We want to convert the speed and power applied on the pedals of a bike to convert it to an digital signal. But we need to apply an variable load to simulate cycling up-/downhill.

My initial thought was to use an DC motor as generator to convert speed to an voltage we can read (with a power resistor to dissipate the power produced) and a second DC motor that would rotate counter clockwise to represent an load. But as I cannot test my theory I'm not sure if the second motor would just burn up as motors aren't supposed to be used like that. (I am well aware of the current peak the motor could use when trying to spin while halted.) Another option for the load would be using a pneumatic/electronic brake. But then we wouldn't be able to measure the power/torque.

I did find an similar question, but the situation isn't completely similar. As we plan on using the motor to continuously turn counter clockwise.

A simple nod in the right direction would suffice.


My initial thought was to use an DC motor as generator to convert speed to an voltage we can read(with an power resistor to dissipate the power produced)

Yes, go with that. If the generator is large enough, and the power resistor is large enough, they can dissipate all the power produced and there is no need for a second motor as a load.

Filament lamps make good power resistor loads. They dissipate a lot of power by being able to run hot, and they give a nice built-in indication of power! A useful side effect of their temperature dependent resistance is that they tend to a constant current load, meaning that the back torque they'll produce in the generator tends to stay more constant with varying speed than that produced by a fixed resistor.

There are several ways to vary the load with resistors, which range from switching several in and out, to using semiconductors in a linear way, to PWM'ing with semiconductors, to interposing a DC-DC converter between the generator and a fixed resistor.

The current drawn from the generator will be more or less proportional to the torque produced. The voltage from the generator will sag somewhat with the load current, it may be adequate for speed indication, or you could use another tacho, or small generator, or optical angle encoder, or an optical beam break, or a magnet and reed switch, there are just so many ways to get a speed indication.

  • \$\begingroup\$ Actually my first sketch was to power an lamp with the generator and measure the voltage and current produced by the generator. But I had my doubts about it. Reading the speed isn't really the issue as, like you said, there are multiple ways of measuring speed. But what I forgot to mention is that is has to be sturdy so optical encoders and reed switches might be to fragile(considering the people who will use it wont be very gentle with it). Also how does the current drawn from an generator effect the speed of the axle? As we need an physical load on the axle. To represent going up/down hill \$\endgroup\$ – DerekLF Mar 17 '20 at 11:00
  • 1
    \$\begingroup\$ The current drawn from the generator doesn't affect (not effect) the speed of the axle directly, it affects the torque that the rider has to pedal against. A higher torque load might cause the rider to slow down, like he was going up a hill, or he might just increase his power output to maintain the speed. Torque===current, and speed===voltage. The deviation from strict proportionality of torque and current is down to friction ( a constant torque), windage (a torque component rising as speed squared) and saturation (tendency to limit the torque at high currents) \$\endgroup\$ – Neil_UK Mar 17 '20 at 11:12
  • \$\begingroup\$ So what your saying is. If we were to get some (high )power resistors place them across the generator terminals it would affect (not effect, my bad) the torque needed to maintain its current speed. Allowing us to create an variable 'physical' load by switching the resistors in parallel or unconnect them. If we were to use the ideal motor. \$\endgroup\$ – DerekLF Mar 17 '20 at 11:36
  • 1
    \$\begingroup\$ Resistors are simple, robust and cheap, so ideal to use. The current flowing through the generator is essentially proportional to the torque the rider has to work against. For normal resistors, the current drawn is proportional to voltage, so if the speed varied, the torque would vary. Using filament bulbs flattens this curve, less torque variation with speed. Using a constant current load isolates torque control from speed. Varying the current drawn by switching resistors in and out is the simplest way of getting a variable physical load. No need for ideal motor, real motor will do. \$\endgroup\$ – Neil_UK Mar 17 '20 at 11:40
  • \$\begingroup\$ If we decided to use filament bulbs, how would we proceed on using them as an variable physical load? Using resistors would indeed be the simplest way. But using them, calculating and compensating the change in voltage when the torque is increased and visa versa(not to mention the temperature which could also affect the readings), is something we would want to avoid if we can. \$\endgroup\$ – DerekLF Mar 17 '20 at 12:15

Just a few thoughts to help you on your way:

  • Bicycle speedometers use a magnet on the front wheel to trigger a magnet-sensing switch (a "reed" switch) on the fork. By measuring the time between pulses and using the wheel circumference you can calculate the speed.
  • You can use the same sensor arrangement on the pedals if you want to know the "cadence" (number of revolutions per minute) of the rider.
  • You could investigate the various types of indoor cycling trainers. Some good ones will use an electrical load such as that you have described. I'm sure that you should be able to find a DIY "rolling road" type that will use an electrical load that you can measure.
  • \$\begingroup\$ (thanks for the edits ;) ) We thought about using the magnet method and/or an optic encoder. But we decided not to go with both because methods are rather fragile. (Considering the people who will use it..))And also, there aren't gonna be any wheels on it. just the pedals. (the prototype will be an home trainer). \$\endgroup\$ – DerekLF Mar 17 '20 at 10:52
  • \$\begingroup\$ Your English is very good. Your English teacher would like you to remember that we use "an" before a vowel - so it's "an electron" and "a proton". Have fun! \$\endgroup\$ – Transistor Mar 17 '20 at 10:54
  • \$\begingroup\$ You can magnetize one of the pedal cranks and detect its passing with a hall effect sensor for example. They can't break that... \$\endgroup\$ – Solar Mike Mar 17 '20 at 12:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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