An adult human can produce about 100 Wh of mechanical power whilst cycling, which is a not inconsiderable amount if you can convert that to electricity with reasonable efficiency. In a recent article on EDN [T. K. Hareendran] goes over a few ways that you can turn the rotary motion of pedaling into usable electrical power.

A basic form of this is already widely deployed, in the form of a bicycle dynamo that is used to supply power to the front and rear lights. These typically put out something like 3 watts at 6 VAC, so with a simple bridge rectifier and some smoothing this can power a pretty bright LED or two. To get more out of it, you need to use a more capable generator, which can also be a brushed or brushless DC motor in a pinch, with ideally a flywheel in the whole contraption to balance out variations in the human power input.
As for the potential here, a commercial solution like the K-Tor Power Box 50 is specified for ‘greater than’ 50 Watt, with a nominal 12 VDC output. Its target market is emergency generators, with enough capability to keep phones, radios and flashlights charged. Considering the $435 asking price, there is probably quite a lot of DIY potential well within that price bracket, especially if you already have many of the requisite parts lying around.
Fortunately this is not a new idea, with us having covered using bicycles as well as gym equipment to generate electricity in the past.
That should be 100 Watts of power.
100 Watt-Hour is an amount of energy.
Came to verify this was the first comment, glad to see it is.
Energy is a relevant metric for a reason: humans get tired.
You can sustain 200 Watts of mechanical output for an hour before you get tired (NASA’s estimate for average “healthy men”), and counting the mechanical and electrical inefficiencies of a small pedal generator coupled to a conventional battery, it turns out roughly 100 Watt-hours of useful electricity.
In practice, a person riding a bike without breaking sweat would be producing around 75 Watts or less, so time and boredom becomes the limiting factor.
I think you mean Watts, not Watt hours…
for the author to have meant one thing or the other, they’d have to know the difference.
???
This is hackaday, not your local newspaper’s “tech” section. The author definitely knew.
That one went right over your head didn’t it?
And the phrase “not inconsiderable” was probably meant to be “considerable”.
(After all the boolean logic in college, the double negative just grates on my eyes.)
Using litotes is a well-known mechanism to highlight emphasis on a statement. Saying “considerable” in that situation carries a different connotation than “not inconsiderable”
“Ah ha ! A double negative leads to proof positive !” (Wadsworth the butler – Tim Curry in the movie Clue )
Ah, thank you for mentioning that film and brightening my morning. It’s a real comedic gem of the 80s.
For me the highlight was Mrs White (Madeline Kahn)’s incoherent monologue – “I hated her soooo much… Flames. Flames. Flames… from the side of my face… flames… burning.. heaving.. flames…” as the rest of the cast just stare, nonplussed. Apparently it was ad-libbed.
Yeah right!
Probably meant watts, power, but for a lot of people, watt-hours, energy, would also be accurate (like, after they had done 100 watt hours, that’s it for them for the day).
3 watts from Sanyo Dynapower powering a halogen bulb in a parabolic reflector was awesomely bright and completely reliable, no bridge rectifier. Didn’t need to smooth it, either: I’m pretty sure the nonlinear characteristics of a tungsten filament more than sufficed.
You were pretty tired when that 3 watts was noticeable.
I remembered those old wheel contact dynamos to be so horrifically inefficient that they absolutely WOULD wear you out during a ride – just to power a tiny halogen bulb. I wouldn’t be surprised if they were little more than crummy DC motors, or at least the ones we could afford when I was a kid. Anybody take an old one apart?
You are correct, about almost all of the above. They would easily suck 5-10% of your power, or more, given the horrid rubber-to-rubber spring loaded wheel interface, and the inefficient bulbs (they were typically incandescent rather than halogen though).
They were eithe ac or dc motors and would typically have no regulation to the bulb but be designed such that they would saturate before overpowering the bulb too much.
There are much much better options now, in the form of “dynamo hubs”. They are very popular in the cycle packing community (since people will often spend many consecutive days without access to power, but they can charge usb banks during the day and have lights at night). Sinewave Cycles (get it lol) makes quite a few that are well reviewed:
https://www.sinewavecycles.com/collections/dynamo-hubs
Nitpick: the hubs at sinewave are designed and manufactured by SON or shutter precision resp.
I am sure, many of us have. 🙂
Different to a DC motor they typically consist of a rotating permanent magnet and a stator coil. This way commuting / pickup contacts with their unavoidable wear can be avoided.
Low efficiency, but cheap and relatively robust.
Of course I took my bike dynamo apart! When I was a kid I took EVERYthing apart!
I was about to describe the internals to you, but this image sums it up pretty well.
https://www.homemade-circuits.com/wp-content/uploads/2019/03/Dynamo-internal-layout.png
Its a simple two-pole alternator. Mine was from an age before rare-earth magnets; Alnico was king. My bike lamps were incandescent.
mine had a flywheel inside with the motor, so it stayed lit for about 5 seconds after i stopped. did not like how it tore up my tires but it sure was bright!
ELI5 the purpose of op amp and MOSFET in the schematic?
It’s used to dump extra power to a dummy load (large resistor) so it doesn’t exceed the input maximums of the voltage regulator.
I didn’t RTFA but that’s my read on the schematic
guess you didn’t RTFA either lol:
“The core principle is straightforward: the raw input voltage (VT) is continuously monitored and compared against a stable reference voltage (VR). When VT exceeds VR, a power MOSFET activates the dump load, which must be capable of safely dissipating the excess energy.
Conversely, when VT falls below the reference, the dump load is deactivated. To prevent rapid switching near the threshold, it’s advisable to incorporate a small amount of hysteresis into the comparator circuit.”
It’s a really, really crap circuit tbh. I’m not even remotely surprised the dude proposing this in 2025 is “self taught”. In 2025 you should be using a modern brushless motor and some switch mode power staging…. I guess I’m a curmudgeon.
Its a linear regulator to bleed away excess voltage (and a lot of power).
I’ve only now checked the schematic, horrible idea, and the schematic is badly drawn too, bleh.
Pedal power is a lot like wind power, quite variable from moment to moment.
A much, much better solution to “excess voltage” would be to store up that excess in a super-capacitor. If the One Who Pedals finds out that some of that effort is just being dumped as heat just to keep the load happy, you’re going to have to encourage them a lot more to keep at it. Whereas if every now and again that extra effort is saved and put to work a bit later, that’s far more agreeable. You could even arrange some feedback to your Prime Mover so that they know when they’re over doing it, and help them try to keep an even output.
This is the most wildly inaccurate post I’ve seen in a while, and that’s saying something given the quality of articles here lately (given most of them are just a plug for some braindead ad riddled youtube/tiktok garbage – wtf happened?)
BTW, an “average” human generates 100W of heat sitting around pondering their existence, and far more while moving around vigorously trying to ignore it. This can be (and occasionally is) captured and reutilized, particularly when there are large dense gatherings of moving humans, for example SWG3 in Glasgow:
https://swg3.tv/explore/news/pioneering-bodyheat-system-switches-on/
1) Confusing watt hours with watts, that’s a high school freshman level mistake (and I’m being generous…)
2) Being very unaware of cyclist power output
a) Almost any reasonably fit person can put down more than 100W, and do it for >1 hour
b) Fit / athletic people can put down 250W-350W for several hours straight
c) Olympic athletes can put down 1600W+ in a sprint and keep up around 450W for hours at a time
I often think about modern gaming consoles / graphics cards and their power demands. Imagine if anyone who wanted to use 500W to play games, had to cycle for it hahaha. I think they might go back to PS2.
Bay area make faire had a cycle powered band thing many moons ago. I believe it was this crew:
https://rockthebike.com/event-services/pedal-powered-stage/
Hee’s a fun video about toasting with legs:
https://www.youtube.com/watch?v=S4O5voOCqAQ
Here’s some data on power output and racing class:
https://www.cyclinganalytics.com/blog/2018/06/how-does-your-cycling-power-output-compare
Maybe do a little research for a more interesting article, and definitely don’t mix up power and energy. That’s just elementary my dear watson (the misquote is part of the humor, you pedants out there)
Thanks for posting that. I’m a cyclist myself, quite aware of power-output levels, and was trying to think of how to write up this information, but you already did it. Most bike riders out there however put out very little compared to us cycling enthusiasts who’re obsessed with power.
A possibly more efficient way is to use a geared BLDC motor (for example, an ebike mid-mount motor) and an ESC configured to regeneratively brake it. The motor coils will act as a boost converter for the output voltage and the gearing provides high enough RPM to get reasonable efficiency.
I used a tiny alternator and a v belt to the rear rim of my bike in a stand. Connected the alternator through a bridge rectifier to the solar input of an ecoflow portable battery.
What I found is that at even modest charging rates (e.g. 50w), I really need a flywheel because the effort I can put into the pedals is so variable based on the vertical vs horizontal orientation of the pedals. Fun project though! Might make the improvements before winter inside billing season :)
Er, biking season :P
I found something similar: the setup is like a steam engine, where we have lots of torque at one point and at high power levels you end up slipping the tire at your max crank torque angle and having not enough power at the min torque angle. I can operate at a steady 215W for an hour on my bike on the road, but I could barely do 150 on the bike set up as a generator because of the spiky nature of my power output.
flywheels are good! I cannot ride a stationary bike without a big one.
non-circular chaingrings are actually pretty good for this too, see shimano’s ovoid chainrings marketed as “biopace” (https://en.wikipedia.org/wiki/Biopace), though nowhere near good enough to replace a flywheel in your stated application.
Sadly I wore out all my biopace chainrings (they stopped making them in 93…). Would be nice if someone started CNCing them.
I once visited a 1950’s civil defence post (built into a much older fortress but now upgraded to be gastight in case of chemical or faraway nuclear attacks).
The guide opened one door and there was a row of 5 stationary bikes. One of the tour guests (a bit of an upper class type) asked “Oh, is this the gym?”
The tour guide responded “In the ’50’s, life was the gym. This was the ventilation system, everyone would die without”.
The ventilation was purely mechanical, with an axle with 5 bike chain sprockets running underneath and attatching to the bike’s chains if they were coupled to the stands.
All the bikes were fully functional bikes, quickly detatchable to act as messenger bikes, but they found they would not fit through the gas-tight doors without first removing the front wheel.
There was also a hookup to the bikes dynamo’s but this was literally enough to power 10 bike lights in the various rooms, which was thought to be enough in case you were locked in — although they found it would take about 15 minutes to get used to it after you came in on a sunny day.
Or you could use an alternator off an old car, remove the rubber from the exercycle wheel and introduce an appropriately sized v-belt. Excess DC voltage charges 12v battery. Invert post battery.
yeah i think the battery is the most important part. then you could really run something off of it. get a little car-grade 400W inverter and i could even run my gas boiler and pump (less than 100W) for a while.
and it lets you ditch the power-waste stage of this design too
Old car alternators are quite problematic: the rectifier easily drops 1 .. 1.5 V, which is about 10% off efficiency at 12 V. Next is the V-belt, which introduces resistance when squeezed into the wedge by sliding in while compressed. The oftentimes very crude fan introduces a large amount of losses when spinning at the speed needed to generate the exciting current (which may be 2 .. 3 times the idle rpm of the motor it was on, compare the diameter of the wheels for the V-belt).
Newer alternators may have better efficiency at nominal load, but if the regulation circuit uses 20 W at 2 kW output, it is only 1% off efficiency, while at 60 W from the pedals its a third of what you put in.
(we tried that at school, then immediately replaced the alternator with a really old wiper motor after the first test)
I’ve been operating a generator bicycle like this for about 3.5 years. The “generator” is an old discarded brushless servomotor from work, which generates about 36V when run at 420 RPM, which is a comfortable pedaling speed at a 7:1 gear ratio. I charge supercaps with it and run a 36V inverter (also a discarded UPS).
The main thing I use it for is to watch Youtube/TV/movies, which I’d otherwise be sat on my arse for. With the setup I have, I don’t find it hard to generate 60Wh in an hour. I only run a laptop with it, but purposely use an old inefficient one so it draws more power.
Based on my measurements, I’ve generated 30kWH over the past 3 years, goal is 10kWH/year which isn’t a terribly intense regiment but better than sitting (and for which I’m embarrassingly behind on this year).
I did write it up a while ago, but I haven’t been updating it. Latest changes are that I went back to a chain drive for the whole thing because the belt it came with kept slipping.
https://hackaday.io/project/191731-practical-power-cycling
now that’s a hack!!
Hurray, and as a dedicated cyclist: I command you for the effort, sir! BUT… how much is 30kWh worth?
In my area, about 6 bucks. I haven’t done the exact math, but I’d estimate it’s around 150x more expensive than the power from the wall.
Human metabolic efficiency from food to crank is about 25% and if we assume 70% from crank to laptop, you need roughly 5.7 calories per calorie of output, or roughly 171 kWh worth of food.
That’s 147 067 kcal or approximately 129 Big Mac meals, at $11 per pop it would cost you $1419 or $47 per kWh to generate. That’s 430 times more expensive than the average electricity price in the US. With better food choices, I bet 150x would be achievable.
If you were eating just plain white rice for the extra energy you burn, you’d need about 113 kilos of rice and drop to around $7-8 per kWh which is much cheaper at only 70x the price of grid supply.
That’s not accounting for the amount of energy needed to cook the rice. At minimum you’d need around 0.5 kWh to cook a kilogram of rice, so you would use around 60 kWh of electricity to cook the rice to have enough food to pedal for 30 kWh of electricity.
And that’s why, if you want to minimize energy consumption in a society, it’s more efficient to give people scooters than bicycles.
The most efficient food to eat, in terms of requiring the least energy to prepare per kilocalorie, would be potatoes. For foods that require no cooking, rolled oats (granola) soaked in water comes out as pretty much the only feasible option. You could eat raw carrots or fruits to some extent, but you’d get digestion problems.
Though that doesn’t account for the fact that rolled oats are in fact steamed before they’re dried and compressed, so they too require a significant amount of energy to prepare. Just not by you.
Think of all the gym spin classes with bikes connected to generators…..free energy, fewer polluting power plants and healthier people.
Now connect a TV to that generator. No exercise? No Netflix.
Re: My earlier comment of spin classes as a (just kidding) free power source lead me one step further to when the AI Overlord Colossus/Guardian assigns us to daily pedal times to power its needs or face penalties is just one step before being plugged into the Matrix. We might want to rethink this AI thing….
As a human I can attest to this: this human can produce about 300W for a hour, or 250W for 4h. Forget 100Wh, make that 1kWh in a long session 😀
This seems pointless? A single solar panel that you plop on the ground will exceed this, and converting food calories to energy is not exactly efficient. Even if you religiously workout and this is therefore “free” energy (since you’re working out regardless) it just seems negligible? Even in a offgrid setting, the aforementioned solar panel + modest battery would be far more efficient
Nothing practical about the photo. A tire bearing deep into a small diameter shaft for traction, that’s hard on the tire too.
Are we that close to Soylent green?
I built a 350W generator not using a bicycle at all, but a foot crank driving a 400W 3 phase 220v 25:1 reduction gear induction motor. It will happily generate with 10uF capacitors across each phase. You need a double pole relay to drop in to connect the load, when the voltage is over 200vI, because it won’t start if the load is always connected . I used a flywheel from an air compressor weighing about 6lbs on the high speed shaft protruding from the back of the motor . Because of the flywheel, I added a sprag clutch on the crankshaft end to avoid damage to the gearbox if the crankshaft was suddenly jammed as well as damage to the operators foot if it slipped off the pedal. Since 350W is a lot, a second crankshaft can be added to the other end of the first, so two can operate
Waaaay back in the day, I found myself at the Glastonbury festival with my good friend Paddy. After accidentally swallowing a couple small scraps of paper with pretty pictures on them (la la la) we took to strolling about the immense grounds of the festival and happened upon a tent that featured a band playing something just right for our state of mind. The amplification was powered by a couple of bikes on stands. Noticing that the existing pedalers were looking a bit tired, we happily took over so the band could play on while they rested up a bit.
After a couple songs though, a new band took to the stage, and to my ergot-addled ears, they were awful. I remember having some very deep thoughts about whether it was acceptable for the power company to cut the electricity because they had an issue with the quality of the entertainment. I looked over at Paddy and i could tell he was struggling with the same dilemma. After a few moments of what could only have been lysergic telepathy, we both jumped off the bikes at the same instant. The music immediately sputtered to a stop, and half the people booed and the other half cheered. But we were already halfway down the hill and on to the next adventure!
them was the days!