The Final Assembly

The final step of the assembly was to design and print a new version of the chassis. My original design did not factor in the expansion of the plastic as it it printed and so the holes to bolt the motors onto the chassis which too small. Also the structure of the mounting points were compromised structurally as they were only thin walls with no extra support. This new version takes into account the material expansion with widened holes and added strut support to hold the mounting face to the platform.

As you can see in the image below the motors and wheels were able to be attached with the appropriate support and clearance. 

Below show the assembled prototype. The three motors are powered by their own individual 9V battery. The Arduino is also taking power from it's own 9V battery. The Arduino and motor controllers are being held to the chassis by bolts. 

The Omnimac fully assembled and in the operational position. The wheel clearance, approximately 2 cm,  makes for an adequate rise above most smooth surfaces for traversal. The nuts holding the various circuit boards in place are visible from the top. 

The below video demonstrates the working condition of the Omnimac prototype.

 

As you can see by the video the prototype does not move in the desired fashion. This though is far from a failure. As a prototype it has shown me a great deal about how to approach a larger scale robotic platform. The main factors in the design that tripped up this prototype were the motors, batteries and wheels themselves. 

The motors did not possess enough torque  to rotate the wheels at a constant rate even when the speed of the rotation was manipulated. The batteries were chosen for their low weight, small physical footprint and their relative safety though this only amounted to under powered motors. Finally the wheels. As they were 3D printed with PLA plastic they possess very little traction. 

With all this very valid information I will be able to improve on my design to create a platform that will carry the rest of the robot itself.

Cuircuits and Wiring

I now move onto the circuitry of the project, one of the most important parts of the project. The motor controllers, XBee module and the motors themselves need to communicate effectively and be constructed in a discrete maner.

I start by collecting my parts. Two XBee series one modules, three L293D Motor controller H-Bridge Chips, three Kysan 1124101 stepper motors, matrix board and four 9V batteries plus 9V battery clips. I chose individual 9V batteries as an added safety feature. As this is essentially a prototype I just needed to test if the system would work together while being controlled with an XBee module. If I were to use any higher voltage or even just one power source for the motors and Arduino I run the risk of damaging the Arduino and other components. This is a low risk low cost method of prototyping a concept device.
 

 

I started by drawing up the circuit in Eagle so that I could work off the schematic. 

I then moved onto wiring up just one stepper motor on a breadboard. to confirm the circuitry worked with the XBee module. This allowed me to test out all of my components as some were second hand. After some variation I found the successful set up and was ready to continue with all my components in good working condition.

The wiring, I soon realized, was going to be difficult to manage due to the sheer amount of wires needed to run the L293D. I have since revised my PCB design for the final version (see later post). 

The soldering was not overtly difficult though I found that as the wires were in such close proximity there would be adverse effects. The metal core would start to melt the plastic cover and I had to replace a couple wires as the exposed metal would short the circuit.

This is the current working version of the circuitry. A revised version is in the works though I am happy that this version is working and I have been able to continue the programming of the Arduino. 

 

The Design and Fabrication

The design and fabrication of the Omnimac has progressed and started to take shape. I had to build the wheels which were 3D printed and also design and print a appropriate chassis for the Omnimac.

I started with the printing and building of the omni wheels used in the project. These where sourced from thingiverse.com. The design was by user cbiffle (http://www.thingiverse.com/thing:705). I took the original design and scaled it to the appropriate size. 

Cutting the axle for the roller wheels of the omni wheels.

The two halves of the omni separated. The bolts are for holding the motor mount.

 

The wheel fulling built with rollers installed.

Three wheels complete.

I then moved onto the chassis of the device.  This process took some time to design as the particular dimensions needed to be precise. I was able to use the data sheet of my motors to use the true measurements.

I designed the chassis in Fusion 360. This is the first iteration. The motors will be bolted to the three corners.

This is the first test print done for the chassis. I printed with PLA and as you can see I forgot to mark out the holes for the circuit boards. This is something I will address in the next version. 

The Project Begins

The project has begun. I am endeavoring to design and build a wirelessly controlled omnidirectional vehicle.

The omnidirectionality with be provided by omni wheels (see Fig 1). These wheels can not only produce lateral but also rotational movement. I plan to 3D print these wheels so that I can save costs and develop a wheel that is directly suited to my project.

Fig 1.

I have started to build a rudimentary motor controller to be used in tandem with an arduino (see Fig 2), this controller will be using the L329D. I will also be employing the use of Xbee radio transceivers.

Fig 2. Motor controller for speed of stepper motor.