Feasibility bike -> e-bike conversion


I want to convert my existing cargo bike into an electric bike.

Ok, this project assumes
a) that odrive can work sensorless (without encoder) => i think it’s been implemented last summer (https://www.youtube.com/watch?v=-T2dhQKIgUA)
b) that I manage to mount v-brake to change the braking system (http://www.identitibikes.com/FKITRBA/Rebate-V-adaptors) => i will do it anyway

This is my wonderfull Babboe City => http://babboecargobike.com/product/babboe-city-cargo-bike/

I use it to bring my kids to school in the morning, then go to work. and directly go back home once the day is over. That takes me about 50-55min in the morning and 35-40 min in the evening.

I was using a pulling cart with my regular bike before but my kids grew older and wouldn’t fit in anymore. I bought the non-electric version (~1500eur) a few years ago because i wanted to make some daily physical exercise too. However, the electric version of this Babboe bike is not just an update of the classic version, there is a different gear drive system, and the frame accommodates for a battery and different brakes too. it’s about 2500-3000 EUR.
I could sell my own and buy the e-bike version. But it would be little fun and i doubt i would find a buyer anyway.

I think it’s a great opportunity to use odrive for an outdoor project. Toying around the rear wheel is a no-go. Way too complex, I even struggled to change a flat tire. The less intrusive way to do it, is to swap the front wheel for an e-bike wheel. A bit like a the Evowheel (https://www.indiegogo.com/projects/evowheel-the-ultimate-e-bike-revolution-bike-electric#/), they even have a 20 inches version that would be a perfect match for my wheel (for 800 USD). But again, it’s no fun.

I was thinking of using

  1. odrive in speed mode and gimbal motor mode with a single motor
  2. an arduino for sending commands through uart interface + reading some potentiometer (manual throttling on the bike handle for speed) and probably a bluetooth interface to change the settings on the go from a simple smartphone app i would develop. I assume arduino could be powered from a VCC 5V pin somewhere on odrive (from the encoder pins maybe?).
  3. i then need to find a suitable motor (must fit in the 105mm available space. Could be this one => https://www.ebay.com/itm/Bafang-8Fun-36V-48V-350W-Geared-Hub-Motor-for-Ebike-Front-or-Rear-Wheel/162820743214?hash=item25e8df602e:m:m45JK9iNILdF389X1NwUWgA ) + a 20" wheel and make the spokes to connect with motor + find a suitable battery.
  4. find suitable water/dust proof connector & enclosure

What do you think? What kind of nasty surprise have i missed?


The upcoming ODrive firmware v0.4.0 will have support for hall sensors, so I would strongly suggest that you get a hub motor with hall sensors. This will improve startup torque significantly over sensorless. There are a bunch of sensored hub motors on Aliexpress.


thanks for your reply. i’m going forward regarding the braking system anyway. most likely i will have to change not only the front wheel but the front fork as well. we’ll see.

I have also been advised in terms of power to forget about 250W motors (too weak). 500W should work. Best case scenario would be to use… wait for it… a magicpie motor! surprise surprise! but if i do so, i must use a much wider tire to increase adherence of the wheel.
It should also be safer to use dedicate brake levers that embeds the sensors for disabling motors.

Now I’m left wondering…

How could I deliver an assistive solution (ie without the manual throttle = you must actually provide an effort for the motor to activate) with odrive 3.5 fw 0.4.0?


How could I deliver an assistive solution (ie without the manual throttle = you must actually provide an effort for the motor to activate) with odrive 3.5 fw 0.4.0?

=> any idea ?


Why do you want this? It’s much easier and nicer to have a thumb throttle?


one reason would be legal. where i live, in order for a e-bike to qualify as a bike, there are two conditions:

  1. there must not be any kind of such throttle, the engine must help as the user actually pedals
  2. any kind of assistance must be cut off above 25 km/h

failing one of those condtions makes the “vehicule” falling in the motorbike category. and then, it is legally mandatory to suscribe to a specific insurance contract. not having such an insurrance in case of an accident is a serious offense here.

another motivation – assuming that i manage to have everything (including battery) fitting inside rhe wheel – would be the simplicity of installation of such a system: prepare the wheel in advance, swap wheels. done in 30sec. corrolary= no need to change the braking handles with contactless sensors either to detect braking.

my thrid and last motivation would be to better understand the underlying logic of electric motors in general. such a system would require to detect forward acceleration (user is pedaling) and backward acceleration (user is braking), all the while providing energy when required. the last reason – knowledge and understanding – is the strongest

is it doable?


Any update/timeline on the release of the hall effect sensors?


Hopefully with the release next week.