I'm currently trying to build myself an electric bicycle with a regenerative braking system.

I am however stuck on the regenerative breaking part. The bike will include a simple DC motor, powered by two 12 Volts batteries connected in series. The motor will ALWAYS be connected to the wheel via a chain.

From there, there are 3 possible scenarios:

  • I want to accelerate using the motor, in which case the batteries will simply be connected to the motor, which in turn will rotate and apply some extra torque on the wheel and thus the bike will accelerate.

  • I don't want to accelrate nor brake (using the electric bicycle as any other normal man-powered bicycle), in which case the batteries will simply be disconnected from the motor via a switch.

  • I want to brake, using the regenerative breaking system.

Now this 3rd scenario is where I need help. The idea I have is that the motor will be driven by the wheel which will be rotating quite fast. The motor would then be used as a generator, used to re-charge the batteries. However, I have trouble figuring out how to do this/what circuit to use. Am i right in assuming that the generator (the motor driven by the wheel) will never produce more than the 24 V of the batteries, and I thus need to find a way to reduce the P.D. of the batteries, if I was to simply inverse the circuit?

  • 4
    \$\begingroup\$ You will probably not be happy with just connecting the batteries to the motor with a switch. This will cause an abrupt acceleration to maximum speed any time you turn on the switch. It would be more normal to have an electronic speed controller of some sort. So, since you need one anyway, just choose one that supports regenerative braking. Do a google search for E-bike speed controller or something similar. Also brushed DC motor speed controller. \$\endgroup\$ – mkeith Jul 30 '17 at 21:04
  • \$\begingroup\$ Am I correct in understanding that your motor will drive the wheel through a second sprocket? Does this sprocket have a freewheel mechanism? If so you can't regenerate. \$\endgroup\$ – Transistor Aug 1 '17 at 6:38
  • \$\begingroup\$ Thank you very much for your answer. I just ordered one! @mkeith \$\endgroup\$ – Cyril Gliner Aug 4 '17 at 17:38
  • \$\begingroup\$ @CyrilGliner, take a look at the actual posted answers and see if any of them deserve an upvote or could be accepted. That could help out people who took more time to answer your question, and also improves the statistics as far as percent answered questions. \$\endgroup\$ – mkeith Aug 4 '17 at 19:03

There is a tiny problem: regen braking for bicycles doesn't work very well.

I don't know what type of batteries you use. Since you mention 12V, they seem to be lead acid. So, two 12V SLA batteries in series. These have a capacity of, say, 50 Ah. I'm being pretty generous here.

SLA batteries charge at C/10 (roughly) and only if they are sufficiently discharged to accept this current. If they are almost topped up, then they won't accept any charge, so regen braking is not possible.

Anyway. Let's suppose you just climbed a mountain. At the top, your batteries are quite discharged, so they're ready to accept their full C/10 charging current.

C/10 is 5 amps, at 24V this is 120W.

You're going downhill fast. In fact, in a realistic downhill riding scenario, you'll get 1-2 kW from gravity. But you can only recover 120W from this to charge your batteries. So, it's pretty useless. Once you get back down at the bottom of the mountain, you'll have recovered a tiny bit of charge... but just a tiny bit.

Now, you could use modern Lithium cells which can take a much higher charge current. In that case, that could perhaps work, but you'll have to use a smart electronic charger, overdesign the motor to be able to use it as a generator, and go downhill slower, which is no fun.

If you want to use regen braking in the city between traffic lights, it's even worse. On the flat, most of the energy expended is used to push air. The amount that is recoverable at a traffic stop is negligible.

Note things are different for a 2 ton car. In this case it is worth to recover kinetic energy. Not so for a bicycle.

Point being, regen braking is only interesting on a bicycle if you can do it for free (both in terms of energy and money).

Using a motor with a freewheel, which will allow you to not have the motor drag all the time, even when not using it like when coasting, is much more likely to result in a better range than regen braking.

  • 3
    \$\begingroup\$ You can charge SLA batteries much faster than C/10 unless they are already almost fully charged. However your point is still valid. Even with lithium batteries, the maximum discharge current may be 5x or 6x higher than maximum charge current. This creates problems for low-range vehicles which wish to do regenerative braking. \$\endgroup\$ – mkeith Jul 31 '17 at 0:14
  • \$\begingroup\$ Yup, also you need a more powerful motor. The alternative would be to arrange the gears to go downhill just as slow as uphill in order to have time to charge and not require a kW motor... \$\endgroup\$ – bobflux Jul 31 '17 at 11:29
  • \$\begingroup\$ Yes, I don't know where you got the idea that sealed lead acid batteries can only be charged at C/10. Taking the Odyssey PC680 from my Lancair as a random example, it has a nominal 16 Ah capacity and if it's about 50% discharged it will happily accept a charge current of 30A or more. \$\endgroup\$ – pericynthion Sep 1 '17 at 4:36

I assume your DC motor has a permanent magnet for creating the field. For such a motor, regenerative braking is actually very simple.

  • When you put a fixed voltage from a battery to its terminals, the motor will accelerate your bike until it reaches its free running speed, which is tightly connected to this voltage.

  • When your bike runs faster than this (e.g. downhill), the motor produces a higher voltage at its terminals than the connected battery does, so the current direction inverses. This means the motor has turned into a generator automatically.

So, this works for the speed set by the battery voltage at the motor terminals. Regenerative braking to a lower speed requires having a lower voltage applied to the motor. A simple approach would involve a switch to change the series/parallel arrangement of individual battery cells so you have a wide selection of voltages and thus, free running speeds.

If you want to try that, please only use lead cells and don't use two 12V batteries, but individual 2V cells instead. Lead batteries are forgiving too high charging voltages, but 12V overvoltage is a bit much.

If you want to do it the modern way, you need a boost converter which turns e.g. a motor terminal voltage of 6V (low speed) into more than the battery voltage of 24V and allows current flow from motor/generator to battery this way.


There's a simple approximate rule you can use to get started thinking about most electric motors.

  • The voltage across the motor's terminals is proportional to the speed at which the motor is turning.
  • The current through the motor is proportional to the torque exerted by the motor.

This ignores various non-idealities, most notably the resistance of the motor windings, but is a good place to start when you're trying to figure out which way around the voltage and current are.

I want to accelerate using the motor, in which case the batteries will simply be connected to the motor, which in turn will rotate and apply some extra torque on the wheel and thus the bike will accelerate.

This is true, but you are going to want some control better than on-off. Otherwise, either it will be a very abrupt acceleration (lots of torque applied) or it will not supply enough power to be useful. Thus, you need a motor controller, which can have a “throttle” input for you to control by hand.

Am i right in assuming that the generator (the motor driven by the wheel) will never produce more than the 24 V of the batteries, and I thus need to find a way to reduce the P.D. of the batteries, if I was to simply inverse the circuit?

This is almost correct, but you need to clear up your thinking about it a bit. Follow the rule I described above: if the motor is turning faster, then you can conclude there is more voltage.

Let's say you're out riding your 24-volt-direct-connection bike, currently on a flat road in Ideal Physics Problem Land — then the voltage in the system will be 24 V and no current will be flowing.

Now, say you start going down a slope. Gravity will be accelerating you, and so the voltage at the motor will be increasing beyond 24 V. This reversed voltage difference causes current to flow in the other direction, charging the battery. Thus, the voltage does stay around 24 V, but only because the battery is acting to regulate it.

So you have in fact got the right general idea about regenerative braking — if you want to slow down below the 24-volt-speed, you need a converter (a motor controller, in fact) which makes it look like you have a battery voltage somewhere below 24 volts. That, by itself, is enough to cause braking.

A motor controller can be thought of as a variable power supply specialized for motors. It works much like a switching power supply, using PWM control and big capacitors. To perform regenerative braking, the controller does in fact (or can; there are various designs) connect the battery “backwards” (using an H-bridge circuit) for very brief intervals to the motor. The inductances (in the motor) and capacitances (in the controller) translate this into a smooth application of reverse torque and a smooth reduction in voltage and speed.

(If you were to wire the battery to the motor backward with a switch, you would still get braking — or more precisely, the motor would try to reverse direction as fast as it could, throwing you off and possibly burning out from stall current.)

High-power motor control is not an easy problem — you need big MOSFETs, big heat sinks, big gate drivers, big capacitors, and control firmware that doesn't make mistakes which either blow up the motor controller or, as I have mentioned several times already, throw you off the bike.

If you want a practical, rideable device, you should buy an off-the-shelf motor controller. If you want to understand power electronics and control systems, you should build your own. (And either way, this blog is a fun read mostly on the subject of Stuff With Electric Motors In It.)


A bicycle is typically driven by the rear wheel, and (in heavy braking) the main retarding force comes from the front wheel. So, the 'regenerative' braking will be only partly effective, unless you motorize TWO wheels.

Since the motor is expected to accelerate the bike, with the applied 24V, we ought to assume that the motor (as generator) will generate a back-emf somewhat less than the battery voltage, at normal speeds. So, to brake you need to attach the motor (now acting as a generator) to a boost converter that raises the DC voltage, and use that boosted voltage to charge the battery.

This means you want an unregulated DC/DC boost converter (a flyback type might be appropriate) that can be switched ON only when the brakes are needed. The boost converter input will be the motor/generator, and output (through a diode) will charge the battery during braking. By conservation of energy, all that charge power implies braking torque at the motor, on the rear wheel. The regenerative brake will have to sense the rider's hand pressure on one or both brakes (you do NOT want the rider to move his hands to a special switch).

Battery temperature and charge state may require inhibiting the braking current, on long downhills with a small battery. Intermittent braking with most battery types can exceed the 'normal' charging current without doing damage.


I would look at the question another way. When the motor is being used as a generator for regenerative braking, do remember that you asked about braking, and not regenerative recharging.

Unless you plan on blasting past speeding motor cars on the highway on your electric bicycle, your best speed is very likely low enough that 2 seconds at most of regen braking will bring your bicycle to a halt. Regen braking can be very effective on bicycles, in my practical experience of designing and building controllers for low-momentum, lightweight electric-powered carriers of one adult person.

Braking vs recharging: It's a bicycle. it's hardly got momentum to keep rolling under braking for any useful recharge to occur. If you wanted regenerative recharging then I envisage you having to pedal for long periods, using energy to both move your bicycle along and charge your battery to boot. There is not a lot of logic in doing that, IMHO. All you would have is a hot and sweaty battery charger.

The parameters of your question make it apparent that a simple answer is what is required. OK, here is simple:

1: Use unsealed lead-acid batteries if you have been advised that SLAs would be more finicky.

2: Get some nichrome wire that will handle the peak current that will occur with regenerative braking. Stick a low resistance of nichrome wire in series with the regen braking circuit to limit the current to what the spec. label on the motor says its normal current draw is. And, given a possible duty cycle of "on" for 2 seconds and "off" for a minute, heating and ventilation concerns over a maybe not-too-hot nichrome wire will be easy to address. EDIT: Oops, forgot. Better use a series diode with the nichrome resistor, otherwise if you switch to what you think is regenerative braking AND you are already going slow AND the "generator" is producing below 24V, then you won't have braking, you will have mild acceleration instead.

3: Why are you worrying about EMFs or PDs when the basic way to look after a LA battery on charge is to limit the current to the amounts already suggested in some answers and comments?

4: And yes, I agree that for high reliability in security, industrial, military and commercial situations, EMF across a LA battery is important and sometimes deliberately set to maximise certain subtle conditions of reliability for life, genuineness of charge, etc (eg. I used to do "equalising charges" in sensitive reliability environments on LA batteries, which required a specific voltage across the batteries).

5: And the best result, compared to getting extremely finicky with how to charge a LA battery, and just whacking some current which is within the C/10 or C/6 or whatever (depends what school you went to) limits, I suggest makes little difference to the overall life of a LA battery in a real-world situation, which never mimics laboratory bench test ideals anyway.

6: So how big a nichrome current limit resistor? Too many unknowns for us to answer here. Time to break out your multimeter, your diode(see above edit) and various lengths of nichrome wire, and find out yourself. And read the current draw on the motor spec. label.

7: Oh, BTW, never mind the EMF you measure on an open circuit motor used as a generator. In fact don't concern your self with EMFs even if if you haven't measured anything with your voltmeter. It's a 24V motor at the end of the day. It isn't built to run without a 24V source across its terminals. Whether a motor or a generator, it is always going to have the 24V across it - what's the problem if the current is kept within limits in consideration of both motor limits and battery current limit recommendations? Don't get complicated unnecessarily.

8: But if you do use any sort of active charge controller as suggested in other answers - and nothing wrong with that - make sure you put it on the battery side of the power switch, not the motor/generator side. That way you still don't care about peak EMFs on a motor running as a generator at speed.


Yeah you should reinvent the wheel while you're at it and 60 pounds of SLA batteries is a pretty good idea too.

Why do you think 99 % of anybody with an e-bike uses lithium ion or LiPo batteries. The brushless controller below is $29. It's also 1500 W 45 A, and 48 to 84 V range. It has regenerative breaking, regenerative charging, reverse, cruise control and it is also a smart controller that will configure itself to your phase angle and order. Not to mention LVC circuit protection and other gadgets you can use for displays or lighting.

You also have to ask yourself just how would reverse work on a bicycle? My average speed on my bike is 32 miles an hour. Sending an improper signal to reverse in attempt to the slow the motor down, would send you over the handlebars and shred your forks beyond repair.

Here's the link to that controller:

Along with some links to some forums that specialize In E bikes:

  • \$\begingroup\$ Second paragraph: "breaking" or "braking"? Your first sentence could do with improvement - is it supposed to be funny? Note that the question is two years old. Welcome to EE.SE. \$\endgroup\$ – Transistor May 19 '19 at 15:12

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.