28 Feb 6
This small line follower robot, was designed to be easily built at home without any special equipment, and using a minimum number of mechanical parts. You wont need more than two small motors, two free wheels and a piece of pcb (to hold the micro-controller, the motors driver and the line sensor) and sure.. your soldering iron!
The main trick making this design simple and affordable, is that the robot’s chassis is actually the main board of the robot, where some supports for the wheels – also made of small parts of copper boards – are soldered to it. All the motors, and the skids are mounted on the main PCB. For an electronics hobbyist, PCB manufacturing is a skill that will be learnt sooner or later, so this design lets you use your experience in PCB manufacturing to design a high precision chassis for your robot.
In case you’re not familiar with line following algorithms, it is recommended that you read that tutorial about line tracking sensors and algorithms before reading this article.
Figure 1.A shows a 3D graphical representation of the robot, where different parts can be clearly identified according to the following table:
|1||The base of the robot, also the main PCB.|
|3||Free Wheel, shaped as a pulley|
|6||Pipe clamp use to hold the motors|
|7||Ni-Cd 7.2V battery pack|
|8||1200 rpm 6V motor|
It is clear that the drive train of this robot is differential type, meaning the two rear wheels are responsible of moving the robot forward and backward, but are also used to turn the robot in any required direction depending the difference of speed between the right and left wheels.
The first thing that need some explanation is the fact that there are only two wheels, Well, while not being the best thing to do, a caster wheel can sometimes be replaced with a skid, when the robot weight and size are not important, and when the robot is designed for indoor environment, where the robot can move on relatively smooth surfaces, where friction wont be a serious problem.
It may seem strange that the battery was placed on the top of the robot, and it is actually an important mistake, as a battery at that height totally destabilize the robot because it raises the center of gravity, increasing the moment of inertia. For more information about robot stability and moment of inertial read this tutorial. For this size of robot, a smaller li-ion battery, placed beneath the robot, would have given much better results.
Again, in figure 2, a graphical layout of the main PCB was used rather than a picture to make it is easier to differentiate between different parts of the PCB.
As you may have notices, the main board has a dual function: Electrical and mechanical. From the mechanical point of view, this boards is the chassis of the robot, where the motors, the wheels and the electronics are mounted. You can see in figure 2 that the holes to be used to fix the motors are present on the layout, as well as the holes to mount the front and read skids.
Using PCB layout software to design the chassis, as well as PCB techniques to manufacture it, gives a lot of accuracy which is very important for the mechanical system to work correctly. You can see that the line sensor is integrated in that same main board. It’s important that the line sensor be as far as possible from the drive wheels in a differential steering robot. This principle is explained in detail in this article about line tracking sensors and algorithms.
There are many kinds of materials from which the copper plated boards are made. Try to choose a relatively thick one for this chassis, to be able to bear the weight of the motors and the batteries, all concentrated in four points, where the screws are fixed.
The wheels in this design also have a dual function, they act as a wheel and as a pulley, with which power is transmitter from another smaller pulley using a rubber belt.
Those wheels were originally free wheels used in sliding doors and windows. they are small, cheap and can bear very important loads. They have been modified as shown in figure 3.A so that they can be fixed to the chassis using those 4mm standard screws. Note that the wheel is still free to rotate around the axe of the screw, so the only way to transmit power to that wheel will be though a belt directly mounted on it, as you shall see later.
You can also notice that the wheels are mounted on the chassis using small rectangular pieces of copper board welded the main board using a regular soldering iron, and where the center of the wheel is etched on it for maximum accuracy, this way, both the right and left wheels are at the exact same height. You can also notice that another small piece of PCB is added to cary any eventual shear stress on the main part holding the wheel. (see figure 3.B and 3.C)
Motors and power transmission
The motors, which are DC motors originally made for cassette players, are cylindrical and thus very difficult to mount and firmly fix to a chassis. So this unique technique was used, which is to use pipe clamps, originally used to mount water pipes all along the walls of buildings (see figure 4.A). Those pipe clamps are easily available for all the diameters you can imagine, at least you will easily find a pipe clamp whose diameter fits the diameter of your motor.
You can notice a small black plastic pulley fixed at the end of the motor’s shaft, which will be then used to transmit power to the wheels using a belt. This small pulley can be found from the same store where you can buy those motors, the rubber belts, as well as all kind of accessories of cassette players.
When the motors and the pipe clamps are assembled as shown in figure 4.A, they can finally be easily inserted in their place in the chassis (main board), then all you need is to add a rubber belt to obtain the transmission system shown in figure 4B.
This pulley / belt assembly acts exactly as as the gearbox added to a DC motors to reduce speed and increase torque.
Depending on the size of the belt you have, you can adjust its tension by adjusting the height of the motor itself, which can easily be done by changing the position of the nuts on the screws holding the motors to the PCB. The optimum tension in the belt can be easily found by trial and error.
Being powered from a 7.2V battery, the regulator U3 provides regulated 5V for the microcontroller and for the logic gates of the motor driver. You can add a capacitor between the output of the regulator and the ground to absorb the noise caused by the presence of motors in the system, but I didn’t use any, and didn’t face any problems regarding this issue.
When the switch SW1 is switched OFF, the battery can be charged using the jack J2.
The line sensor is composed of 4 cells, and is based on the IR emission/reception technique described in this tutorial. D1 to D4 are IR LEDs used as receivers, D9 to D12 are also IR LEDs, but used as emitters this time. The output of the line sensor is directly fed from the Op Amps to the microcontroller. Only two outputs are connected to the LEDs D7 and D8, giving a direct indication of the output of the sensor, making the calibration process very easy through the potentiometer R6. For more information about line sensors, check this tutorial specially dedicated to line tracking sensors and algorithms.
Figure 5.B shows the four emitter and four receiver LEDs at the front of the robot. Note that this is the optimal position of the line sensor, as you can see in the tutorial above about line sensors.
It is also clear that they are mounted on the copper side of the board, even through they are regular LEDs (not SMT type). The Leads of the LEDs are used to adjust the height of the sensor from the ground. 10 to 20 millimeters proved to be a fair height for the sensor to function properly.
The connections around the microcontroller are standard in most of our 8051 based projects, they are the crystal resonator along with the two decoupling capacitors, the debouncing circuit attached to the reset pin, and the ISP (In system programming). Upon switching on the robot, The software loaded on the microcontroller simply directs the robot to the line, using standard line following algorithms described in the following article. You can download the C code along with the HEX file to be loaded into the microcontroller at the end of this article.
The two motors of the robot are driven using the reliable L293D Motor driver IC, the motors are connected to the wire connections W3, W4, W5, and W6. Being controlled by the microcontroller, the speed of the motors can be easily adjusted using PWM pulses fed to the motor through the Enable PINs of the driver. Note that each channel has it’s own independent Enable PIN, making it very easy to control the speed of two different motors simultaneously.
REMINDER: Operating the L293D motor driver
Using the L293D motor driver, makes controlling a motor as simple as operating a buffer gate IC. It totally isolates the TTL logic inputs from the high current outputs.
Putting a logic 1 on the pin In1 will make Out1 pin go to Vpower (36 Volts MAX.), while a logic 0 will make it go to 0V.
Each couple of channels can be enabled and disabled using E1 and E2 pins. When disabled a channel provide a very high impedance (resistance) to the motor, exactly as if the motor wasn’t connected to the driver IC at all, which makes this feature very useful for PWM speed control.
Figure 5.C shows different ways to connect a motor to the IC.
One way is to use 2 channels to build a bi-directional motor driver, another way is to use 1 channel per motor, building a unidirectional driver. In this project, we will be using the four channels to drive the two motors in both directions. To get more specific information on this very useful IC, you can always download and inspect the datasheet
[Note: I use ExpressPCB(FREEWARE) to design the schematics and the PCB]