Getting the most out of your hobby servo

Overview

ServoProject

by Adam Bäckström

Getting the most out of your hobby servo

Robot

Theory

The control system of a regular hobby servo looks something like this:

loop1

This control scheme is good at holding and moving fast between position, but not so good at following a smooth predefined motion. Which is what you want when you build a robot.

Industrial servo controllers use a cascade based control scheme instead, which looks something like this:

loop2

This is the control scheme used in this project.

The main benefit of this approach is that it also takes the velocity and torque into account. This allows for a much tighter motion tracking.

Backlash control

When dealing with cheap gearboxes, backlash is always a problem. The hacked servos in this project uses two encoders to compensate for backlash. One AS5048a magnetic encoder or a potentiometer on the output shaft and one custom optical encoder inside the DC-motor. The benefit of having the extra encoder inside the motor is getting higher resolution and a main control loop which is backlash free. If one would only use the AS5048a for controlling, the backlash would enter the control loop as a time delay. This limits the control loop performance.

The backlash compensation is done by moving the main control loops reference-position so that the output encoder reaches the correct position.

Project structure

ArduinoSketch

This folder contains the Arduino project for the Adafruit ItsyBitsy M0 Express boards.

When compiling, the active configuration is selected by modifying the #include "*.h" line in the config/config.h file to include the desired config file. The configSelector.py and the configurationWizard.py scripts can be used to to simplify the config selecting process by giving a drop-down list with all configs in the config folder.

The folder also holds a Makefile which can be used to compile and transfer the project. One benefit of using the Makefile is that the configSelector script always is executed prior to compiling. To be able to use the Makefile you must first install arduino-cli.

Setup

setupStep1

  1. Create the base configuration

setupStep2

  1. Make the new configuration the active configuration by selecting it in the drop-down and transfer the project to the microcontroller. Either with the Arduino IDE or the make script.

setupStep3

  1. Calibrate the optical encoder

setupStep4

  1. Calibrate the pwm nonlinearity compensation

setupStep5

  1. Generate the system model (auto tune control loops)

setupStep6

  1. Calibrate the output encoder nonlinearity compensation

MasterCommunication

Holds the example master project.

To compile run make. This creates the program ./executable with the following options.

Allowed options:
  --playPath            play the path defined in createPath()
  --output arg          data output file
  --simulate            simulate servos

Obsolete options:
  --servoNr arg         servo nr
  --recOpticalEncoder   record optical encoder data of given servo
  --recSystemIdentData  record system ident data of given servo
  --recMomentOfInertia  record moment of inertia data of given servo
  --amp                 amplitude for recMomentOfInertia
  --frq                 frequency for recMomentOfInertia
  --pwmAmp              pwm amplitude for recOpticalEncoder and recSystemIdentData
  --fricPwmAmp          pwm amplitude to overcome friction for recSystemIdentData

MasterCommunication/view.kst

Kst2 project file which can be used to view the recorder data in test.txt from --playPath option. Run

./executable --playPath --output test.txt

to start robot movement and update plots.

USB_To_Serial

Basic Arduino project to turn an arm based Arduino into a fast usb-to-serial-converter. The resulting communication delay becomes much lower than FTDI based converters. This is important since the master has to be able to send and receive from all six servos every 18 ms.

CadFiles

Holds all .stl and the original Freecad files

PcbDesignes

Holds all KiCad projects for the electronics

Dependencies

ArduinoSketch

ArduinoSketch/configSelector.py

  • Python >= 3.8.2
  • python3-gi

ArduinoSketch/configurationWizard.py

  • Python >= 3.8.2
  • matplotlib >= 3.2.1
  • numba
  • numpy >= 1.18.5
  • python3-gi
  • scipy >= 1.5.0
  • serial

ArduinoSketch/Makefile

  • GNU Make >= 4.2.1
  • arduino-cli

MasterCommunication/Makefile

  • GNU Make >= 4.2.1
  • gcc >= 9.3.0
  • boost >= 1.71.0
  • Eigen >= 3.4.0

Hardware

Parts list

Modified normal size servo

Cad and stl files for 3D printing:

CadFiles/Servo

Electrical schematic:

PcbDesignes/Servo

Modified micro servo

Cad and stl files for 3D printing:

CadFiles/MicroServo

Electrical schematic:

PcbDesignes/MicroServo

Gripper

  • 1 x SG90 servo

".stl" files for 3D printing:

CadFiles/Robot/stl/GripperBase.stl
CadFiles/Robot/stl/GripperClip.stl
CadFiles/Robot/stl/GripperFinger1.stl
CadFiles/Robot/stl/GripperFinger2.stl
CadFiles/Robot/stl/GripperMount.stl

Robot

Cad and stl files for 3D printing:

CadFiles/Robot

Normal size servo modification (Turnigy MG959)

Build MainPcb and AS5048aEncoderBoard

A video on how to build the MainPcb and AS5048aEncoderBoard can be found at https://youtu.be/YQpAAr5RPSE?t=60. This video is for an older version but the building instructions are the same.

Servo gearbox with 3D printed parts

PCB Assemblies

How to mount the PCBs to servo

Build the Optical Encoder

  1. Take the motor, use pliers to bend out the four indents in the metal housing and bend up the motor lid.
  1. Remove the two steel brushes from the lid carefully by cutting the lid into pieces. Take the 3D printed motor lid and insert the two steel brushes and two ITR8307 into the new lid. Solder the ITR8307 according to:
PcbDesignes/Servo/OpticalEncoder/OpticalEncoder.sch
  1. Take the motor with the lid removed. Paint the rotor with black nail polish and glue in the optical encoder wheel on top as shown in the middle image. The dimensions of the encoder wheel can be found in:
CadFiles/Servo/OpticalEncoderWheelDimensions.png

Then carefully put the new lid on the motor without bending the steel brushes.

  1. Solder the four leads from the optical encoder on top of ItsyBitsy board as shown in this image.

Micro servo modification (MG90S or SG90)

  1. Open the servo, desolder the motor and the potentiometer
  1. Bend down the motor terminals and remove the plastic back axis cover so the motor looks like the one in the last image
  1. Place one of the holes on a servo horn over the axis. Then use a screw and a vice to push out the motor axis about 1 mm so it looks like the motor in the last image.
  1. Then solder two wires to the motor terminals. Use super glue to mount the optical encoder wheel to the back axis and mount the 3D printed "OpticalEncoderSensorMount" part over the back of the motor.
  1. Place the two IR sensors next to each other on a piece of tape and put some hot glue over them. Fold pin 4 of the lower sensor like the middle picture shows. Then fold in the rest of the ground pins, cut them to length and solder them together.
  1. Place the four resistors next to each other on a piece of tape, in the order 6.8k, 120, 1.5k and 120 ohm from right to left. Put some hot glue over them to hold them together and remove the tape. Fold the top leg of the 6.8k resistor like the middle picture shows. Then fold in the rest of the 3.3V pins and solder them together.
  1. Connect the two 120 ohm resistors to pin 1 on the IR sensors and the 6.8k and 1.5k ohm resistors to pin 3.
  1. Solder three wires to the potentiometer.
  1. Connect the potentiometers 3.3V and ground to the optical encoder sensors while soldering on a 5 wire flat cable to the assembly. Then connect the third wire of the flat cable to the potentiometer sens wire and connect the fourth and fifth wires to the IR sensors third pin. This wires up the flat cable so that from right to left we have: 3.3V, GND, output shaft potentiometer, optical encoder channel A and lastly channel B.
  1. Glue the optical encoder sensors to the "OpticalEncoderSensorMount". Cut the screws to a length so that they fit with the servo lid removed.
  1. Put back the motor and the potentiometer into the servo. See "PcbDesignes/MicroServo/MainPcb/MainPcb.sch" for info on how to connect the microcontroller and motor driver.

License

Open Source License

ServoProject is free software. You can redistribute it and/or modify it under the terms of Creative Commons Attribution 3.0 United States License.

To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/us/

Comments
  • Accuracy achievable

    Accuracy achievable

    Hi Adam, An awesome project. I was looking greater accuracy in servos for my automatic turret. Can you please tell me how accurate are the modified servos in degrees ?

    opened by vkumarsinghnoida 9
  • No current sensor?

    No current sensor?

    What a great project! I noticed that previous versions had current sensors and corresponding control codes。 However, in the latest version, there is no current sensor on the circuit diagram. However, there is still a current sampling part in the code. Is that right?

    opened by pzh11001 6
  • Alternative boards and code

    Alternative boards and code

    Hi My questions are regarding how to make the servo without custom pcb as given in coreless servo build. How many boards are required to control 2 servos? How exactly do we connect the board and use which software ? And can we use the original potentiometer instead of the magnetic encoder ?

    opened by vkumarsinghnoida 5
  • Compatibility with DS3218 servos

    Compatibility with DS3218 servos

    Hi adam,

    First congratulations for your work, your results with theses cheap servos are really impressive

    i'm verry interested in building five of them but before checkout on JLCPB, i would like to have your opinion about the PCB compatibility with DS3218 servos.

    i think its ok because the only difference with 3225 servos seems to be height , and obviously the final torque.

    what do you think ?

    Regards,

    opened by PJfixer 2
  • Brushed servos

    Brushed servos

    For building the robot you mention to use 3x coreless modified servos + 3x micro modified servos. Brushed servos are not required? Finding ATSAMD21G18A-A at a reasonable price for a custom PCB is very difficult this days. So, can we build the robot just with off the shelf boards?

    opened by Telepresent 1
Releases(v3.1)
Generic automation framework for acceptance testing and RPA

Robot Framework Introduction Installation Example Usage Documentation Support and contact Contributing License Introduction Robot Framework is a gener

Robot Framework 7.7k Jan 07, 2023
Test python asyncio-based code with ease.

aiounittest Info The aiounittest is a helper library to ease of your pain (and boilerplate), when writing a test of the asynchronous code (asyncio). Y

Krzysztof Warunek 55 Oct 30, 2022
Python Webscraping using Selenium

Web Scraping with Python and Selenium The code shows how to do web scraping using Python and Selenium. We use as data the https://sbot.org.br/localize

Luís Miguel Massih Pereira 1 Dec 01, 2021
Voip Open Linear Testing Suite

VOLTS Voip Open Linear Tester Suite Functional tests for VoIP systems based on voip_patrol and docker 10'000 ft. view System is designed to run simple

Igor Olhovskiy 17 Dec 30, 2022
Load Testing ML Microservices for Robustness and Scalability

The demo is aimed at getting started with load testing a microservice before taking it to production. We use FastAPI microservice (to predict weather) and Locust to load test the service (locally or

Emmanuel Raj 13 Jul 05, 2022
Multi-asset backtesting framework. An intuitive API lets analysts try out their strategies right away

Multi-asset backtesting framework. An intuitive API lets analysts try out their strategies right away. Fast execution of profit-take/loss-cut orders is built-in. Seamless with Pandas.

Epymetheus 39 Jan 06, 2023
A Demo of Feishu automation testing framework

FeishuAutoTestDemo This is a automation testing framework which use Feishu as an example. Execute runner.py to run. Technology Web UI Test pytest + se

2 Aug 19, 2022
Airspeed Velocity: A simple Python benchmarking tool with web-based reporting

airspeed velocity airspeed velocity (asv) is a tool for benchmarking Python packages over their lifetime. It is primarily designed to benchmark a sing

745 Dec 28, 2022
Python tools for penetration testing

pyTools_PT python tools for penetration testing Please don't use these tool for illegal purposes. These tools is meant for penetration testing for leg

Gourab 1 Dec 01, 2021
This repository contains a testing script for nmigen-boards that tries to build blinky for all the platforms provided by nmigen-boards.

Introduction This repository contains a testing script for nmigen-boards that tries to build blinky for all the platforms provided by nmigen-boards.

S.J.R. van Schaik 4 Jul 23, 2022
Object factory for Django

Model Bakery: Smart fixtures for better tests Model Bakery offers you a smart way to create fixtures for testing in Django. With a simple and powerful

Model Bakers 632 Jan 08, 2023
A set of pytest fixtures to test Flask applications

pytest-flask An extension of pytest test runner which provides a set of useful tools to simplify testing and development of the Flask extensions and a

pytest-dev 433 Dec 23, 2022
Percy visual testing for Python Selenium

percy-selenium-python Percy visual testing for Python Selenium. Installation npm install @percy/cli: $ npm install --save-dev @percy/cli pip install P

Percy 9 Mar 24, 2022
The async ready version of the AniManga library created by centipede000.

Async-Animanga An Async/Aiohttp compatible library. Async-Animanga is an async ready web scraping library that returns Manga information from animepla

3 Sep 22, 2022
Load and performance benchmark tool

Yandex Tank Yandextank has been moved to Python 3. Latest stable release for Python 2 here. Yandex.Tank is an extensible open source load testing tool

Yandex 2.2k Jan 03, 2023
A Python Selenium library inspired by the Testing Library

Selenium Testing Library Slenium Testing Library (STL) is a Python library for Selenium inspired by Testing-Library. Dependencies Python 3.6, 3.7, 3.8

Anže Pečar 12 Dec 26, 2022
GitHub action for AppSweep Mobile Application Security Testing

GitHub action for AppSweep can be used to continuously integrate app scanning using AppSweep into your Android app build process

Guardsquare 14 Oct 06, 2022
A framework-agnostic library for testing ASGI web applications

async-asgi-testclient Async ASGI TestClient is a library for testing web applications that implements ASGI specification (version 2 and 3). The motiva

122 Nov 22, 2022
MultiPy lets you conveniently keep track of your python scripts for personal use or showcase by loading and grouping them into categories. It allows you to either run each script individually or together with just one click.

MultiPy About MultiPy is a graphical user interface built using Dear PyGui Python GUI Framework that lets you conveniently keep track of your python s

56 Oct 29, 2022
Just a small test with lists in cython

Test for lists in cython Algorithm create a list of 10^4 lists each with 10^4 floats values (namely: 0.1) - 2 nested for iterate each list and compute

Federico Simonetta 32 Jul 23, 2022