Yet-another Unicycle Interface
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

About yui

Yet-another Unicycle Interface - yui


General communication with Matlab:

First speed test:


Yui is a unicycle project, probably one of many. Goal is to design a small, two-wheeled robot with as few electronic elements as possible and using only off-the-shelf components. And, of course, link it to Matlab for control.

Basic elements:

  • two micro servos (9g, continuous rotation)
  • wifi (esp8266 with at least 2 GPIOs)
  • lipo battery 3.7v
  • voltage regulator 3.3V for esp8266 (up/down regulator)

Other needed tools:

  • lipo charger
  • usb-uart communication


My current generation 0.1 of the robot is:

  • two micro servos 9g modified for continuous rotation are already pre-modified, meaning they have been unlocked (little bit of plastic cut out) and the potentiometer is disconnected from the motor/gear axis BUT still is used for control - I soldered two 2.2Kohm 0805 smd resistors in there and the speed regulation is rock-solid now

    problems: this is a 4.8V-6.0V servo, so driving it from a 3.7V battery will not output much powers

    note: yes, electronics can output only a 3.3V pwm signals and it will work

  • wifi - esp8266

now that's a little bugger - you need pins, the more the better

why? in theory esp8266 has ADC pin and several GPIO/PWM/SPI pins (can output a max of 3 PWM signals at a time), but not on all boards they are ready for usage

also, if you will buy a cheap 2$-one from China there is a good chance it it will be soldered incorrectly, upside-down or just in a bad manner

I am using NodeMCU firmware, so a board with a good amount of SPI flash is recommended

my choice: is good enough

  • lipo battery

    my initial choice was a 3.7V 1400mAh battery but for the final build I will use 1600mAh (no reason really, just easier to obtain) - my first experimental estimation of robot uptime on the battery when running in circles (using both motors AND all-the-time communication over wifi) is 6h50min

  • voltage step-up/down regulator

    I honestly don't know if it is really necessary, but I added it as a precaution

    I'm using because I will need a good 3.3V for the electronics - and since I am using a lipo battery I can even face 2.9V-4.2V voltage range (not recommended, but possible scenario) - thus a step up/down version

    and if the next generation of robots will get a lipo 7.4V upgrade - the regulator can be the same

Other tools I needed: