available from the hub (under load) is not as high as the led spec sheets would state as needed to provide that output. Two, their marketing is depending upon the flawed lumen output claim (which isn't measured) instead of the easier to measure lux ratings which would allow direct comparison with other dyno lights. Three, I have never found any independent measurement of the power output from the E3 when supplied by a dyno...

Yes, and real life comparisons of light level by humans is fatally flawed. (You do realize that your iris adjusts the ammount of light you perceive continuously?) It isn't my ego you would satisfy. I don't care what you claim you are getting, I know you are mistaken. If you actually are diy'ing to learn something the light meter is essential for you to understand what works better and what doesn't. Your perception is never going to do that.

A kit of parts made by someone else is not diy... Further, I suspect that the brighter claim is as false as I have demonstrated your claim is about your diy light.

Yes and that is another factor that REDUCES the light output below the curve you like to make your claims from...

You clearly don't understand those German laws. They place no limit on the power output from the light, physics does that. They specifiy a minimum illumination required (which was achievable by old fashioned halogen light bulbs) and the shape of the light. Any LED light will EXCEED the mimimum light output required by the German laws. The aspect of the laws I like is that it requires that the output from the light be shaped properly to avoid wasting much of the output as your diy and nearly all commercial battery lights do. And as I have explained, the physics is clear, none of your diy lights are brighter than any commercial light that uses the same emitter. Indeed it is likely the commercial product would be brighter since it would likely be using a more efficient electrical design that can deliver more of the dyno power to the actual emitter.

As to throw further, since your light is not producing the output you claim it is not possible that it is producing a further throw (which is dictate by light output) than any other commercial product with the same power input...

You clearly don't understand the basics if you believe you are getting significantly more power at 60 km/h than you do at 25 km/hr. Take a look at the graph above. Under load, the generator output power curve flattens out at just under 4W. You simply aren't getting the output power you believe you are. Measuring open circuit voltages and currents do not produce anything close to actual power output under load you and your source believe they do. And frankly I doubt from your posts that you have even performed those incredibly simple measurements and are instead relying on incorrect information provided by someone who doesn't appear to understand the issues any more than you do; ie, martin (pilom)

If you want the most light you can possibly get, load a couple of pounds (or more) of battery packs on your bike and use a battery set-up. It can supply far more power to the light than any dyno you will ever have on a bike.

You didn't read the link. You can't buy that entire kit of parts together from anybody, you have to source it from different vendors and put it together yourself. By your grain of logic, none of my diy can considered diy just because I bought everything like a kit and put together myself.


Under German law, the e3 triple is banned, that's enough assurance from me that very bright lights aren't welcome. I don't agree with their optic cut off non-sense either so there's that. I don't live in Germany and I don't give a damn about their silly bike light laws. You can tout the benefits of lights made to German specification but those who demand more left the barn door long ago.

If you think I believe I'm getting more lights at 60km/h than 25 km/h then you're mistaken, my light tops out at 28km/h, any faster I go it does not get any brighter, that's on paper and that's also proven in real world testing. I'm saying at 60km/h, I want a lot of light, and my setup provides that.

You really shouldn't say anything until you have experienced from vendors people mentioned, or at least read their opinions, or maybe browse the website. Especially doing the last part would've answered your question about minimum order amount (which is one, and the whole website has free shipping, yes, that means you can order a $1 item and they'll ship it to you for free).

Yes, there's negative cost saving (it cost you more money) if you're rigging up a diy using 5mm leds like those used in dyno taillights (I don't do that and I scoff at people on instructable.com who do that). But for a high powered one like a double or triple xr-e red? There are significant cost savings.

I've overdriven leds with an adjustable power supply for giggles and I know they dim when they overheat, and I made sure they're adequately heatsinked, it may be inadequate in your opinion but riding at the speed where they produce heat, the wind provides enough cooling.

You can argue theoretical limits all you want, except that reality is people who have tried these setups (including the person who drew up the circuits) have battery setups and independent power supplies to confirm that indeed the leds can be powered by dynamos to these brightness.

I don't think you know how to read that chart, that dynamo hub's load is specific to what the light they have attached is capable of driving. If you disagree, fine. You can go over to the forums and threads where they specialize in driving these leds using dynamos to 10w or more and tell them it's impossible.

https://www.candlepowerforums.com/vb/...php?86-Bicycle
https://www.candlepowerforums.com/vb/...ghlight=dynamo
https://forums.mtbr.com/lights-diy-do...em-627814.html
https://forums.mtbr.com/lights-diy-do...dh-804636.html
https://forums.mtbr.com/lights-diy-do...ht-755558.html

Read them all, they're all good sources from people who use high powered battery lights and dynamo lights extensively. Nobody in those forums were arguing it impossible to use dynamo to light up leds to those degrees like you. Nobody.

Personally, I don't like wasting time arguing what's possible or not, but I don't like misinformation. I'm just an end user reaping the benefits of knowledge sharing and have battery lights to compare my diy setups to. You're one of the extremely few to argue that dynamos cannot possibly ever put out 4w or 10w when that's already accomplished long ago, and I'm saying it's possible because I and others have made it possible when compared to high powered battery lights (which I don't use any more except on bikes without dynamos), if you keep insisting that this is impossible, well, there's not much to say to you.

As for human eye perception, if I see x-light is brighter than y-light, I'm going to say x-light is brighter than y-light, your lux/lumen non-sense doesn't mean anything to me at the end, they're just spec sheet used to gauge where I should start. So you read the spec sheet and tell me that dynamo lights cannot be brighter than a cree xml (example) driven at 10w while I say otherwise when I have the lights itself to compare to in real life.

And you are misunderstanding what I said. To me a kit is anything that is assembled from off the shelf parts, even if the assembler is the one who obtained those parts. And from the pictures of your lights, you may have started with off the shelf parts, but you have adapted and modified them (particularly housing and such).

Again, you clearly have never read the German regs, the e3 is banned not because it is 'too bright' but because of the optics. It is using the same conical optics that put light in places it shouldn't go, like the eyes of those coming from the opposite direction. In short it wastes a considerable portion (up to 40%) of the light it does generaten and does so by blinding oncoming folks. This is a rare case where a government regulation makes sense. The regs would allow a carbon arc lamp (FYI a light source nearly as bright as the sun itself) if the optics were to spec.

I though you were implying that. And I have no doubt that your setup provides a 'lot' of light. It is your claims to measured(or even estimated) values that I believe to be false.

I said nothing about your vendors, but about my experience with similar vendors. And I NEVER questioned your listed costs for the parts you did list. I pointed out that there were parts, tools, and 'miscellaneous' that you didn't include in the cost, which if added would easily cause the OP to spend more for your diy tail light than he would spend for a commercial product that would meet or exceed the performance of your diy product. And that doesn't even count additional costs needed by mistakes or other problems. As I said, it is extremely unlikely that someone would be able to produce a decent tail light for less than the cost of a commercial one IF ALL of the costs are calculated. You have yet to actually calculate ALL of the costs of the lights you displayed.

Given your refusal to perform even basic measurements of your setup I really doubt your devices are heatsinked to the level they need to be in order to meet the power output specifications upon which you are basing you false claim for light output--even assuming they were in fact getting the power required to meet those specifications in the first place. Indeed that is yet another easily checked proof. Do a search for the actual test setup used for those light production curves. Compare the size of the heat sinking used to what your using and see if you don't begin to understand that your LED's aren't receiving the power you think they are.

Those aren't theoretical limits on that chart. Those are actual measured performance for the typical device. As to those people you mention, I have a strong suspicion that their claims are as based in reality (as opposed to perceptions and mistaken beliefs) as yours clearly aren't...

Sorry but I don't have time to waste reading other peoples on a subject that you are saying they are making patently mistaken claims that they are obtaining power levels from dynos that are physically impossible. Especially when what you have shown me so far clearly indicates that such claims are based in power estimates from an unloaded (ie, one with nothing connected to it) dyno...

The dyno's chart I provided shows the MAXIMUM power curve for that device. changes to the load from that curve actually reduce the power available...

Not liking misinformation is why I am wasting my time trying to explain this to you. You don't need to trust me. Buy or borrow a light meter and a good multimeter. Do a little reading of a decent text book (for god's sake stop taking internet sites as gospel) and do a few measurements after learning how to do them properly. If your in the D/FW area I will be glad to arrange to meet with you to show you how to perform the measurements with my equipment.

The one site of yours that your circuits are based on and which provided 'data' that claimed powers of 10+W were possible from dynos is one I read. Taking the time to read it and check its results is why I stopped posting on your other thread with wild claims of light output levels. At first the writing and claims seemed plausible, but when I compared those charts of power versus speed with the specs published by the dyno manufacturers it became clear they weren't performing the measurements they claimed. It appears that they were performing voltage and current measurements with either no load or a very lightly loaded dyno output. Thats great but the power you get from those measurements bears no relationship to the power you get when you put a significant load, such as a high power led circuit on the dyno. The chart I provided show the maximum power output curve of those SON hubs. They are designed to reach maximum power with an output current of around 500ma. That is expected to be achieved with a load of about 12 ohm. Choosing a smaller load will increase the voltage and reduce the current; however, the power so obtained WILL BE LESS than the curve shown on that chart. That will also happen if one chooses a larger load, only the voltage will decrease and the current will increase, but the power (combination of the two) will still be less than that curve. In short the SON 28 hub maxes out at just under 4W UNDER LOAD. And if your friends are claiming otherwise, they simply aren't performing the measurements correctly (or most likely not at all just reading data sheets)...

You can place all the faith in your perceptions and beliefs you want. As I have attempted to explain to you, the basic physics demonstrate you are incorrect. You could easily demonstrate this for yourself with a simple and inexpensive measurement tool, but feel free to continue your delusions.


BTW, I asked in the Luxos thread what you had against Peter White and you never said. I really am curious.

Anyway, I am shutting the computer down and going away for a long weekend.

You're missing out on this discussion when you're not acquainting yourself with the knowledge others have openly shared and accomplished. Reading those links I posted takes a lot less time than all the replying you've done today, if you've done the research ahead of time, you wouldn't have replied to me at all today and just accepted what I have to say.

Just calculating the power required to light up 6 cree xp series at a forward voltage of 3.6v per led, 4w across 6 led is not much, you would think that folks who posted on these forums for years who also use battery setups would know how bright their light is but you come along with your luxos b and tell everyone who did these setups that they're wrong and their dynamo can't put out more than 4w and you adamantly say it's impossible. As I've said before, there's nothing to say to you.


As for the cost of my builds, I will listed the full cost.

Material:
Heatsink housing for my setup all cost around $0.50 from a local computer supplier.
LEDs for dynamo light #1 on my diy thread, $35 including shipping. ($30 was because I bought from ledsupply for a triple Cree XPG plus shipping, the Cree XR-E is $5 on dx.com)
Optics, $4.
Capacitors, resistors, switch, total $8. $2 for each pair of different capacitor denomination, 2 pairs, $1 for bulk resistors, $1.5 for switch, plus tax.

This was 4 years ago.


Dynamo LED light #2 .
Aluminum tubing in 4 feet length, $12, chopped to size, used 2.5 inches, the rest was used for other projects around the house.
LEDs from cutter including shipping $45, including optics.
Capacitors, resistors, switch, total $8. $2 for each pair of different capacitor denomination, 2 pairs, $1 for bulk resistors, $1.5 for switch, plus tax.
Special 7 strand cable, $8.


Ok, now let's just say I start from scratch and I had no tools and I will build the taillight I just posted.
$7 led, dx.com.
$1.5 switch, ebay.com.
$4 driver, dx.com.
$3 optics, dx.com
$20 ubiquitous chinese light batteries + charger.
$6 cheapo blinkie and mount hacked.
$2 connectors + cable.
$12 Aluminum housing from dynamo number 2, used 2.5 inches.
$6 silicone (if you insist)

Tools:
$20 hacksaw.
$8 soldering iron.
$3 solder.

Total cost of the project is about $95, $31 of it are tools, and most of the components can be reused for future projects. Now, the equivalent light like this on the market is the Dinotte which cost $190. (Yes, you can definitely argue Dinotte is better quality, you'll get some disagreements from me so don't go there)

Fortunately for me, most of the items I've listed have multiple uses (silicone, solder, hacksaw, iron) or free (blinkie) or I already have them from years ago, so to build another one, I just have to buy the following:

$7 led, dx.com.
$1.5 switch, ebay.com.
$4 driver, dx.com.
$3 optics, dx.com
$2 connector plus cable

$18. No shipping. No tax. If I were to modify this setup for dynamo rear light, just switch the driver for supercap and resistor, which cost another $2 more. All the numbers have been rounded UP.

I recommend this if you're trying to build a dual red led taillight that's extremely bright.

But, if you're just wanting to run a single led red taillight and you have no tools to do diy.

Buy one of these and add battery, charger, and mount. Sets you back $40.
https://www.ebay.com/itm/Ultrafire-WF...item4164d1c9c0

(I have one too, don't give me anymore of "you don't know how bright your diy is" because I have the lights to compare side by side, and no, it's not 200 lumens).

There, I just posted something far more informative than your endless "that amount of light is not possible from a dynamo" bickering.

It isn't 'knowledge' that others are sharing. As I said, from the one web site you previously supplied (and appears to be the go to site for designs) all of your claims and ostensibly those of the additional sites mentioned are based upon the flawed supposed analysis of those designs. I can't be any clearer; open circuit voltages and currents (ie, measured without any load) do not provide actual power measurements. In addition to the prior source I provided showing the power curve of a specific model dyno I came across a test at various speeds of a wide variety of dyno's in the August 2012 issue of Bicycle Quarterly. That article tested eleven different hubs from five different manufacturers and 9 different models. All of the results are consistent with the SON power graph, though they did get power levels of nearly 5 watts from one SP dyno model at speeds of 50 km/h (31 mph). So the claims by those who clearly don't know how to actually measure the power (a VERY SIMPLE process) and post claims of over twice what the dyno's are designed to produce are simply not credible.

Thank you for a more complete accounting, which confirms my statement. A commercial product will cost less than any DIY...

Just out of curiosity, if these devices are so reliable, why do you keep building more of them?

A straw man. I have never said "that amount of light is not possible from a dynamo" claim you say I am making. I am sure that as LED performance continues to improve we will eventually get the performance you and others claim ( and yet NEVER measured) to be receiving. What I said was that you (and others) are not getting the light output levels you claim to be getting since you are basing it upon assumption (rather than measurements) of the power your LED circuits are receiving and manufacturer claims of light output when receiving that power.

What you don't seem to realize is that those modern high performance LED's will illuminate quite brightly at power levels far below what the spec sheets use for those light output curves. If you had actually measured the light output you would have noticed the levels are far below what you claim.

Those circuits that were used to test the power curve does not go beyond that power ceiling, whereas circuits for some lights like pilom.com, Exposure Revo MK2, and Supernova E3 triple do, I'm going to take the side of the folks who designed latter.


In my experience (before those cheap chinese lights flooded the market), that's incorrect, I never started out buying the tools specifically for DIY.


Each bike gets a set. Every 3-4 years, the LED techs change, so I build another based on the new leds (not the circuit, just the leds get swapped). (There's a noticeable difference between XR-E and XP-E2/G2, especially when you put 4+ of them together, yes they're bright as 10w magicshines or brighter (when wired with 6), that puts your 5 watt theory to rest))

My estimates account for optic and circuit loss.

Well pilom, as I have repeatedly tried to explain to you, simply don't understand that they aren't getting the power they claim. They measured voltage when no device was connected and current with their device and assume that by multiplying the to they are getting the power that their device is drawing--ignoring that the manufacturers and independent testers (all cited above) claim that the dyno's max out at well below that claimed power output....

For the commercial manufacturers, I suspect you haven't actually seen them claim to be receiving the power you claim from a dyno. If you have please provide a link/reference.


Sorry, but I wasn't citing cheap chinese tail lights, but rather German (which are hardly any ones definition of cheap) ones. I have never seen a Chinese dyno light though I am sure they are made.


You have shown far more 'versions' than any upgrades "Every 3-4 years" would tend to support.

And despite your hysteric claims, the max output of dynos being in the 4-5w range is not my 'theory' It is based upon actual measurements performed by a number of qualified individuals. Not the imagined perceptions of someone who have never even bothered to actually measure their lights/dynos.

Given that you apparently can't even understand the manufacturers (dyno) data sheets I find it unlikely you were able to perform the complicated analysis (since you didn't perform any measurements, analysis is the only remaining option) required to account for such losses.

And of course, I wasn't talking about optic or circuit loss in the post of mine your responding to here. I was pointing out that the LED's will actually glow quite brightly at power levels far below those the spec sheets provide light output for. And your claims are clearly based upon those spec sheet light output levels.

Maybe you're not understanding it correctly. That's also entirely possible.

I would think reputable manufacturers like Supernova/Exposure knows what they're doing, if they claim 800 lumens out of 3/4 leds (which would require 8w+) and it only puts out 5w equivalent of light, that's serious false advertisement (that's what you're accusing them of). I suspect their lights put out maybe 700 or so lumens instead their claim of "800", but that still shows the dynamo can power the leds beyond 5w.


Only in your mind. You haven't been on this forum long enough. I have 3 dynamo DIY front lights (2 of them are posted). Every post on this forum made by a member is logged so when you can go back to my entire posting history and it will show that I've only posted 2 dynamo front lights.

My claims are well below the estimate of the spec sheets (which are reasonable benchmarks to go on and as usual you are more than welcome to disagree).

I came across a this test.
Dynamos are tested against (unoptimized) 12 & 24 ohm load.

https://translate.google.fi/translate...-test/&act=url