Aérobeautique club

Home

Who are we ?

Projects :

Contacts

Partners and links

Dixi'Bot Project

Hello and welcome to the project page Dixi'Bot

We will present our project Dixi'bot. The draft 2 students (Jérôme and Thibault) has the ultimate goal of building a robot plexiglass controlled by computer via wifi.
Several modules will be added as such as:

- The obstacle detection via a transmitter ultrasonic
- A camera module to sending images to the pc

The original robot that we will use, is a robot already underway, but that is far from finished! (For info, we are in June 2007)

Here is the robot recovered from high school
It loads the card with the PIC 16F876, a module wellman for radio transmission that will be replaced by our wireless module (EZL-80C).
The second map is the map of his generation, which generates 6 grace notes to astables montages and an amplifier sound.

The main work to do:

- Establishment of a program with C + + builder:
         - Sending and receiving frames per socket
         - Management of frames received
         - Development of a practical robot with the keyboard

- Management Module EZL-80C (liaison Series / wifi):
         - Cabling module
         - Understanding of the module
         - Configuration module via serial link (ip address, port ...)

- Monitoring of the programme PIC::
         - Get your own programme
         - Management of the receipt and chief engineer (speed and direction, meaning, ...)
         - Creation of a function to deliver data to the pc to reflect the state of robot

- Establishing a bridge card H:
         - Cables component L298
         - Test the proper functioning of the component with a motor

- Creation of the chassis of the robot via Catia final:
         - Logic on the form the type and materials Chassis
         - Design of a chassis and easy to assemble at the same time aesthetics

Test to send the fabric from the PC

First, we simulated the robot by another pc in the local network. Nous avons ensuite relevé la trame par un logiciel prévu à cet effet (WireShark). We then found the fabric by software designed for that purpose (WireShark).
After several hours of tests we have finally managed to send the fabric that we need.

Here is the frame sent:

Here are our issued 5 bytes (robot)
Click to enlarge

And here our program to run the robot (test version)
Click to enlarge

Preparation of the bridge H

We therefore used the L298 (structural diagram below).
This circuit integrates 2 H-bridge. Here is the table of truth that we have established:

IN1 / IN3	IN2 / IN4	Result
0	1	Engine running before
1	0	Engine astern
1	1	Bloquage engine (Brake)

	We apply our MLI ENA on pins and ENB which are pin validation of the circuit We connect SENSE A and B SENSE to ground because we did not use it, these pins are present 2 to measure current in the various engines.
We also added diodes "coasting" to protect transistors cuts supplies After turning off the circuit and test we noticed that only one bridge did, the problem was solved when we noticed that the masses were not connected to the interior and should therefore all connect. We have also observed that the circuit heated lot. So we decided to install a heater to prevent the destruction of the circuit.

Management module EZL-80C

The module wifi ezl-80c's IP address by default: 10.1.0.1 (it is always good to know given that the documentation is relatively short!)

But we can configure via wifi (only in infrastructure) and RS232 (infrastructure and ad-hoc).
The RS232 is thanks to a component to be a MAX232, it adapts TTL levels.
Here is the structural diagram:

	Schematic structural map for the configuration of the module ezl-80c
Be careful, stoke the +3.3 V module, the module wifi running with this tension. We encountered some problems with food. A feeding battery works perfectly unlike food sector does not walked.

But big problem the last couple of months. We fail to establish a stable wireless connection ... Because we have noticed that the card is wifi HS. Remplacemenet of the card and wireless connection established.

November 8, 2007, the robot works dixibot, rolling with control via wifi. Mit we have 70 hours to get to this part.

We are now on the "finishing", ie the design of the robot plexiglass via the CATIA and routing of the e-card via the software PROTEUS.