velorider562
1 µW
- Joined
- Oct 28, 2019
- Messages
- 2
My buddy owns a bike shop and he recently gifted me a Faraday for my wife. My wife loves it because it looks like a bike and most importantly can now keep up on evening rides. From different reviews the Faraday has its share of problems but the implementation is really well executed. I spent the last few weeks reading posts here on Endless Sphere about others DIY eBike projects. Now that the eBike bug has bitten me I think it would be cool to convert on of our beach cruisers into a very simple eBike for errands and when friends come over. Since my other hobby is electronics I think it would be fun to create an open source controller. I was intrigued by the lack of a community supported controller. I found posts regarding reverse engineering controllers, and wire hacking with Cycle Analyst. So I ordered a few parts from Alibaba and while I waited, started designing a controller. By profession I’m a software developer in the electronic space and an avid cyclist, so this is the perfect project for my skills.
I’ve read many posts regarding driving brushless DC motors with different wave forums. Doing some google’ing I found a post on Digikey.com about driving brushless motor and details on some dedicated brushless dc driver chips. I figured regardless of how much time I put into trapezoidal, sinusoidal and field-oriented control software it would not perform better than a dedicated chip from Texas Instruments, Microchip, Allegro, or any of the other multi-billion dollar semiconductor companies. This also allows me to focus on the awesomeness aspect. So my design separates the main logic from the brushless DC driver. There are many advantages to this design, first being the volt and amp requirements can be beefed up for larger system while the control interface remains the same. The brushless driver board is a daughter board to the main control board. Currently, I’m using the commonly available JYQD DC Brushless Motor Controller commonly available on eBay and Amazon for about $15. The board is rated up to 500W when using a heat sink. If in the future more power is required, I or anyone, can design a different daughter board with more amp and heat dissipation capacity.
Using the daughter board allows me to focus on the aspects I think are lacking in the currently available controllers. First being the “bike” computer. I haven’t seen many eBikes that don’t already have mounts for smart phones and “bike” computer. I thought it would be much cooler to have bluetooth connection between the controller and phone so all the stuff currently displayed on the “bike” computer can be integrated into a smart phone app, plus a lot more. All important data is streamed to the phone using a messaging protocol, analogous to CANBus message, the system used by modern cars to connect all the different computer modules. This allows information like bike speed, hub speed, brake on/off, battery level, etc be streamed and displayed on your phone. Additionally the phone can send messages back to the controller for configuration settings and for features like disabling the motor, equivalent to a digital lock.
The current status is I completed the design of the board and am sending the design to OshPark to get 3 copies made. General turnaround time is a couple weeks so hopefully in a couple weeks I’ll have a working and functional controller. I posted some additional design details on the github page, https://github.com/jpetrocik/EloController . I have not posted any firmware yet since I have only tested the different circuit modules separately on prototype boards. Once the board arrives I’ll be able to stitch the firmware together and post on github.
My plan is to post updates here in hope others find it interesting and a place to contribute ideas. I’m sure there are some thinking why reinvent the wheel, but I don’t think I am. I moving things forward. Cycle Analyst is a great product but without true integration into the controller it’s limited, and with controller features coming out of factories overseas they aren’t very attentive to the DIY community needs. This is the first post and my immediate goals are to convert a beach cruiser to a simple eBike with a DIY controller with Android integration.
I’ve read many posts regarding driving brushless DC motors with different wave forums. Doing some google’ing I found a post on Digikey.com about driving brushless motor and details on some dedicated brushless dc driver chips. I figured regardless of how much time I put into trapezoidal, sinusoidal and field-oriented control software it would not perform better than a dedicated chip from Texas Instruments, Microchip, Allegro, or any of the other multi-billion dollar semiconductor companies. This also allows me to focus on the awesomeness aspect. So my design separates the main logic from the brushless DC driver. There are many advantages to this design, first being the volt and amp requirements can be beefed up for larger system while the control interface remains the same. The brushless driver board is a daughter board to the main control board. Currently, I’m using the commonly available JYQD DC Brushless Motor Controller commonly available on eBay and Amazon for about $15. The board is rated up to 500W when using a heat sink. If in the future more power is required, I or anyone, can design a different daughter board with more amp and heat dissipation capacity.
Using the daughter board allows me to focus on the aspects I think are lacking in the currently available controllers. First being the “bike” computer. I haven’t seen many eBikes that don’t already have mounts for smart phones and “bike” computer. I thought it would be much cooler to have bluetooth connection between the controller and phone so all the stuff currently displayed on the “bike” computer can be integrated into a smart phone app, plus a lot more. All important data is streamed to the phone using a messaging protocol, analogous to CANBus message, the system used by modern cars to connect all the different computer modules. This allows information like bike speed, hub speed, brake on/off, battery level, etc be streamed and displayed on your phone. Additionally the phone can send messages back to the controller for configuration settings and for features like disabling the motor, equivalent to a digital lock.
The current status is I completed the design of the board and am sending the design to OshPark to get 3 copies made. General turnaround time is a couple weeks so hopefully in a couple weeks I’ll have a working and functional controller. I posted some additional design details on the github page, https://github.com/jpetrocik/EloController . I have not posted any firmware yet since I have only tested the different circuit modules separately on prototype boards. Once the board arrives I’ll be able to stitch the firmware together and post on github.
My plan is to post updates here in hope others find it interesting and a place to contribute ideas. I’m sure there are some thinking why reinvent the wheel, but I don’t think I am. I moving things forward. Cycle Analyst is a great product but without true integration into the controller it’s limited, and with controller features coming out of factories overseas they aren’t very attentive to the DIY community needs. This is the first post and my immediate goals are to convert a beach cruiser to a simple eBike with a DIY controller with Android integration.
