I want to research how easy it would be to use a small laptop device while powering it by pedaling on a stationary frame. (For example, using a tablet on an exercise bike, or using a netbook while pedaling at a desk.)

I already know that it is possible to run a laptop with pedal-power, because I have found several examples of this being done. [1] [2] I now want to construct such a setup for my academic study. I need to confirm exactly what I need before purchasing the equipment. I do not have a background in electrical engineering, so although I have lots of numbers I don't know what I need to calculate or consider to pick a feasible option.


  1. netbook: adapter's input: 100-240V 1.7A, output: 19 V 65 W [3.4 A]
  2. tablet: adapter's input: 100-240V 0.45A, output: 5.1 V 2.1 A [10.7 W]
  3. smartphone: 3.7V 4.44WH (all the information I have)

POSSIBLE ARCHITECTURES: Apart from the seating arrangement (bicycle, recumbent bicycle, desk) there seem to be several options for how to hook the equipment up:

  1. wheel > bicycle dynamo > regulator > device (powerful enough for the tablet or the netbook?)
  2. pedals > DC generator > DC-DC converter > regulator > device
  3. pedals > DC generator > charge controller > battery (e.g. lead acid) > inverter > device [3]

Additional data:

  • An adult can generate 50-150 W in the course of an hour's strenuous pedalling. I want my users to be pedalling comfortably rather than strenuously.
  • Any device will have its own battery, so could I eliminate the need to have a battery in the setup?
  • Is it a bad idea to hook the device up to the regulator directly in order to avoid using the device's own adapter?


I'm very confused about a way forward at this point (even if it seems like I'm not). What I am looking for most of all in answers is clarity and definitiveness.

Of these combinations of devices (1-3) and architectures (1-3), which will definitely work?

Which combinations should I strike from the list as they categorically will not?

The dynamo architecture looks good to me because it seems easier/cheaper to hook up - is it capable of powering all these devices, even the netbook?

  • 1
    \$\begingroup\$ "is it capable of powering all these devices, even the netbook?" - That all depends on the dynamo. If you get one that is powerful enough, then yes. If you get one that isn't, then no. As for which will work - again, if you get parts that are powerful enough, then all of them are capable of working. \$\endgroup\$
    – Majenko
    Commented Nov 26, 2011 at 13:02
  • \$\begingroup\$ @Majenko so you're saying I need specs for the parts too - can you direct me to a site that would have examples? I assume there is some kind of a limit as to how powerful/efficient these parts can be? is there special terminology that describes how powerful the parts are? thanks for the lead! \$\endgroup\$ Commented Nov 26, 2011 at 13:14
  • 1
    \$\begingroup\$ What's the purpose of this? Motivating the user to exercise? Powering a computer where there's no other electricity available? Other? \$\endgroup\$
    – starblue
    Commented Nov 26, 2011 at 16:13
  • \$\begingroup\$ @starblue the aim is to generate power off the grid for users in rural communities \$\endgroup\$ Commented Nov 26, 2011 at 16:58
  • \$\begingroup\$ I've added a video link to a setup that uses a bike and capacitors but no additional battery \$\endgroup\$ Commented Nov 26, 2011 at 16:59

4 Answers 4


Background: I used to design controllers for exercise machines. The machine drove a 3 phase alternator and a resistive load was applied to suit various criteria.

I have carried out load tests with a specific view to establishing how much power users can readily make over an extended period in order to power electronic equipment or charge batteries.

Consider a moderately fit person to be one who could walk briskly on a level surface for an hour and be tired but not utterly exhausted in the process. ie not super athletic in capabilities and not even "extremely fit" - but well above "couch potato" fitness.

Using a good quality alternator a moderately fit user can supply 50 Watts mean for an hour. This is a level at which you definitely know that you are exercising but it would be bearable for many.

The same moderately fit user could supply 100 Watts for an hour and be extremely tired.

if you were aiming at a frequent powering task, 50 Watts would be far preferable to 100 Watts.

The above assumes a good quality reasonable efficiency system. Ideally with minimal "cogging" or saliency - ie no slow speed jerkiness to drive as magnets approach and recede from coils. Some systems use generators with substantial gearing ratios using chain drive. These could be reasonable but usually aren't. Belt drive to a low saliency alternator can be made to work well.

Note that the claims made on the website provided tend to contradict themself. I'd estimate the mean output as shown on that graph as being 60 - 70 Watts range. Their graph:

enter image description here

But, they say:

  • NOTE: Someone who works out every day can put out between 200 and 300 Watts of power.

    Those who are not in so good shape can put out around 80 to 100 Watts of power as shown on the one hour workout graph above.

    And Finally - Someone who is a competitive cycler can put out up to 500 Watts!!

The 200-300 Watts claim is true but tat's very demanding.
500 Watts + is also true of top top athletes.

Far more in selected cases - such as
Gossamer Albatross - 1st man powered flight across the English Channel June 12th 1979. . !!! :-).
About 300 Watts continuous in still air and no turbulence. "Rises rapidly" with turbulence.
Another day at the office for Bryan Allen. (day job ) -

enter image description here

I can do 500 Watts for about 10 seconds, after which my legs turn to jelly and I'm utterly exhausted.

Your 3 options don't seem different enough to matter:

  • wheel > bicycle dynamo > regulator > device (powerful enough for the tablet or the netbook?)
  • pedals > DC generator > DC-DC converter > regulator > device
  • pedals > DC generator > charge controller > battery (e.g. lead acid) > inverter > device 2

Any of these could work well or be terrible if badly designed.

It's not clear what you mean by wheels / pedals.
Powering from a wheel rim as shown may work OK - and may be what you mean by "wheel". Main aim is low loss, no jerk or unevenness. Ideally a steadyish non cyclical load.

Some of these do look good - on site you referenced.

Best is an AC alternator - preferably 3 phase or more. As shown below, a 3 phase waveform does not drop to zero at any stage. Rectifying this and filtering produces an even smoother result. More than 3 phases gives an even better result but is rare. A DC generator is an AC alternator in which the the rectification is provided by a commutator and brushes - as the voltage levels change a new winding is selected by the commutator rotating a new winding contact under the brushes. While the principle is similar to using an alternator with diodes, with brushes you generally get more mechanical drag and losses than with an alternator. Most generator (DC) offerings will be DC moors being driven as generators - the two roles are interchangeable but a machines designed for one will tend to be less optimised for the other.

enter image description here

The rim-powered motor based dynamo which you cited is OK in principle but likely to be horrendously inefficient in practice due to mechanical issues. As a child the many rim dynamos that I saw were all terrible so I was surprised to see modern units which do this reasonably well - it's a matter of actually knowing what you are doing mechanically with shaft alignment, gearing, surface friction etc. A "dynamo" of sort will typically produce 1 to 10 Watts output. With an efficient system, you'd notice 10 Watts when cycling and would almost not notice 1 Watt except under very light load conditions or coasting.

Some form of battery storage is almost essential. A laptop or tablet etc will have its own unless removed. Towards the lower limit, a more than usually netbook of tablet may get 6 hours from 3 x 18650 LiIon cells. Say 3.5V mean x 2 amp hours x 3 = 21 Watt hours. So operating over 6 hours = 21/6 = 3.5 Watts. Big laptops may get as low as 2 hours with 6 or 8 or 9 or 12 cells so say worst case 20 to 30 Watts. The rate they will charge their battery packs at is not directly connected but obviously a bigger pack takes more energy. A larger charger may specify 19v x 5.5A = 110 Watts. A single 2 AH 18650 LiIon cell (as used in combination in most laptop batteries) requires about 10 Watts peak to charge (4.2V x 2A + some "headroom") so the above 110 Watts is about right for the very largest of battery packs (12 cells). This is for about the first 40 minutes of a charge cycle from fully "flat". After that the cells go into a constant voltage tapering current mode.

More later if needed. Sleep calls alas ..

  • \$\begingroup\$ legend! thanks for all the info so far, although there's a lot I'll have to look up :) this is the bicycle dynamo I saw instructables.com/id/… - I assume a hub dynamo would be more efficient? - no-one ever mentions powering something more powerful than a USB device with a dynamo, that's why I'm concerned about that \$\endgroup\$ Commented Nov 26, 2011 at 15:12
  • \$\begingroup\$ Needs to be edited to have the actual answer stand out a little more; awesome answer nonetheless :). \$\endgroup\$ Commented Nov 26, 2011 at 21:27
  • \$\begingroup\$ Answer still growing :-).- Will try and partition a bit better as it does :-). \$\endgroup\$
    – Russell McMahon
    Commented Nov 27, 2011 at 1:41

I'll probably get slammed for this, but.. you guys are all thinking like, well, electrical engineers. Why hasn't anyone even mentioned the mechanical advantages of large flywheels, gear ratios to driving generators, etc? Spin-down time on a flywheel vs. energy required to get it moving should definitely be a factor, since a human being is not a pure energy in-energy out machine. The Powertap on my TT bike says I can put out ~400 watts or so for an hour, during my best race of the year. In an out and out max short effort, I can hit 16-1700, and bigger guys than me break 2K regularly. However put me on my trainer with it's tiny flywheel (not a very PRO setup) and the spin down effect once I stop pedalling makes me forced to put out more effort to maintain the same speed. The solution to the problem isn't optimizing for a steady output, it's realizing that a constant output is near impossible with a human, and that a fresher human can output more power to get started than it takes to keep going once moving. Same as the stand up and go at a stop sign effect.

  • \$\begingroup\$ Some degree of flywheel is highly desirable. This needs really only be enough to store variations across a pedaling cycle so you get a smooth load regardless of where you are in the cycle - as you say, a simulator or trainer should feel much as a real loaded bike does. Many fail to do this. Having a larger flywheel does store some energy but it is hard to store enough to make it useful long term. The equipment I worked with probably typically stored a few hundred watt-seconds of energy - say about 0.5 "horsepower.seconds". \$\endgroup\$
    – Russell McMahon
    Commented Jan 9, 2012 at 17:13
  • \$\begingroup\$ I'd say that a flywheel of about 10 horsepower-seconds would be getting about as large as was practical in an exercise machine type arrangement and that would be well above what you'd usually see. \$\endgroup\$
    – Russell McMahon
    Commented Jan 9, 2012 at 17:13

I would probably go with number 3, with a slight modification.

Unless you are pedaling at a perfectly constant speed, the voltage will probably vary pretty widely. Varying higher than the regulated voltage shouldn't matter too much, but lower would as the regulator will need a minimum input voltage to operate correctly.
You could get a very wide input range switching regulator, but that still wouldn't allow for a complete stop (or near to)

EDIT - regarding the external battery, for some reason initially I was not considering the device battery. I think that you could do without but since you asked which of the 3 will definitely work, for the reasons mentioned below I would probably stick with it. The MIT link you gave mentions using a 12V battery and a cigarette lighter 12V -> 19.5V converter, this would be an easy way of providing the input voltage if you don't want to mess around with switching regulators.

So some storage of power is a good idea (although you don't actually mention anything about allowing for slowing/stopping pedaling)
You could use some large (super) capacitors (or even a largish flywheel) but at a reasonable wattage it won't give that much leeway, so I think a small battery is the way to go.
I would have the generator -> charge controller -> 12V battery -> boost (or buck depending on device) converter + filtering -> laptop/tablet input.
I wouldn't bother with using the laptop/tablet supply and inverter as it will be more inefficient than simply using a DC to DC converter. With adequate filtering (a couple of capacitors and inductor/ferrite bead) and transient protection (e.g a TVS - transient voltage suppressor) this should be fine.
I would probably choose the tablet, or a lower power laptop as 30-65W is a reasonable amount of power to keep up for a long time, whereas ~10W for the tablet would be much easier.

  • \$\begingroup\$ thanks for the advice, I understand a bit better now that the lower W devices will be easier to power for longer \$\endgroup\$ Commented Nov 26, 2011 at 15:19
  • \$\begingroup\$ would the device battery suffice for the slowing/stopping problem? \$\endgroup\$ Commented Nov 26, 2011 at 15:19
  • \$\begingroup\$ Yes, the device battery would keep the device powered, but I personally wouldn't be too keen on the input voltage varying too much as the devices battery charging circuit might not be too keen on it. It would probably work okay, but without knowledge of the device internals you can't be sure. For the addition of a small battery you can have peace of mind. I would try and provide whatever the device asks for - e.g. if it states an acceptable input voltage range input in the manual or on the label then assume that anything outside this is "not ideal" \$\endgroup\$
    – Oli Glaser
    Commented Nov 26, 2011 at 15:39

I can remember, as a kid, having dynamo lighting on my bike - and nearly getting run over when stopped at a junction because my lights went out!

Conceptually, yes, a pedal-powered generator (whether dynamo or DC/DC) can be used; and I'd most certainly go with option (3) otherwise you could be wasting power if generation exceeds demand, or losing power if demand exceeds generation.

In many respects, (3) maps with the current paradigm in electric vehicals, which are using a small engine as a range-extender generator, rather than as the motive power.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.