Micro-servo S90 hardware interface for RaspberryPi 3B+

The micro-servo S90 is a small, lightweight motorized device commonly used with Raspberry Pi or Arduino boards to control the precise movement of mechanical parts. It is compact, low-cost, and ideal for applications such as robotics, automation, and remote-controlled systems.

The primary objective of this repository is to develop and incorporate the hardware interface, enabling the utilization within the ROS2 (Robot Operating System) framework.

For a better understanding and simple control (outside ROS) of the micro servo, see my other github project S90_servo_motor, or looks at other examples.

Implementation

To control the servo position it has used the "ros2_control" framework, specifically implementing the forward position controller hardware interface.

In addition, it has been integrated the WiringPi library for communicate with the servo on the RaspberryPi.

Configuration and Setup

To enable the utilization of this repository, the micro-servo S90's PWM wire must be connected to PIN 5 (GPIO's pin enumeration, corresponding to physical PIN 29).

Used System

This repository has been designed for use with the following hardware and software:

• Raspberry Pi 3B+
• Ubuntu Server 22.04 LST
• ROS2 Humble

Installation

Once the hardware connection is ensured, access to the Raspberry (e.g., SSH) and:

• Clone the wiringPi repository, since the installation from apt may not work, and build it:

cd ~/
git clone https://github.com/WiringPi/WiringPi.git
cd ~/WiringPi && ./build

• Clone the servo repository both on your PC (to visualize the simulation) and Raspberry (to send actual signals):

cd ~/
git clone https://github.com/mataruzz/ROS2_servo_motion 

and install all the dependent libraries:

rosdep install --from-paths src --ignore-src -r -y

• Build the repository (limitating the number of threads) on the Raspberry:

cd ~/ROS2_servo_motion
colcon build --parallel-workers 2 --executor sequential

Run the example

In the following example, 5 positions are defined and iteratively passed to the controller.

Inside the RaspberryPi:

• Source the ws:

cd ~/ROS2_servo_motion
source install/setup.bash

• Run the controllers:

ros2 launch ros2_servo_motion S90_servo.launch.py

• Run the example:

ros2 launch ros2_servo_motion test_fordware_position_controller.launch.py

Another way to send to the controller the target position is to write directly on the controller topic:

ros2 topic pub /forward_position_controller/commands std_msgs/msg/Float64MultiArray "data:
- 3.14"

The above example script nothing does more than sending to the controller the desired position every second.

On your PC:

• Open Rviz:

rviz2 -d ~/ROS2_servo_motion/src/description/config/config.rviz

You will see the following model:

Expected result

If everything works as expected, you should be able to see the physical micro-servo S90 moving to 5 different positions ([0, 0.785, 1.57, 2.36, 3.14] rads), in loop. In addition, the movement will be also simulated in the RViz environment, as shown below:

Pros and Cons

To provide a clear and structured assessment, you can find pros and cons in the following table:

Pros	Cons
Hardware Integration: The capacity to effectively communicate with physical devices is demonstrated through the construction of a hardware interface for the micro servo S90.	Scalability Concerns: You can run into scalability problems that need to be handled if you integrate more hardware or devices into your project.
Raspberry Pi Compatibility: As a control platform, the Raspberry Pi 3B+ offers a variety of advantages, including as price, a strong community and ecosystem, and usability.	Limited Control Precision: It is difficult to achieve high precision when feedback mechanisms is missing. Due to the lack of a closed loop control system, the current configuration of the project is susceptible to disturbances and inaccuracies in servo positioning, which can restrict precision and robustness.
Open Source Framework (ROS2) compatibility: Embracing ROS2 simplifies integration with hardware interfaces and opens a framework that is modular, reusable, and reliable. As part of the ROS2 community, you have access to a vast network of experts, collaborators, and enthusiasts.	Lack of Trajectory Tracking: The lack of trajectory control means that the servo cannot smoothly follow a predefined path or perform complex motions. This is often important in applications that require dynamic and precise behavior.
Position Control: Achieving open-loop position control, even if simple and inaccurate, is an important part of advanced control technology and indicates progress toward project goals.	Presence of Jitter: A notable drawback is the presence of jitter caused by using an inaccurate square wave for control (generated by the RaspberryPi 3B+). This can lead to unwanted fluctuations and instability.

Future developement:

Looking to the future of this project, I would like to pursue several goals.

First, I want to enhance the functionality of the system by adding other servos and building a simple 2 degrees of freedom (2DOF) structure. This will provide me the ability to control a possible camera's movements in both the horizontal (right and left) and vertical (up and down) axes, which may be helpful for analyze the surrounding environment.

Additionally, I would like to explore other control methods, such as velocity control and, if practical, closed-loop control. Implementing these control approaches requires access to the necessary sensor data, such as potentiometer values, speed and time measurements.

Furthermore, I recognize the need to address system jitter, which can affect camera stability.