Onewheel Build Guide

Post your Onewheel build guides here!

Firmware is not complete, but found code for a self stabilizing robot with the same gyroscope. Was first planning on using a microbit, but it was to laggy. Thinking of dooing the pid controlling in the vesc6, and need to convert pwm signal to ppm to make the code work.

Parts list: 10 inch hub motor with tire: http://bit.do/10inchhub 12s 3p 18650 pack from used laptop cells vesc6: http://bit.do/vesc6 Arduino Gyro: http://bit.do/gyroscup Switch: http://bit.do/switchs Motor mounts: https://a360.co/2CQGT64 End piece: https://a360.co/2CUvRwZ

Bought aluminium profiles, length 30 inches. Drilled holed according to mounts, and mounted all the pieces. In depth video coming.

TODO: Shorten length to 20"-25". Finish code, make stronger motor mounts. Testing

51 Likes

I think you are like the only one

27 Likes

Doesn’t the Vesc6 ship with an IMU? Why use an Arduino?

1 Like

I would definitely build a onewheel but there are no good toturals lol.

7 Likes

There is no gyro in the vesc, it is possible to rewrite firmware for the vesc and alter the pcb layout to connect the gyro over i2c. It seems not worth the hazzle right now. I am working on theese smaller vesc6 that will be perfect for theese kinds of builds.49643205_2241519969298164_8509075540525711360_n

14 Likes

I could have sworn mpu9150 is in the bom

2 Likes

Yeah in the official bom. But even trampa vescs decide to leave it unpopulated iirc, well escapes deffo does.

And Flipsky vesc 6 is not really a vesc 6

2 Likes

Its in the BoM, Vedder had some troubles with it from the first version and decided it wasn’t worth the time at that moment. will probably be implemented in later versions.

1 Like

i exchange a couple of emails with @trampa a while back. he mentioned that the new version will include the TDK 410-MPU-9250

What voltage did you choose for the hubmotor to run 12cells? 48v 600watt?

Definitely populated in the new version that is in production as we speak.

3 Likes

I chose the lowest voltage possible, so i can current limit it in the vesc. It draws arround 1000w ish, going to push it harder

Which way did you feed the VESC?

E.g. there is a ramp up/down for PPM/PWM signals, also for ADC and the nunchuck (I2C). So maybe it is not the ardunio which is laggy but the VESC having ramps which are too slow for balancing tasks. Also there are the PID controllers running at the VESC. So I would try minimizing the ramping or use UART commands (those don’t have a ramp up /down deadtime) and adjust the PID controllers.

Are you using hall sensors or an encoder?

1 Like

Hall sensors, ty for the information about the inputs to the vesc il do more testing. I was sending data through ppm, but are unsure if i was pid tuning it in the vesc. What I ovserved was that when writing ppm to the vesc, the accelleration was jerking. Meaning that it stopped sending data when reading the gyro. To test this we added a delay between eatch time we sent the ppm signal and the wheel stopped eatch time. Meaning the ppm controll from the microbit wasent hardware controlled but rather software controlled, causing a huge problem. We might have done something wrong or miss understood something, but il do some more testing. Im kinda swamped with other projects atm, but il come back to this later

Sounds like a timer interrupt problem. I dont know the microbit but e.g. you cant use neopixel ws2812 leds together with servo/ppm control on arduino uno since one interrupts the other.

1 Like

Yeah, id love to build one. This is awesome :ok_hand:

2 Likes

so you can easaly run that motor with a vesc…?

The vesc is perfect for this!

What kind of witchcraft :woman_mage: is this?

image

Looks cool, what’s the main goal of this design? How much power can they handle?

6 Likes

So just to make sure you ordered the 24v version with 600watt? And you are runnig it on 12s (±50v) and it produces aprox. 1000watt? What are the benefits chooaing the 24v instead of the 48v or even 60v version?